In the first nine months of this year, startups in Southeast Asia raised more than $13 billion. A record. Depending on whose tracker you believe, that is somewhere between a 120% and a 170% jump on last year.
Hold on to that range for a second. Two serious data providers, looking at the same region over the same nine months, cannot agree on the number. That is not an accident, and we will come back to it.
Here is the second number, and it is the one I want you to actually feel. Of that $13 billion, seed and angel rounds, the first real cheque a brand new company gets, came to about $270 million. Spread across roughly 90 companies.
Sit with the gap. $13 billion at the top of the page. Ninety founders at the bottom, splitting $270 million between them. The money in this region has never been bigger, and it has almost nothing to do with the founder in Manila, KL or Bangkok trying to make payroll this month.
I have watched the money here change shape about four times in a decade. Every time, it arrived looking like the boom had finally come. Every time, the headlines said we had arrived. And every time, when you followed the money to the end, it was not here to build a Southeast Asian company at all. It was here for something else.
So this week I want to follow the money through all four of its disguises. Then, at the end, I will tell you why, after taking a blowtorch to all of it, I am more bullish on this region than I have been in years. I mean that. This is not a sad story.
One. The headline is the trick
Thirteen and a half billion dollars, nine months, a record. If that is the only line you read, the drought is over and it is time to go out and raise again.
Here is what the line does not tell you. Three deals were 54% of that number. Three. Widen the lens to every round of a billion dollars or more, and four deals were about 63% of all the money raised in the entire region this year. Before we have even started, nearly two thirds of Southeast Asia’s record year is a handful of enormous cheques going to a handful of enormous companies.
So here is the one habit that will protect you from almost every misleading funding headline about this region. When a number says funding hit a record, your first question is never how much. It is how many. How many companies? Here, more than almost anywhere, one or two giant transactions can carry the whole number and make a starving market look like it just ate a feast.
Look at where the money actually sat. Late stage, the Series C and beyond, the pre-IPO money, more than tripled to over $10 billion. Early stage, the seed and Series A, the part of the market where companies are actually born and where the next decade here