On July 24, I attended Jason Calacanis’s Angel University in Tokyo.

As I’m sure everyone here knows, Jason is a well-known angel investor in Silicon Valley. He became especially famous for his investment in Uber, which reportedly returned 2,500 times his original investment. He was a self-proclaimed loudmouth known for his sharp tongue, but perhaps due to age—or having mellowed out—he had shelved that abrasive persona.

He frequently used startup and investment jargon. Fortunately, having read his book beforehand, I was able to understand most of what he said.

His message was surprisingly basic. But it was also extremely practical.

As I listened to him speak, I found myself wondering when I had first become involved with Silicon Valley and the global startup community.

One turning point came in 2011.

That year, I reconnected with Allen Miner, the first representative director at Oracle Japan, and the co-founder & CEO of SunBridge for the first time in two and a half years. That reunion led me to start working at SunBridge—and to begin traveling regularly between Tokyo and Silicon Valley.

From there, my work and relationships expanded in directions I could never have anticipated.

I met Keith Teare, who was effectively a co-founder of TechCrunch, and he became a good friend and my mentor with whom I have now worked for 15 years. Even today, whenever I take my students from Musashino University’s Faculty of Entrepreneurship to Silicon Valley, Keith generously makes time to speak for them.

I invested in Infarm, a Berlin-born startup, launched its Japanese subsidiary, and ran its operations in Japan for about five years, including the preparation period. I also helped secure an investment in Infarm’s parent company from JR East. Infarm became a unicorn—and ultimately went bankrupt. Through the same company, I experienced both extraordinary success and extraordinary failure.

In San Francisco, I discovered and invested in AnyRoad. Andreessen Horowitz later led its Series A round, followed by BlackRock leading its Series B.

In Japan, I invested in Peatix’s seed round and developed a relationship of trust with its founder, Taku Harada, that went far beyond the conventional relationship between an investor and a founder. I also invested in WHILL at an early stage. And I saw several other portfolio companies—including TeamSpirit, giftee, SHANON, and Betrend—go public.

Looking at this list, it might seem as though everything had been planned from the beginning.

But that was not the case at all.

My reunion with Allen led to another encounter, which then led to another opportunity. Looking back, the events form a single line. But at each point along the way, I had no idea what lay ahead.

It was exactly what Steve Jobs described in his famous Stanford commencement speech as “connecting the dots.”

In my case, however, I did not connect those dots intentionally. They simply connected themselves. There was no strategy behind it.

Some time ago, when I was struggling to decide what to do next in my career, I asked ChatGPT for advice. It said something that stayed with me:

“If the 25-year-old version of you could see who you are today, wouldn’t he be quite surprised?”

I realized it was right.

I was born and raised in Japan and have never lived abroad. I still live in Tokyo. In fact, I come from Koriyama, a typical regional city in Fukushima Prefecture.

When I was 25, I could never have imagined that I would one day work with entrepreneurs and investors in Silicon Valley and Europe, invest in startups overseas, or run the Japanese subsidiary of a foreign company.

There will always be people who have achieved far more. Still, I think I have been fortunate enough to live a remarkably interesting life.

And recently, I have found myself returning to one question:

Could I turn the serendipity that expanded my own life into some kind of system for the next generation?

For many years, I have been preoccupied with the question of how Japan can produce startups that grow successfully on the global stage.

In Japan, it is possible to conduct both everyday life and business entirely in Japanese. The domestic market is still relatively large. A company can be built with its sights set only on Japan and still grow to a considerable scale.

That is one of Japan’s strengths. At the same time, however, it weakens the imperative for Japanese entrepreneurs to venture out into the world.

With that issue in mind, I established a company in Singapore called Go Global Catalyst (GGC) to create a “co-founder matching platform” for Japanese and non-Japanese entrepreneurs.

One of the models for GGC was Entrepreneur First (EF), which originated in London. I was introduced to EF’s co-founders, Alice and Matt, by Chris Wade, who runs Isomer Capital, a venture capital fund-of-funds. At the time, Chris was working on a task force organized by UK Trade & Investment to help establish London as a global startup hub.

Unfortunately, GGC did not succeed. We dissolved the Singapore entity, while the Japanese company remains dormant. I plan to close it when the time is right.

In Europe, people move across borders, and it is relatively natural for people of different nationalities to work together. It is also rare for university graduates not to speak English.

A system that works in Europe will not necessarily work if it is simply transplanted to Tokyo. The underlying social conditions are different.

I have probably been too attached to the specific idea of “matching Japanese and non-Japanese co-founders.”

What I truly want to create is not simply a service that introduces potential co-founders to one another.

What I want to create is an environment in which young people who were born and raised in Japan can meet others across national and linguistic boundaries—and in which taking on a challenge together no longer feels unusual.

I also want to create an environment in which companies are built not solely for the Japanese market, but are open to the world from the very beginning.

I still do not have a clear answer as to how this can be achieved.

I want to make use of my experience and network to create work that I genuinely enjoy—and that also provides me with a proper income. At the same time, I want that work to help young people in Japan venture out into the world, even if only in some small way.

Can I really create something that conveniently satisfies all of those ambitions?

I honestly do not know.

But the chain of serendipitous events that began with my reunion with Allen has taken me to places that the 25-year-old version of me could never have imagined.

Of course, that journey also came with many hardships and setbacks.

Still, I do not want that serendipity to remain merely my own good fortune.

I want to turn it into something that can be passed on—a system that creates similar opportunities for the next generation.

I recently went to Setagaya Park, a relatively large park in Tokyo, for the first time in several decades.

According to Google Maps, it takes about 32 minutes to walk there from my home. It is 2.2 kilometers each way.

There is a small train for children in the park. When my older son was still in preschool, I sometimes took him there. At that time, we lived in Ebisu, a very central part of Tokyo, but as our children grew, we moved to the house where we live now.

For me, Setagaya Park is somehow connected with Masaki Akashio, a former member of the Japanese pop duo H2O.

I met him by chance when I was still a university student. We both happened to live in Shimokitazawa in Tokyo at the time. He had already left H2O and was preparing to study at Berklee College of Music in Boston.

His name sounded a little like Akashiya Sanma, one of Japan’s best-known comedians, so people called him “Sanma-chan.” Apparently, my friends and I are now almost the only people who still call him that.

Before leaving for Boston, he said that he wanted to visit all the major parks in Tokyo, and he went to a different park almost every day.

I do not clearly remember why Setagaya Park became connected with him in my memory. But I vaguely remember him saying that it was a very good park.

The reason I decided to visit Setagaya Park again after so many years was that my younger son had recently gone there.

In September of his first year of junior high school, he contracted COVID-19. He developed long COVID and was unable to go to school for quite some time. His fatigue was so severe that he could not even sit in the living room to eat a meal.

At first, we thought that he would probably recover within six months. Unfortunately, that expectation was completely wrong. Even after he entered his second year of junior high school, he was still unable to return to school.

One piece of good fortune was that my wife found a clinic specializing in long COVID.

The director of the clinic was also a member of a Tokyo Metropolitan Government team dealing with long COVID. The clinic was—and still is—fully booked and extremely busy. Fortunately, there was a cancellation, and my son was able to see the doctor.

I think it was sometime after the summer vacation in his second year that he gradually began attending classes again. At first, it was only one or two hours a day, two or three days a week.

After entering his third year, he became able to go to school almost every day. However, he is still not able to participate in physical education classes. It will probably be difficult for him to return completely to the condition he was in before COVID.

But he has accepted this reality and seems to enjoy his life every day.

Even though he is my own son, I think he is admirable.

This explanation has become quite long, but in late July, he walked to Setagaya Park. He told me that it was a wonderful park, and that made me want to visit it again after several decades.

Unfortunately, I cannot do it every week, but I have my own jogging—or perhaps walking—course. The shorter course is 3.4 kilometers, and the longer one is 3.8 kilometers. There are many steep hills. I run uphill and walk downhill or on flat roads, trying to raise my heart rate and burn calories.

The road to Setagaya Park is almost straight. It is mostly downhill on the way there and uphill on the way back.

The round trip is 4.4 kilometers. I do not really run along the way and mostly walk at a fast pace. Even so, it makes me more tired and burns more calories than my usual jogging course. It also makes me feel very refreshed.

I would like to go there about once a week if possible. But including some time to rest in the park, I need about an hour and a half. I have not been able to go again since that day.

Meanwhile, a typhoon is approaching. Before my business trip to Silicon Valley at the end of this month, I would like to go there at least one more time.

Speaking of doing something again after a long break, we recently had a barbecue on our rooftop.

Before my younger son developed long COVID, we used to have rooftop barbecues fairly often from spring through autumn. But we had not been able to do it for the past two years.

My older son is now a university student and is busy with many things, so it is not easy to find a day when all four of us can be together. On that day, his part-time job finished at 7:00 p.m., so for the first time in a while, we had an evening when the whole family could gather.

We went to a nearby supermarket to buy food and started preparing the charcoal at around 6:00 p.m. The charcoal had not been used for two years, so I thought it might have absorbed moisture. But it caught fire more quickly than expected.

An earthquake registering the maximum level of 7 on Japan’s seismic intensity scale struck Kumamoto Prefecture, causing many deaths and affecting many people. I made a small donation for those who were affected.

Outside Japan, Russia’s invasion of Ukraine has now continued for four and a half years. With the United States attacking Iran, the future of the Middle East is also uncertain.

Being able to live these ordinary days is something very fortunate.

I am grateful for this good fortune. I also hope to continue using the experience I have gained—as an entrepreneur, investor, and educator—to do what I can to help create a better society.

1. The question

Let me admit something. When I began this analysis, I could think of little except valuation (the price at which shares are issued). Having raised money as a founder and written checks as an investor, I have it in my bones that valuation is the variable that matters. How much dilution can I avoid? How many times my money comes back?

But capital efficiency walks straight past all of that. How much did you raise, and what did it become? It is arithmetic a child could do. Valuation gets absorbed inside the division and disappears.

We all read numbers through the bias of where we stand. Only when I saw that did these numbers come into focus.

Vol.1 classified the pre-IPO shareholder structure of Japan’s top 50 listed startups by market cap — whose capital built these companies — from primary sources. The answer: VC- and external-capital-led companies form a clear majority (66%).

But something remained unmeasured. What Vol.1 examined was external shareholders’ ownership — not how much money they put in. The same 30% stake means something entirely different at a ¥1bn valuation than at a ¥10bn one.

So Vol.2 changes the lens. The question here is:

What did the capital actually turn into?

The metric is simple. Capital efficiency = market capitalization ÷ pre-IPO capital raised. Think of it as fuel economy. How much fuel did it take to cover the distance?

The set is ten representative companies, chosen as five head-to-head matchups: Money Forward vs. freee, DeNA vs. GREE, ANYCOLOR vs. COVER, SHIFT vs. Rakus, and Mercari vs. Sansan. Each pair competes — or has competed — in the same market, and each pair also draws a sharp contrast in the amount of capital raised.

And we measure at four points: the IPO (offer price), the first-day close, the all-time high, and today (July 7, 2026). Listing is not the finish line. Without following what happened afterward, the real picture of capital efficiency stays hidden.

2. Definitions and notes

Capital efficiency = market cap ÷ pre-IPO capital raised.

The denominator comes from the “history of capital and capital reserve” table in each IPO prospectus. It excludes capital raised in the IPO offering itself (the point is to measure the productivity of the capital investors put in before listing). It also excludes proceeds from the exercise of stock options granted to insiders. Allocations to existing shareholders are included.

The numerator is market cap at each point — offer price, first-day close, all-time high, and the July 7, 2026 close — multiplied by shares outstanding (adjusted for stock splits).

An important note on Rakus. Rakus’s denominator (¥149.85m, the stated capital as of the listing approval date) consists of the founders’ own capital; investments from Nojix (whose representative served as a Rakus director until just before the IPO) and Septeni; and increases funded from operating profit (inferred). The precise amount of purely external capital is unverified, but it is unquestionably tiny compared with any other startup here — and not a single yen came from a VC. This matters enough to warrant its own section below.

3. Three findings

3-1. The more a company raised, the lower its capital efficiency

This is the clearest finding of all. Capital raised and capital efficiency are strongly inversely correlated (Spearman ρ = −0.89). Measured at the all-time high, the relationship is nearly identical (ρ = −0.92).

Rank the ten by capital raised, and the light raisers dominate the efficiency table: Rakus (¥150m), GREE (¥460m), COVER (¥790m), SHIFT (¥880m). The heavy raisers sink: Mercari (¥16.5bn), freee (¥16.0bn), Sansan (¥11.4bn).

This may look trivial — a smaller denominator produces a bigger multiple. But it is not a story about small companies winning. Mercari’s market cap is ¥745bn; Rakus’s is ¥367bn. Mercari is the larger company. And yet Rakus dwarfs it on efficiency. Some companies grew by burning capital. Others grew without burning it.

Notably, the correlation between Vol.1’s external investor ownership and capital efficiency as of the July 7, 2026 close is weak (ρ = −0.20). How much you raised predicts efficiency far better than how much outsiders owned of the company. These are two different axes.

One caveat is worth stating. Capital raised is the product of valuation and the share of equity given up. Hand over the same 30%, and a higher valuation means a larger amount raised — and a fatter denominator in the efficiency ratio. Valuation, in other words, is absorbed inside the division. This report does not measure each company’s valuation directly, so its contribution cannot be isolated. What can be said is only this: it is not what share you gave away, but how much you took in, that moves the number.

3-2. The IPO valuation tells you nothing about what comes next

Follow the four points and one thing becomes clear: capital efficiency at IPO has no relationship to what follows.

SHIFT is the emblem. Its efficiency at IPO was 3.9x — the lowest of the ten. At its all-time high, it reached 735.2x: 187 times its IPO level. At the time of listing, the market simply did not understand what it was looking at.

GREE is the mirror image. At IPO, it stood at 158.5x — extraordinary for its time. But its all-time high was 743.0x, only 4.7 times the IPO level. The company, priced richly by the market from day one, had the least room left to run.

Listing is not the finish line. For some companies, it is barely a milestone.

3-3. All ten are below their all-time highs

This was within expectation, but the scale is striking: every one of the ten trades below its all-time high — by 58.2% on average. Not a single company is at its peak today. It is precisely why the Tokyo Stock Exchange is pushing through its Growth Market reform.

And the peaks cluster in 2021: five of the ten (Money Forward, freee, Rakus, Mercari, Sansan). The afterimage of the growth-stock bubble. It is a measure of how much the market of that moment was willing to believe.

The steepest falls are freee (−79.3%) and GREE (−79.1%). freee peaked in February 2021 at ¥652bn and now sits at ¥135bn. GREE’s peak is older still — November 2011, fifteen years ago.

4. The efficiency ranking

5. Five matchups

5-1. Money Forward vs. freee — the fuel gap widens with time

Direct competitors in cloud accounting. Both founded in 2012, with similar business models. But they used capital in opposite ways.

freee raised 3.9 times as much as Money Forward (¥16.0bn vs. ¥4.10bn). Even so, Money Forward was already ahead on efficiency at IPO — 6.9x against freee’s 5.8x.

And the gap widened with every passing year.

Money Forward’s capital efficiency today (as explained at the beginning, as of July 7th, 2026. The same applies hereinafter.) is 7.8 times freee’s. On market cap, the gap is 2x: ¥271bn against ¥135bn.

freee’s drawdown from its February 2021 peak is −79.3%, the largest of the ten — the widest gap here between what the market expected and what arrived.

5-2. DeNA vs. GREE — the cruelty of a fifteen-year-old peak

The two giants of Japan's social-gaming era, rivals to the point of litigation. Their numbers hold the most telling story in this report.

GREE reached its IPO having raised just ¥460m. Its IPO-level efficiency of 158.5x is by far the highest of the ten. Under founder Yoshikazu Tanaka’s control (62.40%), Globis invested, and then KDDI — a cap table that was much discussed in the startup world at the time.

DeNA, by contrast, raised ¥2.28bn, with heavy corporate capital from Sony, Sumitomo Corporation and others; its IPO efficiency was a modest 14.3x.

But both companies’ all-time highs fell in 2011 — the height of the social-gaming boom. Fifteen years on, neither has returned to that level.

The difference is in how far they fell. GREE is down 79.1% from its peak; DeNA is down 50.0%. Two companies that were more than 10x apart in efficiency at IPO now sit almost level: 155.3x against 142.3x. One reading: DeNA held on to its value; GREE lost fifteen years of it.

Tomoko Nanba’s return as CEO, the professional baseball franchise, and Delight Ventures’ backing of Japanese founders taking on Silicon Valley — those moves can be seen as the backdrop to the gap in these numbers.

5-3. ANYCOLOR vs. COVER — same market, same era, lighter capital wins

Direct competitors in VTuber entertainment: Nijisanji (ANYCOLOR) and Hololive (COVER). Both listed within a year of each other, in 2022 and 2023.

COVER leads ANYCOLOR at every point. The reason is simple: capital raised. COVER took ¥790m; ANYCOLOR took ¥2.58bn — 3.3 times more.

ANYCOLOR is the larger company by market cap (¥151bn vs. ¥107bn). But per yen of capital invested, COVER created 2.3 times the value.

5-4. SHIFT vs. Rakus — the far end of capital efficiency

These two represent the extreme of growth without burning capital.

SHIFT listed having raised ¥880m. Its IPO efficiency of 3.9x was the lowest of the ten. It then climbed to 735.2x at its all-time high — 187.6 times the IPO level. Software testing is an unglamorous business; the market took years to price it.

And then there is Rakus — the company this report owes the most words to.

5-5. Mercari vs. Sansan — the ceiling of the heavy raisers

Finally, the heavyweight matchup.

Mercari raised ¥16.5bn; Sansan ¥11.4bn. Mercari leads on efficiency by 2.4x, but neither comes close to the leaders above 100x.

One nuance: Mercari’s drawdown from its peak is −36.1%, the smallest of the ten. The heavy raisers are less efficient — but their value also moves less. Thick capital may dull the climb and cushion the fall.

6. Rakus — the company that raised nothing from VCs

6-1. What actually happened

Rakus’s capital efficiency stands at 2,448x today, and reached 5,646x at its all-time high. That is far and away the highest of the ten — 11 times second-place SHIFT (216x), and 290 times last-place freee (8x).

But this is not a story about being efficient. The denominator barely exists.

Rakus’s stated capital before the IPO was ¥149.85m (as of the listing approval date, November 4, 2015). It appears to have been built up mainly from the founders’ own capital and from increases funded by the operating profit the business itself generated (the latter is an inference).

And not a single yen came from a VC.

6-2. What the cap table says

Look at the shareholder structure at IPO.

The only external holders were Nojix (1.84%) and Septeni (2.21%) — 4.05% in total, both operating companies. Nojix’s representative served concurrently as a director of Rakus until just before the IPO (he stepped down for the listing). Septeni’s reasons for holding the stock — a commercial relationship, or a personal one — are unclear, but either way, neither is a VC.

And after the oyabiki allocation at IPO, management, including the founder, held 77.21% (79.34% including Septeni).

What is oyabiki? In a Japanese IPO, it is the practice of allocating shares from the offering, through the underwriter, to third parties designated by the issuer (business partners, alliance partners, and so on). It sits outside the normal lottery and discretionary allocation, and is typically used to build a base of stable shareholders after listing.

External capital: 4.05%. VC capital: zero. Management holding over 77% after listing. That is the starting point of a company now worth ¥367bn — and ¥846bn at its peak.

6-3. Not "founder-led." Bootstrap.

In Vol.1, we renamed the “Bootstrap” category to “founder-led.” The reason: Bootstrap reads as “zero external capital,” which was inappropriate for a definition that tolerates up to 15% external ownership.

That call was right. But now a company has appeared that fits the very reading we were trying to avoid — literally.

Rakus grew on the capital of its founders and insiders, and the cash flow the business generated. It raised no venture capital, went public, ran ads on television, and became a ¥367bn company. This is not “founder-led.” This is Bootstrap, in the truest sense of the word.

In the history of Japanese startups, that is a rare achievement. Which is why the Bootstrap category deserves to be revived — for Rakus (n=1).

We set out to measure capital efficiency — and found a company we could not measure. That may be the most valuable finding here.

7. Implications

Three.

First, “how much you raised” and “what it became” are different questions. Mercari raised ¥16.5bn and is worth ¥745bn. Rakus raised ¥150m and is worth ¥367bn. Which is the greater success depends on the question you are asking. But on the single axis of capital productivity, the answer is not close.

Second, the valuation at listing guarantees nothing. SHIFT, the least efficient of the ten at IPO, later ran to 735x. GREE, by far the most efficient at IPO, cannot reclaim a peak set fifteen years ago. At the moment of listing, the market has not finished pricing the company.

Third, external capital is not a precondition. Vol. 1 showed that 66% of Japan’s top startups were VC- or externally capital-led. And yet Rakus became a household name through television advertising, with virtually no external capital at all. External capital is an accelerant, not a prerequisite — and Rakus is the living proof.

8. Limitations

Honestly stated.

• The sample is ten companies. The granularity differs from Vol.1’s fifty. What is here is a hypothesis, not an established law.

• The denominator’s definition has edges. Proceeds from insider stock option exercises are excluded; allocations to existing shareholders are included. Disclosure granularity varies by company, so perfect comparability is not guaranteed.

• Rakus’s denominator, as noted, contains no VC money but does mix a small operating-company investment (4.05%) with increases inferred to have been funded from operating profit. It is not literally “zero external capital,” and should not be read as such.

• Market caps depend heavily on market conditions. The 2021 growth-stock bubble, and the correction that followed — the figures here are a cross-section of a single day, July 7, 2026.

Data: each company's IPO prospectus (history of capital and capital reserve), Minkabu, Yahoo! Finance. Market caps as of the July 7, 2026 close. Analysis and charts by the author. For pre-IPO ownership figures, see the accompanying spreadsheet (v4.0).

1. The question

Did Japan’s most valuable listed startups grow with venture capital? With capital retained by their founders? Or with strategic investment from operating companies?

To answer this, I narrowed the field to “startups” — companies founded mainly from the late 1990s onward, first listed on an emerging market, and still listed today — then took the top 50 by market cap and classified each by its pre-IPO shareholder structure. The axis of classification is what I call Investor DNA: the type of capital that primarily supported the company’s growth. Wherever possible, I verified this against a primary source — the major-shareholder table in the IPO prospectus.

One important note: the figures in this report measure external investor ownership — the share of equity held by external investors immediately before the IPO — not the amount of capital raised. These are different measures. An analysis based on how much capital each company actually raised could reveal a different picture.

With that focus on equity ownership, here is the conclusion up front. Once the universe is narrowed to startups — businesses built on technology or a new business model — and external capital is measured from primary sources, VC- and external-capital-led companies form a clear majority. But the conclusion changes significantly depending on two things: how the universe is defined, and how external capital is measured.

2. Defining the universe, and what I excluded

Three formal criteria: (1) still listed today; (2) an emerging-market IPO (judged by the first listing venue, with an exception for VC-backed startups that listed directly on the First Section); and (3) founded, as a rule, from the late 1990s onward.

But formal criteria alone were not enough. Many companies that listed on emerging markets are retailers, restaurant chains, long-established manufacturers, or maintenance businesses — companies whose nature is difficult to describe as that of a startup. So I added a qualitative judgment and excluded the following companies, regardless of market cap.

A few principles guided the exclusions. Where an operating company had invested and had large ownership, but, with founders still in control, I kept the company (MonotaRO, Monex, M3). PE shells — vehicles set up to acquire an existing business — are out, but a company that a fund and professional managers built from zero, like GENDA, is in. Even among M&A brokers, I kept Strike because it runs S Venture Lab — an open-innovation M&A platform connecting large corporates with startups — a genuinely different kind of business.

The boundary is continuous, and no perfectly clean line exists. That is why I believe it is essential to disclose which companies were excluded and why. Since the exclusions include companies large enough to reshape the ranking — Nitori at ¥1.4tn and BayCurrent above ¥1tn — the definition of the universe is central to this analysis.

3. Cross-analysis

3-1. Investor DNA distribution — external capital accounts for two-thirds

Re-measuring external investor ownership from primary sources, VC/Angel-type companies number 33 (66%) and the founder-led camp 14 (28%) — the latter being the chart's Founder-led bar (13) plus one Founder/Family company — with the rest strategic, PE, or public. Mixed, a blend of several capital types, is the single largest bucket.

Two definitions matter here. By “VC/Angel-type,” I mean the sum of Mixed, Independent VC, Corporate VC (CVC), Foreign VC, and Angel — everything on the chart except the founder-led greens. “Independent VC” is restricted to companies led by independent funds that are not subsidiaries of financial institutions or operating companies.

And “founder-led” is defined precisely: (1) founders and insiders hold a dominant stake, and (2) external investor ownership — VC + CVC + operating companies + angels + financial investors, excluding officers, employees, and relationship holdings from regional banks or credit unions — is at or below roughly 15%. The question is how much external risk money went in, not whether it was VC specifically.

3-2. Inside the cap tables of “founder-led” companies

The part I learned the most from was this: even when founders held a large stake, the prospectuses often showed that a meaningful amount of external capital had also been invested in the company.

• SHIFT — founder Masaru Tange holds 56%, while external holders (Mitsui & Co., NTT, Draper Nexus, and others) account for about 33%.

• SMS — founder-controlled by Morofuji and Taguchi, yet external capital (Care21, So-net M3, As One, etc.) is about 30%.

• Infomart — Murakami holds a non-controlling 17.75%, while external holders, led by Mitsubishi Corporation (13.88%) and Mitsui & Co. (13.88%), collectively hold more than 60%.

• PKSHA — founders hold 54%, but a family office, strategic investors, and Toyota’s CVC bring external capital to about 20%.

• MIXI — Kenji Kasahara holds 62.75%, yet NetAge-affiliated VCs and CyberAgent push external capital to about 24%.

All five are founder-controlled, but external investor ownership is far from negligible. In B2B and healthcare, in particular, strategic investment from operating companies appears to play a larger role than VC. By contrast, Rakuten, ZOZO (Maezawa 80%), JINS (Hitoshi Tanaka 74%), Bengoshi.com (Motoe 85%), Sakura Internet, Azoom, Rakus, Cybozu, Colopl, Change, and Strike are genuinely founder-led, with external capital at or below 10–11% — again, confirmed from primary sources.

3-3. Generation × capital — the newer the cohort, the more external capital

The share of external-capital-type companies (VC/Angel/Mixed) rises with each generation: 25% for Pre-1995, 55% for Generation 1, 82% for Generation 2, and 86% for Generation 3. The clear shift occurs between Generations 1 and 2 — companies founded in the 2000s. The development of Japan’s VC ecosystem from the mid-2000s onward is also visible in the structure of these cap tables.

3-4. Market Cap × External Investor Ownership — no simple positive correlation

The chart below excludes the seven firms with market caps of ¥500bn or more (Rakuten, M3, ZOZO, MonotaRO, CyberAgent, Mercari, Kakaku.com) to make the ¥100–200bn band of startups easier to read. Three clusters emerge: founder-led firms (left of the 15% line), Mixed and Independent VC (the middle), and a high-external-capital group (right of the 45% line).

The simple story — "the more external capital, the higher the market cap" — does not hold within the top 50. Median market cap is about ¥137.9bn for founder-led (14 firms) and about ¥137.6bn for Mixed (22 firms) — essentially level, and not explained by how much external capital went in. If anything, the very top of the list — Rakuten and ZOZO — are founder-led companies with external capital below 10%.

Among companies with external capital below 15%, another pattern appears. Many were founded around or before 2000 (Rakus was founded in 2000), before Japan’s startup ecosystem had developed. Raising external capital was not as easy then as it is today. One possible reading is that these companies grew largely through their own operating cash flow. (The scatter also plots WealthNavi and Raksul — since delisted — as red-X references; see below.)

For reference, average market cap by sector group is shown below.

3-5. Two off-list companies — WealthNavi and Raksul

The top 50 here is restricted to companies still listed. But two firms you cannot leave out of any conversation about Japan's startup scene went private recently via tender offer or MBO: WealthNavi (2025, a tender offer by MUFG Bank) and Raksul (2026, a Goldman Sachs–backed MBO).

Both are representative Japanese VC-backed startups that grew with substantial external capital, led by independent VCs. Their take-private values were roughly ¥116.1bn for WealthNavi and ¥130.0bn for Raksul. Their external investor ownership, confirmed from primary sources, both exceeded roughly 60%. At WealthNavi, founder Kazuhisa Shibayama held 24.84%, while SBI, DBJ Capital, Global Brain, and other VCs and financial investors formed the external core. At Raksul, founder Yasukane Matsumoto held 21.55%, while WiL, Global Brain, YJ, the Development Bank of Japan, and other VCs formed the external core. Both sit on the right-hand side of the scatter, where external capital is high. Remaining listed is not the only definition of success. Large startups that chose to go private for their next stage of growth also used external capital as a major source of growth — and these two are important examples.

4. Key findings (hypotheses)

• Finding 1. When the universe is limited to startups and external capital is measured from primary sources, VC- and external-capital-led companies form a clear majority (66%) among the top 50.

• Finding 2. Even many companies that appear “Founder-led” have accepted a meaningful amount of external capital. In B2B and, in particular, healthcare, strategic investment from operating companies plays a major role.

• Finding 3. The newer the generation, the greater the dependence on external capital; the shift is clear by Generation 2.

• Finding 4. There is no simple positive correlation between external capital and market cap. Founding-era business and business-model changes alter the relationship between external investor ownership and market cap.

5. Implications

Two implications stand out. First, the conclusion changes with the definition of the universe. If the ranking is created mechanically from formal criteria alone, retailers, restaurant chains, and companies originating from established corporate groups become more prominent, creating a picture in which VC and other external capital appear less important. Narrow the universe qualitatively, and the picture changes.

Second, even many companies classified as “founder-led” have brought in a meaningful amount of external capital.

This may sound obvious, but the composition of the universe has a major effect on the result.

6. Limitations

This classification remains a hypothesis. The limitations are as follows.

• External investor ownership was fixed from primary sources (IPO prospectus shareholder tables) for most companies, but for two — IIJ and GMO Internet — primary sources were unavailable, and I classified them qualitatively (founder control). For Rakuten, I treated the Crimson-affiliated funds as the founder's vehicles, so the possibility of external LPs remains.

• Numeric external-investor-ownership figures are still missing for a couple of firms in the upper band (notably IIJ), so any quantitative correlation with market cap is limited.

• In the ¥10bn–¥100bn band, small differences reshuffle the ranking; positions near #50 are not definitive.

Data: each company's IPO prospectus (major-shareholder table); July 7, 2026 close. Classification and charts by the author. For the underlying figures, see the accompanying spreadsheet (v3.9.1).

I wasn’t able to catch Launch Pad, one of the marquee sessions, this year for reasons beyond my control. That was a shame, but a few other sessions left a real impression. Against the backdrop of the Tokyo Stock Exchange’s ¥10 billion market-cap maintenance requirement, this year’s standout themes were “how startups earn the right to be chosen by shareholders once they clear the IPO bar” and “the future of Japanese startups.” I’m writing this down as a record.

Time to Drop the Idea of a “Japanese Startup.”

The session that stayed with me most was “The Future of Japanese Startups — How Should Founders Face It,” featuring Masaya Kubota, General Partner at Project Coalis.

When we say “Japanese startup,” we unconsciously picture Japanese people doing business in the Japanese market. Kubota said flatly: drop that idea. In a country with a shrinking population, it makes no sense. Startups in Silicon Valley don’t think of themselves as American startups doing business in America. He’s right.

This is exactly what I’ve been arguing for ten years: globally competitive startups built by multinational teams, even from Japan. The question is how to break the village mentality behind “Japanese startups” and step onto the World Cup stage. Japan is a highly homogeneous society whose native language is Japanese — and yet, as the world’s fourth-largest economy, unless you’re either exceptionally capable or a bit reckless, taking on the risk of going global just doesn’t feel necessary. I was the same way: during the dot-com bubble, going overseas never even crossed my mind.

What changed my thinking was getting involved with SunBridge to restart the accelerator and shuttling between Silicon Valley and Japan from 2011 onward. Japan can feel like a warm bath — safe, with great food and reasonably stable politics — and that makes it hard to even notice the issues that matter globally.

That’s what led me to build a co-founder matching program pairing Japanese and non-Japanese founders. When I ran a pilot in July 2025, I received 30 applicants — only 3 of whom were Japanese. Some of that was no doubt due to my own limited name recognition and reach, but once the outreach and the program itself were in English, it became much harder to reach Japanese founders at all.

In continental Europe, when a German and a French person work together, in most cases, neither side gets to use their own mother tongue. English becomes the default by necessity.

Japan, an island nation with 120 million people and the world’s fourth-largest economy, can still get by on its own local rules — for now.

But without the resolve to climb out of that warm bath, I don’t think Japanese startups have a future.

Back to the IVS program. Kubota also talked about wanting to see more pre-IPO startups raise capital to acquire listed startups whose market caps have stagnated, and help them grow again. Or combine several pre-IPO startups to build something at scale before taking it public. That’s why he launched Project Coalis, which also functions as a secondary fund.

At the top of the panel, he’d already laid out the hard truth: during the dot-com years, or even ten years ago, we could just make out Silicon Valley’s back from here. Now we can’t see it at all. The gap between them and us has widened enormously. And precisely because of that, he argued, we need to think seriously about what we can do — and then actually do it.

Two things matter most, he said: the environment you put yourself in, and never stopping yourself from putting yourself out there. Even someone like Anthropic’s Dario Amodei tweets on X every day — because that’s how you keep the world’s attention on you.

“Nobody in this room can email or text Elon Musk or Jack Dorsey and get a reply. We’re nobody.”

Perhaps it is precisely because he has achieved so much that he can look at this harsh reality so clearly.

What’s Beyond the IPO Finish Line.

The other standing-room-only session was “Beyond the IPO Finish Line — Building a Startup Market Where Shareholders Choose You.” The panelists were: Hideto Fujino, founder of Rheos Capital Works; Shintaro Tabata, activist public-equity investor and YouTuber (small world — we ran into each other on the shinkansen platform on the way home); Masayuki Mochida, Managing Director at Tybourne Capital; and Yusuke Asakura, Managing Partner at Animal Spirits.

Asakura said business model and strategy matter, but what really determines a company’s growth is the sheer intensity — the passion — of its founder and leadership.

Tabata’s point was the biggest single takeaway for me. Pre-IPO startups get rewarded for dramatic, come-from-behind stories — a grand-slam home run in the bottom of the ninth after years of twists and turns. Public markets don’t work that way. Companies that grow steadily by 20–30% every year, without fail, are the ones whose market cap actually compounds. What makes CyberAgent’s Fujita remarkable is that he understood this and delivered relentless, uninterrupted growth since the IPO.

IR is marketing aimed at shareholders. Who do you want holding your stock? What do they value? What information do they need to make an investment decision? You need to understand all of that and design your communication accordingly.

IVS CORE and Some Hope.

This year, IVS CORE was introduced, a VIP track with a hefty ticket price. An American acquaintance got me in, and I was lucky enough to catch part of it.

One line stuck with me: Alex Bangash, Managing Director and General Partner at Transpose Platform, was asked by the moderator what Japan needs. His answer was simple: “Accept risk. Build a culture that tolerates failure.”

But this is something people have been saying for ten, even twenty years. And that tells us something: Japan — the country and its people — struggles to change unless it is jolted by an event on the scale of the Meiji Restoration or the defeat in World War II. I would call this a fatal weakness, partly because Japan is such a homogeneous society. The Japanese phrase dōchō atsuryoku — the pressure to conform — says it all.

But there’s hope, too. The younger generation — say, anyone under 40 — feels different. People like Satoru Naito, who wrote “Hi, Silicon Valley (in Japanese only)”; Shunsuke Yamada, who appears in his book; and Satoru Sasozaki — Japanese founders who’ve left home to build in Silicon Valley — are slowly emerging. What our generation owes them is simple: back these challengers with everything we’ve got.

I ran into Naito and Yamada — I’ve put a small amount of money into Yamada’s startup — on the last day of IVS, right there in the venue. Probably not a coincidence. I also ran into Yuka Fujii of AMI PHOTO, who carried my bag as an intern back in her student days, and some of my students from Musashino EMC, the only department dedicated to entrepreneurship in Japan, whom I bumped into at the venue as well. And then, at a side-event angel gathering, I met Daiki, who runs his own startup, the son of Hitomi Takeuchi — founder of Coloridoh, another company I’ve put a small amount into — who’s now building his own AI startup.

That’s this year’s IVS, on the record

.

In this context, let us look back over the quarter-century since the dot-com bubble.

A VC’s contribution to funding and nurturing startups cannot be measured by IPOs alone—M&A exits and other outcomes must also be considered. However, given the relative ease of obtaining IPO data and the shifting significance of IPOs in light of the TSE’s stricter listing maintenance criteria mentioned above, this report focuses exclusively on IPOs produced by Japan’s VCs. Are there differences based on the type of parent organization—bank-affiliated, independent, or corporate? Even though individual VCs post their track records on their websites, there are few resources that organize this information in a way that enables cross-comparison.

This report covers major VCs—including non-JVCA (Japan Venture Capital Association) members—and analyzes their IPO track records across three layers based on each firm’s official website, new listing data from JPX (Japan Exchange Group), and industry-standard data sources such as INITIAL and STARTUP DB. Cumulative IPOs, IPOs since 2000, and startup IPOs (defined in the body of this report) are organized for side-by-side comparison to the extent possible.

Furthermore, English-language information about Japanese startups remains limited. Even when overseas investors and startup professionals try to accurately understand Japan’s ecosystem, there is a scarcity of primary sources they can reference. We hope this report serves as a useful reference for industry stakeholders both in Japan and abroad.

1. Survey Overview and Definition of "Startup IPO."

The Japanese VC industry has a long history. In particular, the "cumulative IPO count" for bank-affiliated and legacy VCs includes many legacy deals from before the 1990s, as well as financing-type investments in medium- and small-sized enterprises (SMEs) that cannot be considered startups. Therefore, to measure the pure results of startup investments, we have structured the track record of each VC into the following three layers:

Layer 1: Cumulative IPOs (All-time / Since inception) The total number of IPOs across all periods officially announced by each VC. If no official announcement is available, estimated figures from external databases are provided.

Layer 2: Estimated IPOs since 2000. A metric to measure the results of modern venture investments since the IT bubble period. For JAFCO, the actual count directly tallied from their public PDF list is provided; for other VCs, estimated figures based on public information are used.

Layer 3: Estimated Startup IPOs (Since 2000) A metric estimating the number of "Startup IPOs" extracted from Layer 2 based on the following criteria:

◦ Criterion 1: Companies established in or after 1998 (excluding legacy SME deals from before the IT bubble).

◦ Criterion 2: Companies with a high probability of being startups at the time of receiving VC investment (investment stage is irrelevant; includes all stages from seed to later).

Note: While syndicated investments by multiple VCs are ideal, single-investor deals are also included if they meet the startup criteria above.

This survey includes estimated figures to provide an overall picture, comprehensively analyzes new listing data from JPX (Japan Exchange Group), startup databases such as INITIAL and STARTUP DB, and the public portfolio information of each VC . [1] [2]

2. IPO Track Record of Major VCs (Three-Layer Analysis)

Editor’s Note: Based on additional research conducted after the initial publication, I have revised and expanded parts of the data and analysis below. This update reflects the information available as of Thursday, June 20, 2026, Japan Standard Time.

Leading VCs in this report are categorized into three groups: "Bank-affiliated & Legacy VCs," "Independent Startup-focused VCs," and "CVCs & Corporate-affiliated VCs."

[Bank-affiliated / Legacy VCs]

Historically, these VCs have included financing-type investments in SMEs, so Layer 3 is narrower than Layers 1 and 2. However, regardless of the investment stage (seed to later), companies are included in Layer 3 if they are "startups established in or after 1998."

Note: Mitsui Sumitomo Insurance Capital is a legacy VC established in 1990, but since no official cumulative IPO count for the entire company could be found (only an excerpt of 42 companies since 2001 is available), it has been excluded from the table. [17]

[Independent Startup-Focused VCs]

Since their inception, these VCs have primarily invested in seed and early-stage startups, and have syndicated co-investments as standard practice. Layer 3 ≒ Layers 1 and 2.

Note: B Dash Ventures mentions "over 100 total investments and many IPOs" in their official note, but is excluded from the table because the specific number of IPOs is undisclosed (past materials mention 16-19 IPOs) [28].

[CVC / Corporate-backed VCs]

While aiming for business synergy with their parent companies, they also make pure investments in startups. Since they were established relatively recently, Layer 3 ≒ , Layer 1, and Layer 2 in many cases.

Note:

*1: JAFCO's Layer 2 "382 companies" is the actual count directly tallied by extracting and analyzing text from their official public PDF "List of Listed Portfolio Companies" (as of April 22, 2026, 833 domestic companies). However, due to the PDF layout, there may be some text extraction misrecognitions, allowing for a margin of error of approx. ±5 companies.

*2: Layer 2 figures for Mitsubishi UFJ Capital, Mizuho Capital, and SBI Investment are estimates based on annual data and company lists posted on their official track record pages.

*3: Nissay Capital’s Layer 1 “278” is from their official site. Layer 2 is estimated by excluding pre-2000 deals from the list.

*4: SBI Investment does not officially disclose cumulative IPOs. The “1,314 total investees” on their site is not the IPO count and differs in nature from Layer 1 of other VCs.

3. Trend of Startup IPOs in Japan by Year (2000–2025)

The table below shows the trend in the proportion of "Startup IPOs" (companies that received VC investment) relative to the total number of IPOs in the overall market. Total IPO figures are based on JPX (Japan Exchange Group) new listing data [38]. Includes estimated figures based on databases like INITIAL [1] [37].

Note: This data excludes the TOKYO PRO Market.

4. Discussion

Through comprehensive surveys of JVCA members and a three-layer analysis, the diversity and structural differences among the players supporting Japan's startup IPO market emerge clearly.

The Essential Difference Between "Bank-affiliated/Legacy VCs" and "Independent VCs" Shown by the Three-Layer Structure. Bank-affiliated and legacy VCs (JAFCO, Mitsubishi UFJ Capital, Daiwa Corporate Investment, etc.) boast overwhelming track records of hundreds to around 1,000 companies in Layer 1 (Cumulative IPOs). However, they exhibit a "diminishing structure" where the numbers shrink as they are narrowed down to Layer 2 (Since 2000) and Layer 3 (Startup IPOs). This indicates that their historical track records included many financing-type listing support deals for SMEs. Nevertheless, they still generate tens to over 100 startup IPOs in Layer 3, surpassing independent VCs, demonstrating that they have adapted to modern startup investments alongside traditional SME investments and continue to be the backbone of capital supply in Japan's ecosystem.

On the other hand, independent VCs such as GCP, Global Brain, and CyberAgent Capital show almost identical numbers across Layers 1-3. This is direct evidence that, since their inception, they have primarily focused on investing in seed and early-stage startups and have walked the royal road of modern startup investing by making syndicated co-investments alongside other VCs. The figure of "almost 100% startup IPO purity" embodies the very nature of their funds.

The Rise and Track Record of CVCs and corporate-affiliated VCs established in the 2000s and later, such as Z Venture Capital, ITOCHU Technology Ventures, and SPARX Asset Management, are steadily producing startup IPOs on a scale of 20 to 40 companies. The numbers show that they are achieving both pure capital gains and startup growth support, not just business synergy with their parent companies, establishing themselves as important players in the market.

Maturation of the Startup IPO Market and Recent "Selection." In the early 2000s, startups (VC investees) accounted for only about 20% of all IPOs, but since 2015, they have consistently accounted for 50-60%, clearly indicating that startups have taken the leading role in Japan's IPO market.

However, in 2025, total IPOs plummeted to 66, and startup IPOs also dropped to the 40s. As pointed out in INITIAL's report, against the backdrop of stricter listing maintenance criteria for the TSE Growth Market (market capitalization of 10 billion yen or more, 5 years after listing), investors are increasingly shying away from "subscale IPOs," leading to a stronger "selection" where capital is concentrated on companies with a clear path to winning [37]. Startups need to rethink where and what kind of business they will conduct in a world where AI is redrawing the boundaries of every industry. There is no doubt that Japan's startup ecosystem is entering a major turning point.

References

[1]: https://initial.inc/articles/japan-startup-finance-2025 “INITIAL. “Selectivity and Extended Matches—2025 Startup Financing Trends.”

[2]: https://lp.startup-db.com/media/articles/investment-report-2025 “STARTUP DB. “[Annual 2025] Domestic Startup Investment Trends Report”.

[3]: https://www.jafco.co.jp/portfolio/ “JAFCO Group Co., Ltd. “Portfolio (List of Listed Portfolio Companies PDF ).”

[4]: https://www.mucap.co.jp/feature/ipo/ “Mitsubishi UFJ Capital Co., Ltd. “IPO Track Record”.”

[5]: https://www.mizuho-vc.co.jp/publicstock/ “Mizuho Capital Co., Ltd. “Listing Track Record.”

[6]: https://www.daiwa-inv.co.jp/ Daiwa Corporate Investment. “Track Record.”

[7]: https://www.resona-gr.co.jp/resonacapital/trackrecords/ipo.html “Resona Capital. “IPO Track Record”.

[8]: https://www.jaic-vc.co.jp/business/pe_ipo/index.html “Japan Asia Investment. “IPO Track Record”.

[9]: https://www.nissay-cap.co.jp/portfolio/list.html “Nissay Capital Co., Ltd. “Investment Track Record.”

[10]: https://orixcapital.co.jp/aboutus/ “ORIX Capital. “About Us.”

[11]: https://www.nvcc.co.jp/portfolio/ipo/ “Nippon Venture Capital. “IPO Track Record”.

[12]: https://www.smbc-vc.co.jp/portfolio/ “SMBC Venture Capital Co., Ltd. “Investment Track Record.”

[13]: https://startup-db.com/investment-companies/Ao31G8nUg1lZkwyX “STARTUP DB. “DBJ Capital”.”

[14]: https://www.mickk.com/about “Mobile Internet Capital. “About MIC.”

[15]: https://www.ksp.co.jp/service/fund.html “KSP. “Investment Funds.”

[16]: https://www.sbinvestment.co.jp/news/ipo/ “SBI Investment Co., Ltd. “List of IPO Companies.

[17]: https://www.msivc.co.jp/results/ipo.html “Mitsui Sumitomo Insurance Capital. “IPO Track Record”.

[18]: https://globalbrains.com/portfolio “Global Brain Corporation. “Portfolio.”

[19]: https://www.globiscapital.co.jp/ja/about/history “Globis Capital Partners & Co. “History”.

[20]: http://www.cyberagentcapital.com/policy/ “CyberAgent Capital. “Policy.”

[21]: https://z.incubatefund.com/n/nb7ae6b2fd74b “Incubate Fund. “SUMMARY OF 2025”.

[22]: https://www.ut-ec.co.jp/ “UTEC”.

[23]: https://startup-db.com/investment-companies/gwlEZN8UN8mGvL2A “STARTUP DB. “WiL, LLC.”

[24]: https://www.xtech-ventures.co.jp/portfolio?and-taxonomy_portfolio_status%5B%5D=ipo “XTech Ventures. “Portfolio.”

[25]: https://tracxn.com/d/venture-capital/anri/__9qyxcQWMZeKFvbtWulyzhjK1OIeBZ2wL0jMoHsII4XM “Tracxn. “ANRI”.”

[26]: https://united.jp/business/investment/investee/ “UNITED. “Investee List”.

[27]: https://startup-db.com/investment-companies/rB5WJbnU365Jo97q “STARTUP DB. “Coral Capital.”

[28]: https://note.com/bdashventures/n/n5ccc0e0119f0 “B Dash Ventures. “B Dash Ventures Official note”.

[29]: https://zvc.vc/fund/ “Z Venture Capital. “Performance.”

[30]: https://www.techv.co.jp/news/2025331 “Itochu Technology Ventures. “News.”

[31]: https://mirai.sparx.co.jp/ “SPARX Asset Management. “Mirai Creation Fund.”

[32]: https://www.nttdocomo-v.com/portfolio/ “NTT DOCOMO Ventures. “Portfolio.”

[33]: https://gmo-vp.com/ “GMO VenturePartners. “Portfolio.”

[34]: https://dentsu-ip.co.jp/company.html “Dentsu Innovation Partners. “Company.”

[35]: https://www.jp-capital.jp/portfolio/ “Japan Post Capital. “Portfolio.”

[36]: https://www.hakuhodody-ventures.co.jp/news97/ “Hakuhodo DY Ventures. “News.”

[37]: https://initial.inc/articles/6qtZbz4VumqsGwd7rmbYRn “INITIAL. “Startup IPO Trends.”

[38]: https://www.jpx.co.jp/listing/stocks/new/index.html "JPX (Japan Exchange Group ). 'New Listing Information'." See also: https://www.jpx.co.jp/equities/listing-on-tse/new/basic/04.html

In 2025, the Tokyo Stock Exchange rewrote its Growth-market maintenance bar — from “¥4bn within ten years” to “¥10bn within five.” This piece asks, without wishful thinking, how that single move reshapes Japan’s startup ecosystem. The conclusion, stated up front: Japan has ended up with America's harsher funnel but none of its bigger winners — and not by importing it, but by self-imposing it: a domestic fix for the small-cap listing problem.

The same symptom, opposite meaning.

The starting point is the anomaly that Crunchbase has documented in the US. The median seed round has swelled to about $3m, three times its 2018 level — yet the share of companies advancing from seed to Series A has fallen from above 55% through 2020 to 24% for the 2023 cohort and just 16% for the 2024 cohort. Time to Series A now exceeds two years. Raise more, advance less: that is the state of US seed today.

Japan shows the same lengthening. According to Speeda’s Japan Startup Finance 2025 (it’s Japanese only), the median time from founding to seed rose from 12 months in 2021 to 21 months in 2025; seed-to-A from 13 to 16 months; A-to-B from 15 to 18 months; and founding to Series D and beyond from 87 months (7 years plus) to 109 months (9 years plus). The number of companies funded at seed fell from 1,245 to 855 — down roughly a third.

The symptoms may look similar, but the underlying structures differ greatly between the US and Japan. Let me start with the US. First, on timing. Two forces are at work here, operating on different layers: interest rates move the total quantity of capital, while AI moves its allocation. What first drove the collapse in Series A conversion was not AI but the Fed’s historic tightening, which began in March 2022. Excess liquidity reversed, and the water level of the whole ecosystem dropped. Because conversion is measured across cohorts, that shock surfaces with a lag — in the 2023–24 cohorts, whose graduation rates fell to 24% and 16% — and it squares with cohort analysis. Through investing in Infarm and running its Japan operation, I understand this firsthand.

One more thing deserves mention: the launch of ChatGPT (November 2022), which happened in the very midst of that tightening. The seed of AI was planted into a headwind, and when rate cuts began in September 2024, the water level rose again, rushing into the emerging terrain of AI. In other words, 2022–23 belonged to interest rates (a contraction in quantity), and 2025–26 belongs to AI (a concentrated recovery). The lead role changed hands; AI was not the cause of the collapse in startup investment.

Second, on concentration: the abundance is more skewed than ever. According to Crunchbase, global startup investment hit a record of roughly $297 billion in Q1 2026, with 81% of it going to AI. Just four companies — OpenAI, Anthropic, xAI, and Waymo — accounted for 64% of global venture investment in the quarter, and the US alone took 83%. Seed funding rose 30% year over year in dollar terms, yet deal counts fell 31% — the base did not widen; if anything, it was filtered and grew larger. As Keith Teare notes in That Was The Week, a handful of marquee firms — a16z, Sequoia, General Catalyst, Lightspeed — write most of the deals, and a company that cannot secure their backing finds it extremely hard to climb the ladder beyond seed.

And these two variables — rates as the tide, AI as the terrain — bring the US–Japan contrast into focus. In the US, the tide of interest rates powerfully moves startups and VCs; so when the tide came back in, the terrain of AI pulled capital back and concentrated it at the summit. Indeed, in the US, seed rounds themselves tripled, while Japan’s median seed has barely moved from around ¥50m. Japan has the talent and the technology to compete with the world. But it lacks a vast single market like America’s English-speaking reach or China’s billion-plus consumers, and although its large corporations’ retained earnings and its households’ financial assets are both ample, the risk that money flows to startups remains — at least for now — limited. The US bottomed out and rebounded; Japan’s contraction continues structurally. So — what should Japan do?

Added in response to a comment from Takeshi Ebihara (Rebright Partners) — that the main driver of the US Series A conversion collapse was the 2022 inflation/rate downturn, felt with a round lag and consistent with cohort analysis, with AI better seen as the 2025 bottoming-out. Data source: Crunchbase News, “Q1 2026 Shatters Venture Funding Records…,” https://news.crunchbase.com/venture/record-breaking-funding-ai-global-q1-2026/

Tokyo Stock Exchange’s 10 billion — rewriting the exit function.

In September 2025, the TSE finalized a revision to its Growth-market maintenance criteria: the old bar of “¥4bn market cap ten years after listing” becomes “¥10bn five years after listing.” It asks for 2.5 times the size in half the time. It applies from March 1st, 2030, to companies five years past their listing; those that miss it face delisting or relegation to the Standard market.

The numbers deserve care. The TSE’s own estimate is that about 30% (some 200 companies) sit in the affected ¥4–10bn band. A separate count finds that roughly 70% of companies remain below ¥10bn five years after listing. The first measures today’s danger zone among listed firms; the second measures the post-IPO hit rate. They are different denominators and worth keeping apart.

What matters most is that enforcement begins in 2030, but the market’s behavior has already changed. The listing criteria themselves were left untouched, yet underwriting now asks whether a company can credibly exceed ¥10bn. Growth-market IPOs in the first half of 2025 fell by roughly half year-on-year (34 to 18). Lead underwriters favor larger deals; smaller candidates struggle for support. It is not the rule’s start date but its anticipation that is already tightening the funnel.

The cool forecast — a barbell, not a recovery

This is not merely a listing rule. It rewrites the exit function that the entire capital stack was priced against. The Japanese model rested on the small-cap IPO as a cheap, early, reliable exit — a low-variance, low-ceiling system. The ¥10bn rule removes that reliable floor and turns the IPO into a conditional gate that 70% fail within five years.

The repricing cascades backward. First, the IPO stops being the default exit; M&A, MBOs, and secondaries become standard, and VC starts to behave like private equity. Second, if listings are delayed until ¥10bn is within reach, the path from founding to IPO lengthens and fund timelines break — the years to reach DPI of 1x have already risen from seven to ten, and a ten-year fund cannot wait fourteen. Third, the top of the funnel narrows: the share of funded companies less than a year old has collapsed from 55.5% in 2015 to 15.9% in 2025.

The ecosystem, therefore, splits into three: a smaller, higher-quality, concentrated venture track; a non-VC, non-IPO SME track; and a widening “valley of death” between them. The biggest structural risk is an LP doom loop. Japan’s LPs are mostly corporates, and their allocations are pro-cyclical — when returns disappoint, they withdraw. Post-2020 vintages sit below a net multiple of 1.0x; against a period when the TOPIX would have returned about 30%, ventures returned roughly 5%. Every rational reason for corporations to retreat is in place. For more details, please refer to the back issues here.

In short, Japan is being forced toward a high-variance model without the fuel that sustains it — a non-cyclical LP base and globally scaled exits. The old model is being dismantled faster than the new one can be built. For now, a multi-year air pocket is the honest expectation.

Where the openings are.

An air pocket is also an opening for those who build counter-cyclically. While everyone crowds the accelerator and seed entrance, the structural gap opens on the exit side. Stalled and at-risk companies accumulate; firms with real technology or IP but no domestic exit will need foreign strategic or financial buyers. Cross-border M&A and take-private bridges are counter-cyclical and uncrowded. Tellingly, advisers offering listings on the Singapore Exchange (SGX) have already appeared — the “overseas exit” escape route is forming.

The second gap is the global seam. Only companies with a path to a large outcome can attract the new, concentrated capital, and for most Japanese companies, that means non-domestic revenue. But a full relocation to Silicon Valley is usually the wrong, low-yield bet. The discipline is to match Japan’s structural strengths — deep tech, robotics, materials, manufacturing-adjacent software, content, healthcare for aging societies — to bridgeable large markets: India, Southeast Asia, the Gulf, and Europe.

And the breakthrough may be a team, not a hero. The odds that any single program produces one global breakout founder are low. A multinational team — a globally-native co-founder paired with Japanese deep tech and market access — is something you can build repeatably. Rather than waiting for a hero, manufacture the structure that can win.

Designing the exit — culture at the root, and the multinational team.

Why can’t Japan create big winners? Picture two startups with the same business model. One serves only the domestic market; the other has expanded into the English-speaking or Chinese market — or at least has the capability to. Investors will back the second, every time.

Over the past decade, investment into startups grew roughly fivefold and the founding environment matured — the “cost of manufacturing a startup” rose five times over. And yet most Japanese startups remain closed within the domestic market. That, I believe, is a major reason capital raised keeps climbing while IPO market caps do not — why capital efficiency collapsed 75% in seven years.

So what does it take to redesign a company that can expand globally? The obvious answer is “the language barrier” — but language is the effect, not the cause. When one company made English its official internal language, people who had spoken only broken English became fluent within two years. Under genuine necessity, capable individuals acquire the language. Germans and French do business in English because, in a contiguous single market, fairness leaves no alternative. The real barrier is not language but culture — whether a company can adapt its ways of working, decision-making, and communication to a different culture.

And that adaptation is unavoidable, whichever large exit you choose. Whether the aim is a large IPO that draws global institutional investors, or a cross-border sale to a foreign acquirer, the requirement is the same: the ability to run a business across cultures. It cannot be bolted on at the exit. A company raised mono-culturally within the domestic market cannot suddenly acquire it on the eve of a sale or listing.

If so, the answer lies in the founding design itself. Building with a multinational team from day one solves the entry problem — the constraints of the domestic market and culture — at the outset. But the benefit does not stop there. A multinational team is born as a company that foreign corporates and startups find easy to acquire. The friction that has blocked Japanese cross-border M&A — language, culture, the inability to integrate — sits, in truth, on the sell side. A multinational team removes that friction at the source. The design of the entrance becomes the solution to the exit: rather than arranging an exit after the fact, the company is born with the exit already open.

So far, the argument has been about going out into the world. But there is another answer — pointing in the opposite direction and homegrown: what Takahiro Mizumoto of ant innovations calls "restoring capital efficiency through industry consolidation." Over the past decade, investment into startups grew roughly fivefold, yet the average IPO market cap and the number of listings barely moved, which is why capital efficiency collapsed. His prescription is to replace the small-cap IPO: to give promising startups that would once have listed both growth capital and liquidity (an exit) for their existing shareholders, all while staying private. And to combine sub-scale companies, through carve-outs, MBOs and roll-ups, into entities large enough to clear the ¥10bn bar and command a higher multiple — what he calls “multiple magic.”

Two strategic responses to the same problem now sit side by side. One is to consolidate domestically until scale is reached (the Mizumoto / AIN path). The other is the one this piece has argued for: to go global from the start, with a multinational team. They are not in opposition; they are complementary. An entity that has restored scale and capital efficiency through domestic consolidation has a stronger base from which to go global, and a team designed to be global is a more valuable target not only for foreign buyers but for domestic consolidation platforms too. What matters is that, whichever path is chosen, both agree on one point: stepping off the old model that treats the small-cap IPO as the goal. The exit shifts from standing on one leg — the IPO — to standing on two: consolidation and the global market.

One design choice must be made on day one. If the base-case exit is cross-border M&A, the domicile of the holding company — a Japanese KK, Delaware, or Singapore — should be chosen to suit the intended buyer. It is almost impossible to change later and it governs which exits stay open. (Specific structuring should be confirmed with legal and tax advisers.) The TSE listing need not be pursued actively; become attractive enough and the exchange and the brokers will come to you.

A note on the data.

The principal sources are as follows. Months from founding to each round, and companies funded by round, are from Speeda’s Japan Startup Finance 2025 (as of January 2026). US seed sizes and graduation rates are from Crunchbase. Venture performance figures are drawn from Preqin and the JVCA. The capital-efficiency collapse and the consolidation argument draw on published materials and remarks by Takahiro Mizumoto of Ant Innovations. Note that earlier editions of the report were compiled by INITIAL rather than Speeda; cross-edition level comparisons are avoided, and this piece uses the internally consistent 2021–2025 Speeda series.

The older I get, the more I find myself thinking about things I never used to notice. On a recent morning jog, I caught myself wondering: Will I be able to see those cherry blossoms next year? That kind of thought simply didn’t exist when I was younger. Now it does.

It must be why 20 years feels both impossibly long and strangely close.

In 2006, I founded DreamVision with the ambition of building an EdTech startup. But after a thorough feasibility study, I put the whole thing in a drawer. The conclusion was painfully honest: the education industry was deeply entrenched, and I wasn’t yet capable of moving it. Looking back, I should have stopped right there.

I didn’t.

I had already started the company and reached out to people, so stopping felt impossible. What I’d call, in retrospect, a “mistake” — rooted not in judgment, but in immaturity. I was mentally weak, and I knew it.

I was 42 at the time. My eldest son had just been born. And I was running Interscope, an online research startup that had grown into something I was genuinely proud of. Alongside Macromill and Infoplant, we were considered one of the “big three” in the Japanese online research industry. I had even co-founded the Internet Research Forum of Japan with Mr. Otani of Infoplant, serving as its chairman, to help shape standards in a field that was still finding its footing.

But I was restless. The early-stage energy that had drawn me to the internet in the late 1990s — through the founding of Interscope and Webcrew, which went public on the TSE Mothers market in September 2004 — had begun to fade. Internet research has become an infrastructure. It no longer felt like a frontier.

What it had given me, though, was something more complicated: a reputation, a position, and the fear of losing them.

What pushed me forward anyway was my son. When I imagined him one day being old enough to ask about his father’s work, I kept returning to the same question: Which version of yourself would you rather explain? The one who protected what he’d built? Or the one who risked it for something he actually believed in?

I also gave myself a deadline of sorts. Most things, I’ve found, can be resolved in three years. Start at 42, fail at 45 — you can recover. Start at 45, fail at 48 — that’s a different kind of wound. I wasn’t sure I could come back from that one.

So I launched. And I failed. I closed DreamVision at 45, and spent the following year doing what I’d call “reading in sunny weather, reading in rainy weather” — a deliberate inversion of the old Japanese saying about diligence. It was, in its own way, exactly what I needed.

Someone recently commented on Facebook: “Congratulations on DreamVision’s 20th anniversary!”

Twenty years. I had to sit with that for a moment.

Around the same time, I watched an interview with Jensen Huang — NVIDIA’s co-founder and CEO. I’ll admit some hesitation quoting someone whose scale is so far beyond mine, but what he said resonated in a way I couldn’t ignore: “If someone asked me to go back and start the company again, I’d be too scared.”

Even Jensen. Even him.

I think most of my fellow entrepreneurs would say the same. What we’ve built is the product of coincidence and good fortune, shaped by decisions we couldn’t fully see at the time. If you asked me whether I could do it all again with full knowledge of what was ahead — honestly, I’d have no confidence. It’s terrifying, even now.

But Jensen also said something else: “It’s okay. You’ll learn what you need to know as you go along. Don’t underestimate the power of ignorance.”

Maybe that’s the only way any of us ever started.

Three Investments, One Pattern

Over the past six months — from last autumn through this spring — I made three new angel investments. On the surface, they have little in common. One is building an AI-powered credit infrastructure for immigrants. Another runs a travel business with operational excellence at its core. The third is creating a peer community platform for early-stage founders. Different sectors, different stages, different geographies.

But when I stepped back and looked at all three together, a pattern emerged — one I hadn’t planned, but one that turned out to be the clearest articulation of what I believe in as an investor.

All three use AI. All three were founded by people who left their home country to build something elsewhere. And all three have a deep connection to Japan — either through their founders’ roots or through where they chose to land.

Dwilar: When the Immigrant Becomes the Founder

I first met the founder of Dwilar at IVS Kyoto in 2024, through a mutual connection. Honestly, I wasn’t immediately drawn in. But when I heard his business model, something clicked.

Here’s his story. As a university student in Kobe, Yoshitaka Nakamura had built and operated over a dozen after-school learning centers for children of immigrants. These were bright kids — talented, capable — but because their Japanese language skills lagged behind, they couldn’t keep up in public schools. The city of Kobe recognized the problem and funded the program through grants. It was working. Then the mayor changed, and the grants disappeared.

He went on to work at Toyota and P&G, then earned his MBA at UC Berkeley. And that’s when the tables turned. Suddenly, he was the immigrant. No credit history. No local track record. Trying to rent an apartment or get a credit card in the US felt nearly impossible.

That lived experience became his startup. Dwilar uses AI to analyze immigrants’ economic data from their home countries — the financial track records that exist but are invisible to American landlords, property managers, and financial institutions. He is solving a real problem, born from real pain, on both sides of the Pacific.

Retreat Technologies: The Quiet Strength of Operational Excellence

I was introduced to one of the co-founders of Retreat Technologies through a mutual acquaintance — another Japanese founder building in the Bay Area. Their business is in the travel space.

Now, you might wonder: travel is a mature industry. Expedia, Booking.com, and other giants already dominate the market. Where is the opportunity? But here’s what the founders spotted — those incumbents are all B2C players. On the B2B side, there is a wide-open gap. Corporate travel, group retreats, team offsites — this is a space where operational excellence matters more than flashy consumer branding. And I thought: isn’t that precisely where Japanese founders have a natural edge?

What tipped the scale for me was the founders themselves, Shun Yamada and Jack Tadami. Two people of deep integrity, the kind you trust to hold steady when things get hard — and in startups, things always get hard.

I didn’t have a grand strategic thesis when I invested. It was more of an intuition — a sense that this company would connect to something larger. And as it turns out, pieces are starting to come together in ways I couldn’t have predicted at the time.

Entention Social: Diamonds in the 97%

The third investment has the longest backstory.

The founder, Philip, is a Vietnamese-American who earned his PhD in nuclear engineering at Kyoto University. He fell in love with the city and stayed. I first backed him years ago, when I was running a fund at Sunbridge in partnership with Hankyu Railway. That company was Gochiso — a platform where dining at participating restaurants would automatically direct a percentage of your bill to an NPO of your choice. It was a beautiful idea rooted in his passion for social impact, but it didn’t scale. He put it into hibernation.

But Philip kept going. His next venture was a product to help small businesses visualize and display their CO2 reduction efforts — a sustainability play. It got accepted into an Estonian accelerator, and things were moving forward. Then Trump 2.0 happened. Across the US, renewable energy startups that had been counting on government grants saw their funding evaporate. Philip had to pivot again.

What he landed on this time, I believe, is the most promising idea yet. His insight is simple but powerful: seed-stage founders are overwhelmed and isolated, with no one to talk to. The accelerators — YC, Techstars, and the rest — serve roughly 3% of startups. But what about the other 97%? Among them, there must be diamonds in the rough.

Entention Social is building a peer community where early-stage founders can support each other — practically and emotionally — without needing to win the accelerator lottery first.

This resonated with me on a personal level. When I founded DreamVision in 2006, I originally wanted to build an EdTech platform — specifically, a platform where practitioners learn from each other. I did my own feasibility study and concluded that Japan wasn’t ready. The idea of embedding entrepreneurship into university curricula seemed impossible at the time. I shelved it.

Fifteen years later, in 2021, Yoichi Ito invited me to help create the School of Entrepreneurship at Musashino University. I said yes immediately. It felt like a message from the universe — the idea I had abandoned had come back in a form I never expected.

So when I saw what Philip was building with Entention Social, I felt the same resonance. The founder who refuses to quit. The idea that keeps finding new shapes. And at its core, a belief I share: where diamonds shape themselves.

What Connects Them

I didn’t set out to build a portfolio around a thesis. These three investments found me — or perhaps I found them — through the network of relationships I’ve built over 25 years as an entrepreneur and investor in Japan.

But when you look at them together, the thesis is clear.

The most interesting founders I meet today are not the ones optimizing for a single market from within. They are the ones who have left home — by choice or by circumstance — and are building at the intersection of cultures, languages, and systems. They carry the discipline and integrity of their roots, but they refuse to be confined by geography.

Japan, as I have written before, has a structural handicap as a relatively mono-ethnic nation in a globalizing world. But these three founders — and many others like them — are the living proof that the handicap is not destiny. The future of Japanese entrepreneurship may not come from Tokyo alone. It may come from Kyoto, from Berkeley, from Tallinn — from wherever founders with a connection to Japan choose to build.

Home is not the point. The building is.

Thanks for reading. If you know founders building across borders with a connection to Japan, I’d love to hear from you.

Between 2025 and 2026, global equity markets and startup ecosystems have reached a historic turning point.

Today, I want to explore the current state and future prospects of Japan’s startup market — drawing on macroeconomic data, comparative analyses of emerging stock markets across Asia, and the latest global VC investment trends.

The Macro Landscape Across Asia

To understand the dynamics of emerging stock markets, we first need to look at the underlying macroeconomic and digital environments. Here’s how five major Asian economies compare [1] [2] [3] [4]:

Note: According to “Japan Startup Finance 2025 (Speeda),” total VC investment in Japan in 2025 (including venture capital, operating companies, financial institutions, and overseas companies) was 884 billion yen (approximately $5.6 billion). Limiting it to VCs, the figure was 317.1 billion yen (approximately $2 billion). Other countries’ figures are estimates based on Crunchbase and national startup databases.

Singapore, with a population of just 6.2 million, boasts a GDP per capita of $88,447 and absorbs 91% (~$4.9B) of Southeast Asia’s VC investment. Eighteen unicorns. A tiny nation with an outsized presence.

Japan has a massive economy ($4 trillion) and substantial startup investment (~$5.6B). Yet only 8 unicorns. For an economy this size, that number is conspicuously low. There is clearly a structural challenge in scaling startups into mega-ventures.

Indonesia, Malaysia, and Thailand are seeing rapid internet adoption, but GDP per capita remains low, and VC investment has been sluggish. In these markets, investors favor established, profitable companies in traditional sectors over pure tech plays.

Japan’s Market “Depth”

These macroeconomic differences directly shape the structure of emerging stock markets in each country. A comparison reveals something distinctive about Japan’s TSE Growth Market [5]. Often dismissed as a venue for “micro-cap IPOs,” the Growth Market actually holds a clear advantage over its Southeast Asian peers.

The TSE Growth Market commands approximately $55.3 billion in market capitalization — dwarfing its regional peers — yet concentration among the top 10 companies sits at just 25.9%. This is a deep, diversified market, not one propped up by a handful of mega-ventures.

Singapore’s SGX Catalist, by contrast, is highly concentrated at 52.0%. And the SGX’s key players are REITs and mega-banks — capital that favors stable, high-dividend yields. As a result, the Amazon-style model of reinvesting all profits into growth and running deep losses for years — the kind of trajectory common in SaaS and deep-tech — simply doesn’t get valued by SGX investors. It gets avoided.

This is consistent with what a friend of mine in Singapore told me a few years ago: “It’s better to define your company as a ‘Financial Service’ rather than ‘FinTech’ — investors feel more comfortable with that.”

In Malaysia and Thailand, the top tiers are dominated by mid-sized companies in traditional industries that generate stable profits. When the macroeconomic environment is still developing, investors understandably prioritize current profitability over future growth potential.

Two Structural Challenges Facing the TSE Growth Market

Despite its scale and diversity, the TSE Growth Market is in trouble. While the Nikkei 225 rallies, the Growth Market is stuck in a severe downturn. Two structural forces explain this bifurcation [6].

First, capital flight is driven by macro shifts. Rising geopolitical risks and global interest rate concerns have pushed institutional investors away from low-liquidity, high-risk small-cap growth stocks and toward safer large-cap names on the Prime Market.

Second, the TSE tightened its listing maintenance criteria. As is well known, companies must now reach a market cap of ¥10 billion within five years of listing. The old playbook — going public at ¥3–5 billion — is dead. IPO hurdles have risen sharply, and the number of new listings between 2024 and 2025 has collapsed. Japan is in an “IPO winter.”

The “AI-Dominant” Global Investment Landscape

The global picture is even more dramatic.

My friend from the SunBridge days, Keith Teare — his wife, Gené Teare (Crunchbase News’ senior data editor) — published the latest Crunchbase report showing that global VC investment in Q1 2026 hit a record $300 billion [7].

But look at the breakdown. A staggering 80% ($242 billion) went to AI-related companies. And 65% of that AI funding ($188 billion) was monopolized by just four U.S. companies: OpenAI ($122B), Anthropic ($30B), xAI ($20B), and Waymo ($16B).

U.S. companies captured 83% ($250 billion) of the global total. This is “AI Dominance” and “U.S. Dominance” — a winner-takes-all reality.

Breaking Away from Isolation — The Challenge of “United Characters of Japan”

In this environment — U.S. AI dominance globally, tightening domestic standards at home — where does Japan’s startup ecosystem find its path forward?

I believe Japan must completely break free from its overwhelming inward-looking mindset and embrace diversity.

Currently, Japan’s stock market is extremely domestic. As of April 2026, of the 3,924 companies listed on the Tokyo Stock Exchange, only 5 are foreign [8]. One on Prime, two on Standard, two on Growth. At its peak in 1990, the TSE hosted 125 foreign companies. That’s a 96% decline over 35 years. Near extinction.

When I asked Claude Opus 4.6 why this happened, it identified six key factors:

1. The inherent limitations of the “dual listing” model

2. IT advances rendering dual listings meaningless

3. The Japanese language barrier — disclosure, IR, analyst coverage

4. Japanese investors not buying foreign stocks

5. The TSE’s retreat from active recruitment

6. Currency risk and the long-term instability of the yen

In a sense, these are all rational. For companies already listed on the NYSE, NASDAQ, or the London Stock Exchange, paying the high cost of maintaining a dual listing on the TSE no longer makes sense. Meanwhile, many Japanese companies are dual-listed on the NYSE or NASDAQ — a fact that speaks volumes about the relative power of these markets.

And it gets worse. PayPay — the digital payment service used by over 65 million people in Japan — bypassed the TSE entirely and listed on NASDAQ. Not only are foreign companies leaving the TSE. Japanese-born companies aren’t choosing it either.

Meanwhile, the NYSE hosts over 530 international companies [9], and NASDAQ lists more than 1,500 foreign firms [10].

The TSE Prime Market probably cannot compete directly with the NYSE or NASDAQ. But the Growth Market — the venue for emerging companies — could absolutely attract foreign startups, particularly from Southeast Asia, if the TSE were truly committed to it. I believe it’s a realistic possibility.

I say this from experience. In 2012, when I first visited Singapore and attended Echelon, the startup conference organized by e27, underwriters for the TSE Mothers market (the Growth Market’s predecessor) had set up booths and were actively competing with Singapore and Hong Kong to attract Asian startup listings. But somehow, those efforts vanished. The market retreated into its shell.

To attract promising Southeast Asian startups, Japan needs English-based screening processes, English-language analyst reports, and IR support with English-Japanese translation. These investments won’t pay off immediately. It may take a decade of persistent effort before they become viable as a business. But in Japan, facing a certain population decline, a stock market that operates only in Japanese will never globalize its emerging market, not in an industry as inherently borderless as finance.

At the root of this problem is an overwhelming lack of awareness about the “handicap” of being a linguistically homogeneous nation.

I know this from personal experience. I once invested in and managed the Japanese subsidiary of Infarm, a Berlin-born startup. It was a highly diversified and globalized organization — approximately 1,000 employees representing 50 to 60 different nationalities. Although the company was founded in Germany, all three founders were from Israel. Less than 30% of the workforce was German. Native English speakers made up only 20–30%. Yet the official company language was, naturally, English.

After experiencing that environment, I’ll be honest: the typical Japanese startup — where everyone is Japanese and speaks only Japanese — feels like going back 50 years to watch a black-and-white TV.

At a public international school in Berlin, children of over 17 nationalities learn together and routinely host school festivals featuring their national cuisines. Children raised in that kind of diverse environment naturally develop cross-cultural understanding and communication skills. They approach others with the assumption that “this person is different from me.” By 18, the world they see is entirely different from that of children raised in a homogeneous environment.

Fundamentally changing this in Japan is extraordinarily difficult. But doing nothing is tantamount to sitting and waiting for death.

Consider this. The Japan national rugby team includes many foreign-born players. Michael Leitch served as captain. No one in Japan complained about that. The business and financial worlds should be identical.

Diverse “characters” — individuals of different nationalities, cultures, and expertise — converging on Japan’s playing field. “United Characters of Japan.” A place where Japan’s unique strengths — its manufacturing base, its robotics, its craftsmanship — fuse with diverse talent and capital from around the world.

I believe that this is the singular, defining challenge for Japan’s startup ecosystem: establishing a global presence in the face of America’s AI dominance.

References

[1] World Bank. (2024). World Development Indicators: GDP and GDP per capita. [2] INITIAL Inc. / STARTUP DB. (2026). Japan Startup Finance 2025 Annual Report. [3] Hurun Research Institute. (2025). Global Unicorn Index 2025. [4] International Telecommunication Union (ITU). (2024). Percentage of Individuals using the Internet. [5] Manus AI. (2026). Comparative Analysis Report of Major Asian Emerging Stock Markets. [6] Manus AI. (2026). TSE Growth Market Outlook and Structural Challenges: Latest Trends in 2026. [7] Crunchbase News. (2026). Q1 2026 Shatters Venture Funding Records As AI Boom Pushes Startup Investment To $300B. [8] Japan Exchange Group (JPX). (2026). Number of Listed Companies/Shares. [9] New York Stock Exchange (NYSE). (2026). International Listings. [10] Ritter, J. R. (2026). The Number of Operating Companies Listed on Nasdaq. University of Florida.

I am one of the entrepreneurs from the so-called dot-com bubble.

The top startups by market capitalization listed on the former Tokyo Stock Exchange Mothers market (now the Growth market) were founded around the same time, such as CyberAgent, DeNA, and mixi, but they have since moved to the Prime market.
So, what kind of startups are making their presence felt in the growth market?

I’ve picked out the TOP 20 companies by market capitalization as of March 31, 2026, and analyzed the growth and decline in their market capitalization and operating profits.
That was very interesting. Unlike in our time, internet-related companies and SaaS companies have lost their presence.

Before we dive into individual companies, however, we must confront an uncomfortable but important truth about the market they operate in.

Note: Exchange rates used for conversions are ¥149.52/USD (March 31, 2025) and ¥159.88/USD (March 31, 2026), based on Mitsubishi UFJ Bank TTM rates.

The Bigger Picture: Growth Market Underperformance

The TSE Growth Market has long been positioned as Japan’s premier venue for high-growth startups. Yet over the past two years, a troubling divergence has emerged when comparing it against the Prime and Standard markets.

Using January 2024 as the baseline (= 100), the cumulative change in total market capitalization through March 31, 2026, tells a revealing story (Source: JPX Market Capitalization Historical Data):

At first glance, the Growth Market’s +27.7% cumulative gain appears competitive. However, the trajectory tells a very different story. Growth Market capitalization surged ahead of Prime in early 2024 and reached a peak of approximately +38% by mid-2025 — only to reverse sharply in the second half of 2025. By contrast, the Prime Market accelerated steadily from mid-2025 onward and overtook growth by early 2026.

In absolute terms, the disparity is even more stark:

The Growth Market rose 13.7% year-on-year, well below the Prime Market’s 28.5%. The entire Growth Market — with its 603 listed companies — has a total market capitalization of just ¥8.84 trillion ($55.3B), equivalent to roughly 0.75% of the Prime Market. This structural imbalance reflects a broader reassessment by investors: capital is rotating away from speculative growth stories and toward companies with demonstrated, scalable profitability.

The average market capitalization per listed company on the Growth Market remains modest:

Against this backdrop of market-wide underperformance, the Top 20 companies profiled below represent a select group that has managed to outpace — or at least hold its own against — a challenging environment for growth-stage equities.

Top 20 Startups by Market Capitalization (as of March 31, 2026)

The following table ranks the top 20 companies on the TSE Growth Market by market capitalization as of March 31, 2026, and compares them with their valuation one year prior.

Notably, three companies in this list — PowerX, HUMAN MADE, and Northsand — went public in late 2025 and were therefore not yet listed in March 2025.

Key Takeaways on Valuation

The Undisputed Leader. TRIAL Holdings (141A) has cemented its position at the top, nearly doubling its market cap to over ¥515.4B ($3.22B) — a gain driven by consistent profitability improvement in its retail-tech business.

Explosive Growth from the Emerging Tier. BuySell Technologies (+133.3%), Next Generation Technology Group (+162.2%), and QPS Holdings (+273.8%) demonstrated extraordinary valuation growth, each benefiting from a combination of strong earnings momentum and expanding investor narratives around Secondhand Retail, AI infrastructure, and satellite technology.

The IPO Impact. Three late-2025 entrants — PowerX, HUMAN MADE, and Northsand — debuted directly into the Top 20, illustrating that the market still rewards compelling IPO stories, even as the overall Growth Market environment has become more selective.

Significant Corrections Among Former Leaders. freee (-42.4%), GENDA (-48.2%), and COVER Corporation (-44.2%) all experienced sharp declines in market cap. These corrections reflect a broader investor shift away from “growth-at-any-cost” narratives. In GENDA’s case, a modest decline in operating profit was compounded by investor concerns about repeated equity dilution, contributing to a 48% collapse in its market cap.

Operating Profit & Margin Analysis

Valuation is only half the story. The table below compares operating profit and operating margin for each company’s most recently completed fiscal year versus the prior year.

Note: Fiscal year-ends vary by company. "Latest FY" refers to the most recently completed full fiscal year as of the publication date.

Key Takeaways on Profitability

The Profitability Pivot. freee (4478) achieved a historic milestone — swinging from an ¥8.39B ($56M) operating loss to a ¥611M ($4M) profit. Yet the market's reaction was a sharp sell-off, because the magnitude of the profit fell well below analyst expectations and the forward guidance implied only modest margin expansion. This episode encapsulates the Growth Market's current challenge: achieving profitability is no longer sufficient; the pace and scale of that profitability must also meet elevated expectations.

Margin Leaders. HUMAN MADE (456A) stands out with an operating margin of 31.7%, reflecting exceptional brand pricing power. Northsand (446A) and LA Holdings (2986) also maintain robust margins above 20%, demonstrating that high-margin businesses can thrive even in a risk-off environment.

Deep Tech Continues to Burn Cash. Space and satellite companies — Astroscale (186A) and Synspective (290A) — along with biotech firm SanBio (4592), continue to post significant operating losses. This underscores the increasingly bifurcated nature of the Growth Market: highly profitable consumer/software businesses on one end, and capital-intensive frontier-tech ventures on the other.

The Profit-Valuation Correlation Tightens. MTG's 225.4% profit surge directly correlates with its 118.1% market cap gain. Conversely, GENDA's modest decline in operating profit — compounded by concerns about equity dilution — contributed to a 48.2% collapse in market cap. The market is increasingly unforgiving of companies that fail to translate revenue growth into bottom-line results.

Conclusion: The Growth Market at a Crossroads

The TSE Growth Market in 2026 is navigating a “critical inflection point”. As the broader data shows, the market has underperformed the Prime Market on a year-over-year basis, with total market capitalization growing just 13.7% compared to Prime’s 28.5%. This relative underperformance reflects a structural shift in investor sentiment: capital is migrating toward companies with proven, scalable profitability — and the Growth Market, by definition, is still home to many companies that have not yet crossed that threshold.

Within this challenging environment, the Top 20 companies profiled here represent a diverse spectrum. At one extreme, businesses like TRIAL Holdings, MTG, and BuySell Technologies have demonstrated that disciplined execution and strong profit growth can generate extraordinary returns even in a risk-averse market. At the other extreme, former market darlings like freee, GENDA, and COVER Corporation have been sharply re-rated as investors recalibrated their expectations around profitability timelines.

The central lesson from this year’s data is clear: in today’s market, growth alone is no longer a sufficient investment thesis. The companies that will define the next chapter of Japan’s startup ecosystem are those that can combine top-line growth with meaningful margin improvement — and clearly communicate that trajectory to the market.

Coming Next: The Great Divergence — Why Japan’s Growth Market Peaked and Stalled While Prime Soared.

In the first half of 2025, Japan’s TSE Growth Market surged with optimism. By the second half, it had quietly lost its footing. Meanwhile, the Prime Market climbed to historic highs — steadily, almost defiantly. What drove this stark divergence? In our next issue, we dissect the three shocks and one euphoria that reshaped Japan’s equity landscape in 2025: the Trump tariff shock, the Bank of Japan’s historic rate normalization, the second wave of the AI infrastructure boom, and the Takaichi Trade. A story about how macro forces — not company fundamentals — rewrote the rules of Japanese equity investing.

Thank you for reading. Subscribe for more insights into the Japanese startup and public market ecosystem. So stay tuned.

References

• [1] Mitsubishi UFJ Bank. Foreign Exchange Rates (TTM) as of March 31, 2025. https://www.murc-kawasesouba.jp/fx/past/index.php?id=250331

• [2] Mitsubishi UFJ Bank. Foreign Exchange Rates (TTM) as of March 31, 2026. https://www.murc-kawasesouba.jp/fx/past/index.php?id=260331

• [3] Japan Exchange Group (JPX). Market Capitalization by Market Segment — Historical Data. https://www.jpx.co.jp/markets/statistics-equities/misc/02.html

• [4] Japan Exchange Group (JPX). Number of Listed Companies (Historical Data). https://www.jpx.co.jp/listing/co/index.html

• [5] Yahoo Finance Japan. Corporate Performance and Financial Data for respective tickers. https://finance.yahoo.co.jp/

Fair enough. For the Reiwa generation — those now 18 to 23 years old — the Showa era might as well be ancient history. My grandfather was born in the Meiji era, two eras before mine. That’s the generational gap I live in every day.

Narrative #1: Fear Sells

Here in Japan, where population declines and labor shortages are daily realities, the prevailing discourse around AI is surprisingly optimistic — AI as a dramatic productivity booster, not AI as a job-killer. At least in my circles, I rarely hear anyone sounding the alarm.

And yet, in the United States — supposedly the world’s AI frontrunner — fear-based narratives are alive and well. A film titled “The AI Doc: Or How I Became an Apocaloptimist” was released on March 27th of this year. (Incidentally, that date happens to be the anniversary of my very first company incorporation. Life is full of strange coincidences.)

The logic is simple: good news doesn’t sell. Pessimism does. That’s Narrative #1.

Narrative #2: OpenAI Is Orchestrating Its Own Story

The second narrative has a clear protagonist: OpenAI.

You may have already seen the news. This week, OpenAI acquired TBPN, a tech-focused show attracting 70,000 daily viewers. What caught my attention is where this acquisition was placed organizationally — not under PR, but under the Strategy division led by Chris Lehane, OpenAI’s Chief Political Operative and the architect of Fairshake, a crypto-funded super PAC designed to influence U.S. elections.

Veteran tech journalist and investor Om Malik called it out bluntly: “You don’t put an editorially independent media property under your political operative.”

So why did OpenAI move now?

Because OpenAI is heading toward an IPO — potentially this year.

On March 31st, the company closed a fundraising round at a post-money valuation of US$852B (approximately ¥136 trillion at ¥160/USD), raising US$122B (approximately ¥19.5 trillion). Amazon committed up to $50B in investment, while OpenAI announced a massive cloud deal with AWS worth approximately $100B.

But the detail that struck me most was that ARK Investment has included OpenAI shares in one of its ETFs, giving ordinary retail investors indirect exposure to OpenAI for the first time.

This matters enormously. ETFs typically hold only publicly listed stocks. OpenAI is not public — and it’s not just any private company you can write a check to.

OpenAI’s corporate structure is deliberately complex. At the top sits the OpenAI Foundation, a non-profit entity. Beneath it is OpenAI Group PBC (Public Benefit Corporation), which is the entity that actually runs the commercial operations — ChatGPT and everything else you interact with. Any transfer of shares in OpenAI Group PBC requires approval from the board of directors. And investing directly in OpenAI — or in the VC funds that back it — is restricted to Accredited Investors. For regular retail investors, getting into OpenAI, Anthropic, or SpaceX is essentially impossible unless you happened to know the founders personally from the very beginning.

That gap is exactly what Cathie Wood’s ARK Investment is trying to bridge — delivering the narrative of tomorrow’s most promising private tech companies directly to retail investors and creating demand before the IPO even happens.

One more piece of the puzzle: OpenAI has secured a ~US$4.7B credit facility (currently undrawn) from a syndicate of major banks, reportedly including JPMorgan Chase, Goldman Sachs, Citigroup, Morgan Stanley, and, interestingly, SMBC (Sumitomo Mitsui Banking Corporation). For the investment banks, this is their entry ticket to becoming lead underwriters on what could be the most significant IPO in years.

None of these moves is isolated. As Om Malik wrote in his essay “The New Announcement Economy,” it’s all about distributing the story. This is Sam Altman’s carefully orchestrated demand creation strategy in preparation for the IPO.

Narrative #3: The Democratization of Private Markets — and Its Dangers

The third narrative flows directly from the second: financial institutions selling “access” to private markets to retail investors.

Let me introduce a company called Fundrise, a fintech firm founded in Washington, D.C. in 2012. Its mission: open up private asset investment — historically reserved for institutional investors — to ordinary individuals. The core of its model is the democratization of private markets powered by technology. Through its app and web platform, anyone can invest with as little as $10 (roughly ¥1,500), with an annual management fee of 1.85% and — notably — no carried interest.

In July 2022, Fundrise launched the “Innovation Fund,” and on March 19, 2026, it listed the fund on the NYSE as a closed-end fund under the ticker symbol VCX — tagline: “The public ticker for private tech.”

The fund holds shares in companies such as OpenAI, Anthropic, SpaceX, and Databricks — names otherwise accessible only to institutional investors and top-tier VC funds.

What happened next was extraordinary.

Within three days of listing, VCX’s market cap surged roughly 30x. Within just seven trading days, the share price hit $575 — a +1,840% gain from the listing price. The problem? The fund’s net asset value (NAV) was only $18.26 per share. The market was pricing VCX at roughly 31x NAV. An impossible premium.

Then Citron Research stepped in, publishing a report warning of a 93% downside risk. The stock dropped 45% in a single day to $262. Today, it trades between $113 and $ 130 — still carrying a premium of roughly 6–7x NAV.

Why? Technically, you can point to the closed-end fund structure — unlike a conventional ETF, the number of shares is fixed, so new shares can’t be issued to arbitrage the price back toward NAV. There’s also a six-month lockup for early investors. But the real reason is more fundamental: retail investor behavior driven by the narrative of “access” — access to private markets that were never available to them before.

What This Means for Japan

Back home, the Tokyo Stock Exchange has tightened the listing maintenance standards for its Growth Market, requiring startups to reach a market cap of ¥10 billion within five years of IPO. The old playbook — listing at ¥3–5 billion — is officially obsolete. The numbers reflect the shift: while the number of new VC funds established in 2025 ticked up slightly, the total capital raised by those funds collapsed to less than half of the 2023 figure. More funds chasing far less money. That's not a healthy ecosystem signal.

With global investment attention almost entirely captured by AI — and the center of gravity firmly planted in the U.S. and China — Japan simply cannot compete on LLM infrastructure. We need to think seriously about where Japan’s winning moves actually lie.

A note: This piece draws on analysis and reporting from ThatWasTheWeek, the newsletter by my deeply respected Keith Teare. The research was conducted with the assistance of Manus AI, ChatGPT, and Claude. Keith’s newsletter is long-form and demanding to read, but if you’re serious about understanding the global tech and venture landscape, I strongly recommend subscribing.

Thanks for reading “Be Bold or italic, never regular.” Subscribe for free to receive new posts and support my work.

For years, excess liquidity mass-produced unicorns, and the US market celebrated its spring. But it has been a few years since the Russian invasion of Ukraine and the Federal Reserve’s interest rate hikes shifted the market into reverse. The IPO market cooled, M&A slowed, distributions to LPs stalled, new fund formations dropped, and startup fundraising became extremely difficult.

Amidst this, OpenAI released ChatGPT-3.5 on November 30, 2022. It recorded the fastest growth in history at the time. Fast forward to a year ago, in the first quarter of 2025, OpenAI raised US$40B at a staggering US$ 300 B valuation. That single deal accounted for more than half of all US venture capital investment and a third of global investment in that quarter, pushing the Q1 global startup investment total to $113 billion (Crunchbase).

Meanwhile, the fundraising environment for typical startups remains incredibly harsh.

So, what does the IPO market look like today?

According to a blog post by Jason Lemkin of SaaStr, 174 companies IPO’d in the first half of 2025 (including non-startups), the highest number since 2021, raising US$31B.

Everyone said 2025 was the year the IPO window reopened. And technically, it did. 174 companies raised over $31 billion in the first half alone — the highest since 2021.

However, most B2B and tech companies that went public in 2025 are currently trading below their IPO prices. Can we really say that 2025 was the year the “IPO window” reopened?

Below is a graph of market capitalizations for major B2B and tech startups as of February 23, 2026, from Jason Lemkin's blog post.

What does this reality tell us? I used Manus AI to investigate the details. Here is the summary:

2025 was touted as the “Year of the IPO Revival,” but the reality was almost a complete defeat. Out of the 13 major B2B companies that went public, only three (CoreWeave, Hinge Health, and Circle) are trading above their IPO price. The vast majority have plummeted, with Klarna sinking to 69% below its offering price. The market only values pure AI infrastructure; simply claiming to be an “AI-leveraged” or “AI company” no longer resonates with investors. Furthermore, first-day pops have almost universally vanished, and rapid sell-offs by institutional investors have become the norm. Solid revenue growth alone can no longer justify high valuations; GAAP profitability or fundamental involvement in AI infrastructure are now absolute prerequisites.

What kind of companies are the three that have exceeded their IPO market cap?

1. CoreWeave (CRWV) — The Overwhelming Winner in AI Infrastructure

· Business: GPU-specialized cloud infrastructure (for generative AI)

· IPO Date: March 2025 (NASDAQ) at $40

· Current Status: ~$80 (+100%)

· 2025 Revenue: $5.13B (YoY +170%), the fastest cloud company to reach $5B.

· Evaluation: The only company in the 2025 IPO class recognized by the market as pure “AI infrastructure.” They clearly distinguished themselves from others just claiming “AI utilization.”

2. Hinge Health (HNGE) — The Quiet Winner in Digital MSK Care

· Business: AI-powered MSK (musculoskeletal) rehab/physical therapy platform

· IPO Date: May 2025 (NYSE) at $32

· Current Status: ~$41 (+28%)

· 2025 Revenue: ~$550M with an 85% gross margin.

· Evaluation: SaaStr called them “the quiet winner nobody talks about.” They achieved GAAP profitability in Q1 2025 and won public market trust through solid financials, not AI buzzwords.

3. Circle Internet Group (CRCL) — Riding the Stablecoin Wave

· Business: USDC stablecoin issuance and blockchain financial infrastructure

· IPO Date: June 2025 (NYSE) at $31

· Current Status: ~$105 (+240%)

· Evaluation: A pure crypto/stablecoin play driven by regulatory tailwinds from the Trump administration, though highly volatile due to legislative movements.

By the way, do you believe the figure of “174 IPOs” in the first half of 2025? I couldn’t quite believe it, so I asked Manus to dig a little deeper.

It turns out the “174 companies / $31B” figure includes a broad definition of IPOs, including SPACs, micro-caps, and foreign companies listing in the US (ADRs).

Looking at the breakdown of the 347 total IPOs for the full year of 2025, traditional IPOs accounted for about 60%, while SPACs made up about 40%.

Even among “Traditional IPOs,” definitions vary. Professor Jay Ritter (UFL), whose data is most widely cited in academia, excludes SPACs, micro-caps (<$5), ADRs, REITs, and banks. By his strict definition of “pure operating company IPOs,” there were only 90 IPOs.

SPACs are essentially “empty boxes” at the time of listing, designed to find a merger target within two years. In the 2021 SPAC boom, 50% were eventually liquidated (funds returned to investors), and the 2025 vintage carries the exact same risk.

I can say this now, but Infarm, a company I invested in and managed the Japanese entity for, was actually slated to list on the NASDAQ via a SPAC in 2021.

However, we watched our peer, AppHarvest, complete its SPAC merger in February 2021, surge to $37.64, then plummet to under $6 by October (it eventually went bankrupt in 2023). We also saw AeroFarms attempt a SPAC listing, only for the deal to fall through before listing because the sponsor couldn’t raise the minimum required cash. Given these realities, Infarm canceled its SPAC IPO. (Infarm later raised US$200M in a Series D, but unfortunately filed for bankruptcy in September 2023.)

41% (141 deals) of the 347 IPOs in 2025 were SPACs—empty boxes with no merger target. And among the 206 traditional IPOs, small deals (under $50M) were almost entirely wiped out by year-end. The fact that only “large, high-quality deals yielded a +21% return” shows just how critical selection has become.

In Japan, the Tokyo Stock Exchange (TSE) is reforming its Growth Market, requiring companies to maintain a market capitalization of 10 billion JPY ($67M) within five years of their IPO. This is drastically changing the environment for Japanese startups, but the US stock market is incomparably more ruthless. The survival of the fittest is severe.

For the Japanese startup ecosystem to develop further, it must attract global capital. And to do that, startups cannot just target the 120 million people in the Japanese market. They must target the 1.7 billion people in the English-speaking market, or the 1.4 billion in the Chinese-speaking market. I don’t need to tell you, though LOL.

The exchange rate is now around 160 JPY/USD, and the yen will likely weaken further. In other words, a business model that earns foreign currency (dollars) is absolutely essential. Clinging to the Japanese home market will not bring a bright future.

As an entrepreneur and investor born and raised in Japan, I intend to face this heavy proposition head-on.

The key to everything is to take action. It all starts from there.

Personally, I have high hopes for the new political party, Team Mirai (Future), to make great strides, but it seems that the existing opposition parties have reached their “best before” date. The Liberal Democratic Party remains the only major political party.

In the world of startups, Silicon Valley is the dominant force, so it is nonsense to be conscious of Silicon Valley or try to copy a Silicon Valley-like ecosystem in your own country.

I have previously cited Japan Startup Finance (in Japanese) reports several times to introduce the fundraising landscape for startups in Japan.

This report, created using Manus (AI) and adding my analysis, provides a comparative analysis of the startup investment landscapes in five key economies: the United States, the United Kingdom, France, Germany, and Japan.

Using 2024 data, it examines structural differences and characteristics across these ecosystems from multiple perspectives, including total investment scale, deal quality and quantity, investor demographics, and the composition of capital sources for venture capital (VC) funds (LP composition).

As is well known, the most striking feature of the Global Startup Landscape is the United States’ (most likely Silicon Valley’s) sheer dominance. In 2024, total US investment reached $215.4 billion, approximately 6.5 times the combined total of the other four countries (around $33.3 billion) [1].

This “US dominance” pattern illustrates fundamental differences in the maturity of startup ecosystems, market size, and the supply of risk capital.

In terms of the number of deals, which indicates the “quantity” of investment, the United States stands out with 14,320. In Japan, 2,869 companies have raised funds of known amounts, the same as in the UK (approximately 2,800 deals) and significantly higher than in France and Germany. [2] [3] This is due to the high rate of startups in Japan and the active fundraising at the seed and early stages.

However, the picture changes dramatically when looking at deal size, a proxy for investment quality and scale.

Average Deal Size: US (15.0M) and France (15.1M) lead, indicating markets driven by large, growth-stage rounds. Japan’s average is the smallest at $2.3M.

Median Deal Size: The median, which better reflects the typical deal, reveals an even starker contrast. France leads with 4.5M, followed by the U.S. (3.5M), the UK (3.2M), and Germany (3.0M).

Japan’s median of $0.35M is an order of magnitude lower than the others, highlighting the market’s heavy concentration in seed and early-stage investments [4] [5].

Also, I believe the more fundamental reason is that, with very few exceptions, Japanese startups target only the Japanese market, which keeps their Total Addressable Market (TAM) small and prevents them from attracting growth capital from overseas.

Who is funding these startups? The breakdown of investor types reveals the underlying structure of each ecosystem. (Note: The following comparison uses the effective share of identified investors, excluding the “Other/Unknown” category, which varies by data source.)

US (Independent VC-Led): Independent VCs are the core of the ecosystem, accounting for 55% of investment. While CVCs are active (23%), independent firms clearly lead the market.

UK (Global Financial Hub): While led by independent VCs (40%), the UK has a high share of foreign VC investment (20%). This reflects London’s role as a European financial hub attracting global capital [6].

France (Government-Led): Government and public agencies (like Bpifrance) have the highest share among the five countries at 16%. The high proportion of foreign VC investment (21%) suggests that strong government initiatives are successfully attracting international capital [7].

Germany (Industry-Integrated): Led by independent VCs (35%), Germany also has a significant corporate presence, with CVCs (15%) and corporate direct investment (8%) combining for a 23% share. This points to strong collaboration between large industrial corporations and startups [8].

Japan (Corporate-Led): Japan is a clear outlier, with corporate investment (CVC at 15% + direct at 22%) reaching a remarkable 37%. Furthermore, financial institution VCs account for another 20%, leaving independent VCs with just 24% of the share. Unlike in the West, corporations and financial institutions are the central players in Japan’s ecosystem [5].

Focusing on corporate investment (the sum of CVC and direct investment) highlights Japan’s unique structure.

At 37%, Japan’s share of total corporate investment is significantly higher than that of the next closest countries, Germany and the United States, both at 23%.
This suggests that Japanese corporations are actively using startup investments as a vehicle for open innovation—seeking strategic synergies and exploring new business domains.

However, with a few exceptions, it cannot be said that significant results have been achieved.

The differences in investor landscapes are rooted in who funds the VC funds themselves—the Limited Partners (LPs). Comparing LP composition reveals the fundamental capital flow structures of each ecosystem.

US: The ideal model of patient capital. Pension funds are the largest source of capital at 30%, followed by university endowments and foundations (15%). This long-term, return-seeking capital is the bedrock of the US VC market [9].

UK: Family offices and high-net-worth individuals are the dominant LPs, accounting for 41%. This may be driven by tax incentives such as the Enterprise Investment Scheme (EIS) and the Seed Enterprise Investment Scheme (SEIS) [6].

France & Germany: Government and public agencies are the largest LPs, accounting for 30% and 25%, respectively. Public funds, often channeled through bodies like the European Investment Fund (EIF), are crucial for sustaining the continental European VC market [10].

Japan: Banks and insurance companies (31.7%) and corporations (25.4%) together account for over half of all LP capital. These institutions fill the role that pension funds play in the US. This structure may contribute to Japanese VCs' tendency to seek shorter-term returns or strategic synergies, potentially explaining the smaller median deal sizes [11].

Finally, a look at the historical trend of investment in Japan by investor category (based on SPEEDA data) confirms the consistent dominance of corporate-related investors. While there was a spike in foreign investment (tourist investors) in 2021, corporate investors (incl. CVCs) have remained the largest single category. The share of financial institutions has also been growing, particularly in 2024.

To attract foreign investors, it is necessary to increase the number of Japanese startups that expand into the global market, rather than focusing solely on the domestic market.

That is exactly my theme for the rest of my life.

PS: What is the US<>Japan VC Playbook? It is published by 500 Global, in partnership with JETRO, which analyzes and describes the differences in the startup ecosystems between the US and Japan. Highly recommended.

When I was young, I wasn’t very interested in visiting Kyoto. However, as I’m getting older, I’m getting to understand why Kyoto attracts many people, not only foreigners, but also Japanese.

It was about 20 years ago, but I still remember seeing an advertisement for Remy Martin at a train station that had the tagline, “Something that marketing can’t create.”

That’s exactly true.

Kyoto is filled with historic Buddhist temples and shrines, and the city itself is like an art museum. As you get older, you will come to understand its historical significance.

For foreigners, it must feel similar to how Japanese people feel when they see historical buildings in Rome or Paris.

It was in Kyoto that I met for the first time in about 10 years one of my fellow entrepreneurs, Takahito Iguchi, who is well known to all Japanese entrepreneurs and venture capitalists. This is the Wikipedia page (in Japanese) with rich information about him.

I was honored to have invested in his startup through the Umeda Fund, a VC fund that I ran while at SunBridge.

He founded an Augmented Reality technology startup called Sekai Camera (World Camera), which made a spectacular debut at TechCrunch50 in 2008, just one year after the iPhone was released in Japan.

The product was a device that displayed information about stores and other businesses within its field of vision when you held up an iPhone containing a “digital sticky note” called an Air Tag (unrelated to Apple’s lost property tag).

It can be said that the Sekai Camera was one of the catalysts that led to the creation of many augmented reality apps, including Pokémon Go.

Unfortunately, Sekai Camera was not a commercial success, but a few years later, he founded a smart glasses startup called Telepathy in Silicon Valley, taking on Google Glass head-on.

As a result, neither Google Glass nor Telepathy One succeeded. Many of you who read my newsletter are familiar with such tech products, so I don’t think further information is necessary.

After selling his shares in Telepathy, Iguchi worked on promoting voice SNS, developing the Baby app in 2016 at DOKI DOKI, INC., which he founded in 2014, followed by its successor app Ball in 2017 and Dabel (chatting) in 2018.

He launched Dabel two years before Clubhouse was launched (in 2020).

DOKI DOKI, INC. was the startup that we invested in.

I won’t go into detail about what happens next, but he has moved from San Francisco to Kyoto, launched a startup, and is about to release a product that will take on the world.

I visited him in Kyoto for the first time in 10 years, and if I understand correctly, he seems to believe that the world can realize serendipity more efficiently.

It not only tells me the route to get from point A to point B, but also where I should go today. This “map of the future”, based on an understanding of my humanity, will apparently be unveiled at SXSW 2026 in Austin this coming March!

If you are around, don’t miss it!

I wish I could be the witness!

Exactly a year ago—technically on December 30, 2024—I went to see a movie in Tokyo with two lively girls.
No one could have imagined that such a thing would happen. Even now, I still can’t fully accept that reality.

22 Years

The meaning and value of life may not depend on its length, but still, 22 years is far too short.

Her Presence

I wasn’t particularly close to her.
Yet her lively personality and sparkling smile—something like the sunshine of life—made her loved not only by me, but by everyone around her.

Why I Write This in English

If I were to write this in Japanese, people close to me would immediately know who she was.
That is precisely why I hesitated and chose not to write this story in Japanese.

What I Learned

Ever since I attended her farewell ceremony a week ago, I have hesitated every single day about whether I should write this blog.
I wasn’t able to accept the reality or organize my feelings—not for her sake, but for my own.

We met in a university setting—she as a student, I as a professor—yet it was I who learned the most from her.

Thank you for everything you taught me, even without knowing it.

Why I Started GGC

As readers of my newsletter may know, in January of this year, for the first time in my life, I established a company outside Japan: GoGlobal Catalyst (GGC) Pte. Ltd., together with my co-founder, Dan Brassington.

In the postwar era, when people pursued material wealth, Japanese companies such as Sony, Honda, Panasonic, and Toyota dominated global manufacturing.
Since then, with only a few exceptions, such as SoftBank and Uniqlo, Japan has yet to produce a truly global company.

I fully understand that my contribution is small.
Still, I don’t believe I am powerless.

I thought that what I could do was to help compensate for Japan’s handicap as a mono-ethnic nation by creating a platform that connects Japanese founders with foreign co-founders. That idea became GGC.

Failure and Acceptance

I had been thinking about GGC since around 2015.
In other words, I had been nurturing the idea in my mind for nearly ten years, without taking action.

I didn’t believe anyone would invest in such an idea.
And even if they did, I thought the chances of success were extremely low.
More than anything, I didn’t want to cause trouble for investors.

But, as we all know, life is finite.

So I decided to give it a try with our own money.
And as some of you already know, it turned out much as I had feared.

If I had the ability, fame, and financial resources of the successful founder, for example, DeNA’s Ms. Namba, the outcome might have been different.
But I am always myself—and I cannot be anyone else.

What Remains

One good thing about going through GGC is that I am no longer afraid of failure.
I realized that people don’t care about others as much as we sometimes think they do.
Of course they don’t. Everyone is busy living their own lives.

Even if you are not responsible for anything, your life may suddenly end one day.
If there is something you want to do, you should have started yesterday—not today, but yesterday. (Yes, I mean it.)

This past year has been a time for me to explore many possibilities and to think deeply about what I want to devote the rest of my life to.

Fortunately, I feel I was able to see that more clearly.
I now have a better understanding of what I cannot do—and what only I can do.

Life Is Finite

I sincerely hope that the coming year will be a wonderful one for everyone.

Life is finite.

For those who work in the startup or VC world, it probably doesn’t need to be explained that the excess liquidity of the last decade has driven up valuations for later-stage startups and spawned plenty of unicorns.

However, with the Russian invasion of Ukraine and the US Federal Reserve’s interest rate hike, the money started to reverse.

For example, in an industry I’m familiar with, Bowery, a leading startup in the Vertical Farming industry, which raised over $700M and was valued at US$2.3B, shut down in November 2024, according to PitchBook.

And, Infarm, one of my portfolio companies and also one of the leading Vertical Farming startups, had become a Unicorn but unfortunately ended up in bankruptcy in September 2023.

According to a Crunchbase report, there may be around 40 unicorns that have gone bankrupt since 2022. This suggests that many bankruptcies go unreported because they involve private companies.

And, a study by CB Insights reports that over 80% of unicorns are in the red.

However, as the graph below shows, VC is a classic power law business, even in the US. Or perhaps because it’s the US.

So, what is happening in the Japanese market?

As with the title of this post, the amount invested in startups peaked in 2022, reaching almost 1 trillion Japanese Yen (JPY) = $6.7 billion (JPY150/USD). It has decreased to JPY700B, or $4.7B, expected in 2025. See the graph below from Japan Startup Finance for the 1st Half of the year.

And the number of startups that got funded dropped from its peak of 4,368 in 2022 to roughly 2,800 expected in 2025. It’s about 64% of 2022.

Also, looking at the number and the total amount of newly established VC funds, we see the same trend.

The number of newly established funds and the amount of funds sharply dropped. The peak in the number of newly established funds was in 2022, at 154, and the peak in the amount was in 2023, at about $6.5B.

However, when it comes to the Seed-Early stage-focused funds, both the number of newly established funds and the total amount peaked in 2022: 60 new funds and $2.4B. One year earlier than the total VC market.

The Tide has changed.

One characteristic of Japan’s startup ecosystem is that IPOs are possible at small market capitalizations.

Everything has both Positive and Negative effects.

As a positive effect of such market conditions, many entrepreneurs, including me, challenged themselves to start their startups.

Such momentum attracted a lot of VCs, CVCs, and direct investment from large corporations. Then, Japan’s startup ecosystem has been fostered.

However, as the above graph by Speeda shows, it generated many small CAP IPOs, and about 70% of them remain under 10B JPY ($67M) in 5 years since IPO.

The bifurcation is 2 years after the IPO, as the graph below shows. The conditions that exceed 10B JPY in 5 years are described in the chart below.

But some of the small CAP IPO startups have grown dramatically.

SHIFT is the typical case. They debuted in the market with 3.4B JPY ($23M) in 2014. However, their market CAP has grown to 467.7B JPY ($3.1B) as of June 2025. It is 138x! There are some more startups like them.

In light of the above, the Tokyo Stock Exchange (TSE) has decided that the market CAP required for a company to remain listed must reach 10B JPY ($67M) within five years of its IPO.

Going back a little, investment in startups, the number of new VC funds established, and LP investment in VCs are all on a downward trend, partly due to the influence of the US market mentioned above.

However, it is also attributed to the poor performance of VC funds in Japan.

Below are the essential points from the excellent report published by Prequin and JVCA (Japan Venture Capital Association).

1. The report says that the Net Multiple in 2012 was 3.99, which is excellent.

2. The Net Multiple has been under 1.0 since 2020. But it’s still young, so it may have a chance to go over 1.0.

3. However, when it comes to “PME (Public Market Equivalent)”, from 2016 to 2021, it’s under 0%. Furthermore, for 2020 and 2021, the figure is minus 15-16%.

4. If you invested in TOPIX in 2023, the total return was about 30%, whereas VC funds in Japan returned only 5%, even giving up the liquidity.

I believe JVCA conducted this survey and made the results public because they believe that, to further grow Japan’s VC market, they must first be open about the industry’s current state and then discuss with more people how to achieve even higher performance.

And I think one of the reasons the VC performance in Japan is so low is that most startups go public even though they are still immature and have a small market cap. And as the report revealed, 70% of the listed startups still have a market cap below 10B JPY ($67M), even 5 years after their IPO. VC has to sell its shares and cannot wait for 5 years.

Also, the report hasn’t mentioned that point. Still, I think the other BIG reason VC performance in Japan is low is that most Japanese startups operate solely in the domestic market, meaning they can reach only 120 million people. It implies that a BIG valuation cannot be expected. If they target English-speaking countries, there’s a market of 1.7 billion people. I don’t need to tell you, though LOL.

That’s why I’ve supported and invested in young Japanese entrepreneurs tackling foreign markets.

Stay tuned. More reports, my analysis, and even my new project will be published.

And this time in NYC, I had another unexpected encounter.

One of Peatix’s local team members lives in Brooklyn, and we had dinner together. Before that, I visited a young Japanese founder staying at a property owned and operated by a well-known Japanese entrepreneur, which served as a “Hacker’s House” for emerging founders from Japan.

I first met Nao Yukawa (I don’t know the kanji), founder of the healthcare startup Lifestack, through a mutual friend—another young entrepreneur—. Still, we had only spoken online until this moment.

Back when I was a high school and university student, my impression of the University of Tokyo students was that they were “nerds who study all day.”

But after I founded Interscope and began receiving interns from ETIC, that stereotype completely disappeared.

At Interscope, we had countless interns from Japan’s top universities—The University of Tokyo, Hitotsubashi, Tokyo Institute of Technology (now Tokyo Institute of Science), Keio, Waseda, and more. They weren’t just academically strong—they were stylish, competent, and incredibly productive.
Daisuke Sasaki, founder of freee, which is a very successful listed startup, was one of them.

A quarter of a century has passed since then, and today, especially among engineering and graduate students at the University of Tokyo, the most desired career path after graduation is founding a startup, not getting a job. This is a massive cultural shift, and one of the reasons why I still believe in Japan’s potential.

The challenge, of course, is Japan’s deeply rooted, risk-averse culture.

On my business trip from NYC to SF and Silicon Valley last week, I rode in a Waymo for the first time. It was extremely comfortable, and I felt no safety concerns whatsoever.

I didn’t get a chance to ride it, but I did see Zoox—Amazon’s autonomous robotaxi startup—cruising confidently through San Francisco's streets. With its fully symmetrical design, no steering wheel, and complete focus on accurate autonomous driving, Zoox is striking. I’d love to ride it on my next trip.

From left: Nao Yukawa, Yuki Miyazaki, and I. In Brooklyn on Nov. 18th (Tue), 2025

Returning to the story of the Hacker’s House in Brooklyn—while chatting with Yukawa-san in person for the first time, another young founder named Yuki Miyazaki (again, I’m not sure of the kanji) dropped by.

When I introduced myself, he responded with a familiar line:
“I believe we’ve met before.”

Later, when I checked his LinkedIn profile, I saw the names Antti Sonninen (TAKEOFF Tokyo main organizer) and Lyuki Kumagai (熊谷流気). Ryuki is a student at the Entrepreneurship Musashino Campus (EMC) at Musashino University, where I teach as a professor.

I knew from Lyuki himself that he had co-founded an AI startup, but I never imagined I’d meet his co-founder in New York, of all places.

Since Lyuki isn’t an engineer, I had wondered whether he could really contribute meaningfully to an AI startup. But Miyazaki-san told me, “He handles everything except development. It’s been a huge help.”
Hearing that as an educator made me genuinely proud.

The next day, I flew from NYC to San Francisco.

There’s a program based on the “Silicon Valley Tour” we originally started during my SunBridge days, which is now a required course for second-year students at EMC. Since these students are still sophomores, their prior knowledge of startups and innovation varies widely compared to founders or aspiring founders.

To improve their learning outcomes and the overall quality of the program, we’re planning to make it more experiential next year. This trip to the Bay Area was meant for preliminary research and discussions with people involved.

Riding Waymo and even taking BART to UC Berkeley after a long time were all part of that research.

On BART, three Black performers boarded, blasted music, and suddenly began breakdancing—hanging from the overhead rails and dancing impressively. When they finished, they walked around with a hat saying “Donation!” and several passengers tipped them. At the next stop, they got off as if nothing had happened.

If this were Japan, what would have happened? Likely someone would complain, call it a nuisance, or activate the emergency button to stop them.
There are pros and cons, of course, but this kind of freedom—and society’s tolerance for it—is part of what allows innovation and new ideas to flourish.

View from the rear seat of a Waymo vehicle.

Montgomery Station, San Francisco BART.

By the way, San Francisco also has a place called Tech House, a living-and-working space for young Japanese founders aiming to succeed in the world’s major league of startups—SF and Silicon Valley.

There, I had candid conversations with Kiyo Kobayashi, known locally as the “Godfather” of Japanese founders in Silicon Valley, and Shun Yamada, whom I introduced in the earlier post, along with several other young Japanese entrepreneurs working hard in the Bay Area.

With Tech House members. On the far right is Shun Yamada. On my left is Kiyo Kobayashi. Nov. 21st (Fri), 2025

What they all share is a remarkably positive mindset.

They believe wholeheartedly that the future is bright. They’re grateful for the environment they’re in, and they work with incredible dedication every day.

Japan has yet to produce a globally recognized breakout founder, but if these entrepreneurs keep stacking their efforts, I’m confident we will see Japanese startups succeed worldwide.

Through my conversations with them, I became convinced of that.

And one more thing I rediscovered on this trip:
I genuinely love young founders—especially those aiming for the world. And perhaps it’s a bit self-indulgent, but I feel happy knowing that they seem to appreciate and trust me, too.

This is precisely the essence of DreamVision’s mission.

Keith Teare — Founding Shareholder of TechCrunch, Founder/CEO of SignalRank

Finally, I want to share a message from Keith Teare, someone I deeply respect. He offered me this advice, which I now take as a personal reminder:

“Ikuo. You should get rid of your ‘Silicon Valley Obsession.’”

Meaning: Silicon Valley is not synonymous with “global.”

That said, being validated by the US market—particularly Silicon Valley—remains a tremendous advantage for anyone aiming to become a global standard.

My unexpected reunion with Satoru Sasozaki, one of Snap’s B2C only Accelerators in NYC, after 11 years, my conversations with young Japanese founders in SF’s Tech House, and Keith’s advice—all of these experiences reaffirmed what I should dedicate the next chapter of my life to. And gaining that clarity and confidence was the greatest takeaway from this NYC → SF & Silicon Valley trip.

My first visit to New York was when I was a senior in university.
Just before graduation, I spent a little less than a month staying at the home of a New Yorker I had met in Tokyo.

Ever since then, I’ve been captivated by the unique charm of NYC—something beyond words—and I’ve returned roughly thirty times. Most recently, I brought five students from the Department of Entrepreneurship at Musashino University last March. But in recent years, my schedules have always been so packed that I couldn’t find the time to stroll through Central Park, which has long been one of my favorite places in the city.

This time, my flight ETA at JFK was 9:50 a.m., and I expected to reach my hotel around noon. So I decided, for the first time in more than a decade, to take a proper walk through Central Park.

The purpose of this trip was to connect with the “Accelerator” to foster and nurture Emerging Managers operated by the VC firm Coolwater Capital, which I had been participating in Tokyo.

An “Emerging Manager” refers to someone launching their first venture fund.

Coolwater’s program aims to nurture Japanese fund managers capable of raising capital from overseas LPs—an effort strongly supported by JETRO, which wants to accelerate the globalization of Japan’s startup ecosystem. (JETRO’s press release has more details.)

The requirements for applicants were surprisingly strict:

• Must be Japanese

• Must have an entrepreneurial background

• Must have successfully exited (IPO and/or M&A)

• Must be preparing to launch a new fund

• Must be able to conduct business in English

To be honest, finding people in Japan who meet all five criteria is no easy task. So one day, I received a cold email directly from Coolwater.

Since I wasn’t familiar with the firm, I reached out to Keith Teare—co-founder (Founding Shareholder) of TechCrunch and someone I’ve known for well over a decade—to ask if he knew their founder, Winter Mead. He replied that he did, and he introduced me right away.

After a 30-minute online conversation with Winter, he encouraged me to join the program. I applied immediately.

The first three sessions were held in person at JETRO’s Tokyo office. After that, for about two months, we met on Zoom every Wednesday morning, Japan time.

To keep this from becoming too long, I’ll skip the details—but the program was extremely valuable. I’ve run accelerators many times, but this was my first experience as a “participant.”

Each week, overseas LPs and VC professionals joined as guest speakers, offering insights into global VC trends and LP expectations. Several sessions included opportunities for us participants to pitch our own fund concepts. Wanting to make the most of it, I always volunteered to pitch first whenever a slot opened.

It’s not easy to objectively view your own strengths. So after incorporating their feedback, I revise my pitch deck every time before presenting again.

The Coolwater Investor Summit in NYC

Coolwater hosts a seasonal “Coolwater Investor Summit” in New York. Participants in the program could join, although travel expenses were self-funded.

I have a few friends in NYC, and I had a feeling something good might happen on this trip. So I decided to go.

And indeed, something very special happened.

From 2012 until June 2017, I ran an accelerator called SunBridge Global Ventures.
As part of that work, starting in 2013, we operated the “Silicon Valley Tour” for young people under the Osaka City–backed “Osaka Global Innovation (Osaka GI)” initiative.

A few months ago, I met a young entrepreneur named Shunsuke Yamada, and when I mentioned that I’d be in NYC, he said he wanted to introduce me to someone: Satoru Sasozaki.

When I sent Satoru a brief “Nice to meet you” message on Messenger, he replied:

“Actually, we’ve met before!”

To my surprise, he told me he had been one of the participants on our Osaka GI Silicon Valley Tour. He sent a photo—and there we were: a much younger him as a student, and a much younger me. I forwarded the picture to Masaaki Yoshikawa right away, who was the director at Osaka GI.

Satoru then told me something even more remarkable.

He said he decided to start his company—and do it in the United States—because of the English speech I gave during the kickoff session at Stanford University.

I don’t remember exactly what I said, but I do remember giving that speech.

He said it had such an impact on him that he felt, “I can’t build my future in Japan. I have to start my company in America.”
And after graduating from university, he moved to Silicon Valley without ever taking a job and began building his startup.

He lived in San Francisco at first, but because his business was B2C, he eventually moved to NYC, a city with far more diverse consumers, not only tech people.

He also managed to get accepted into Snap Inc.’s extremely selective B2C accelerator, which selects only 10 companies each year. (There’s a detailed story in Diamond Online. The article is in Japanese, just in case.)

A long-awaited return to Central Park

Satoru insisted on welcoming me properly and spending the whole day with me on the Sunday I arrived.
We met at 3 p.m. in front of The Plaza, and together we took my first proper walk through Central Park in more than 10 years.

From there, we crossed diagonally to the Upper West Side, then back toward the Upper East, where I wanted to see 425 Park Avenue, an incredibly striking building developed by Tokyu Land Corporation. I heard that my friend, Fumiaki Sato, was involved and had always wanted to see it in person.

After that, we headed downtown—this time by subway—and visited Hudson Yards, the extraordinary development by Mitsui Fudosan. One of the buildings, the architecture is impossible to ignore.

That evening, the two of us were joined by Mio Asatani, another introduction from Yamada-san. She spent her early teen years in the UK, graduated from the University of Tokyo, worked at Deloitte, earned her MBA from Columbia, and is now building a startup in the art space.

It was a wonderful evening.

A reunion that made the trip worthwhile

Reuniting with Satoru was truly heartwarming.
If the program we operated under Osaka GI had sparked something that led him onto this path—if it had played even a small role in shaping his journey—then I couldn’t be more honored.

This entry has become quite long, so I’ll end here for now.

But there was another moment on this trip that made coming to NYC absolutely worthwhile.

I’ll write about that in the next post.

That's because I thought it probably wouldn't work out. In fact, when I tried it, it didn't work out so well. In a sense, my hypothesis was confirmed.

In that sense, I'm glad I tried it. I feel so much better now lol.

Some of you who have subscribed to my Substack may remember that the “Catalyst” of mine was Dan Brassignton, who became my co-founder at GoGlobal Catalyst.

If we had not met, I would not have brought my 10-year idea to life.

July 12th, Friday, 2024, Dan and I met in Daikanyama as usual and talked about my idea. To be precise, it was a meeting to announce that I had decided to put into action an idea that I had been nurturing for the past 10 years.

Since then, the GoGlobal Catalyst project has started. Long story short, we launched our first Co-founder Matching program in July 2025 and conducted our first Demo Day on July 22nd.

Although it was our first time, we had no prior truck record; we received 30 applicants and accepted 11 founders.

However, as I anticipated, the Japanese founders who applied were only 3, and we accepted only one Japanese founder.

It means, although our program was “co-founder matching” to connect Japanese and foreigners, it didn’t function as the name suggests.

We have so many talented entrepreneurs in Japan. However, when it comes to the English-speaking entrepreneurs, the population suddenly becomes smaller.

That's why I procrastinated trying my idea; Dan was the reason I finally stopped procrastinating. He convinced me to give it a shot.

A few weeks after the GGC's first program, we met at a cafe in central Tokyo.

Dan has run his own AI-centric consulting firm called INCITE Advisory Group in Singapore, Bangkok, Sydney, London, and Tokyo. Thanks to AI, his firm has become very successful, and he has become busier and busier.

He said he wanted to resign from the board and transfer his shares to someone else who would benefit GGC. He doesn't speak Japanese and is too busy helping me keep GGC going.

I accepted his offer.

About a week later, I left Tokyo for Silicon Valley to run a program for our students at the Entrepreneurship Department of Musashino University (Musashino EMC = Entrepreneurship Musashino Campus), where I’ve been appointed as a professor.

He suggested that I consider, while I am in Silicon Valley, which direction I should guide GGC.

The EMC program features lectures by my entrepreneur friends and the founders of my portfolio companies, presented to our students. Their stories were very thought-provoking and insightful, not only for the students but also for me personally.

The final day of the program every year is at Plug and Play, where students present their business ideas in English. And after their presentation, I'll be sending them a message.

Every year, I explained to them Japan's financial situation and how far behind the rest of the world it was, to make them feel a sense of crisis.

This year, I shared a story about the decisions I made during my week in Silicon Valley, specifically letting GGC go, shutting it down.

My week in Silicon Valley gave me a lot to think about. In particular, my conversation with Keith Teare forced me to make two decisions.

One is grounded in my professional awareness, the other in my personal passion. Yet together they reveal the source of my identity—the origin of who I am.

Back in my junior high school days.

The teacher introduced the student exchange program with a state in the United States. And I raised my hand!

However, he said that you should wait until high school or university. If so, why would the teacher introduce the program?

Ever since then, I have always wanted to live abroad. And I want to be a Cosmopolitan Person.

I guess many of you who have subscribed to my newsletter may remember that my primary professional concern is why there isn't a next Sony or Honda from Japan yet.

Keith once said to me, “Japan’s true potential is consistently underestimated compared to its real strength. That’s why you believe there should be another Sony or Honda.”

That’s true. And the Co-founder matching is one of the means, not the purpose.

I know many young Japanese entrepreneurs who started their startups in Silicon Valley and continue to chart bold new paths there.

These people come to me asking for angel investment, but unfortunately, I don't have enough money right now to invest in them.

As an entrepreneur and angel investor for many years, I have invested in Silicon Valley startups and managed a small fund, placing me in the company of many well-funded entrepreneurs and fellow investors.

So I'm introducing them to these entrepreneurs.

But soon, I want to go beyond that—investing directly in them and helping their businesses grow in Silicon Valley. Life is short.

The next question is: Will you create the next Sony or Honda yourself? Or will you achieve it by supporting someone else?

The first guest speaker at this year's EMC Silicon Valley program was Robert Sharp, a pro-Japan American who runs the US subsidiary of a Japanese startup called HENNGE.

HENNGE, a listed startup at the Growth sector of the Tokyo Stock Exchange, used to be called Horizen Digital Enterprise. It was established back in 1996 when the founder, Kazuhiro Ogura, was a student at Hitotsubashi University in Tokyo.

When HENNGE was still at the seed stage, they received an investment from SunBridge, and their office was at SunBridge Venture Habitat, an incubation office in Shibuya, Tokyo.

SunBridge is a Venture Studio kind of company, co-founded by several Oracle OBs, both in the US and Japan, back in December 1999, when it was just the Internet Bubble in Japan.

I joined SunBridge in March 2011 and subsequently founded SunBridge Global Ventures, a spin-off that served as a Seed Accelerator based in Tokyo, Osaka, and Silicon Valley. I was privileged to get in touch with the Silicon Valley ecosystem through SunBridge Global Ventures.

Robert grew up in Japan but was drawn to work in Silicon Valley upon graduating from university, just as the Internet was taking off. Since then, he has spent the past 30 years working in Silicon Valley.

Now that his children were grown and had left home, he had more time on his hands and decided to visit Japan for the first time in a long while.

Feb 29th, 2024, he reached out to me via LinkedIn. At that time, he was searching for a startup ecosystem in Japan and was meeting many founders at seed-early stage startups and venture capitalists every single day.

I had forgotten that we exchanged messages since he reached out to me just before I left for Boston on a business trip for EMC.

Long story short. One of our mutual friends mentioned him, and I checked my LinkedIn and remembered that we had discussed it.

Then, he delivered a talk for my students on the first day of the program.

I’m sure that all the people who have subscribed to my Sabstack remember Steve Jobs’ speech at the graduation ceremony at Stanford University, where he made the famous expression, “Connecting the dots.”

When I “connect the dots” of my journey, each experience has shaped me in its own way—Interscope, where I achieved an M&A exit, and WebCrew, which went public, marked my beginnings as a founder. DreamVision, which didn’t work out, taught me the humility of failure. At SunBridge Global Ventures, I learned how to build and run an accelerator and investment platform. Later, I invested in Infarm, which grew into a unicorn (but ended up in bankruptcy), and brought them to Japan, where I founded Infarm Japan and served as its Managing Director. Each of these chapters, whether success or setback, has been a defining part of who I am today.

During my week in Silicon Valley, I reached a turning point: I decided to let GGC go and establish DreamVision for Japan as part of the DreamVision group to help “Japanese founders & startups” who challenge the US market in Silicon Valley.

In the next episode, I will explain what kind of business I've designed to launch.

So stay tuned!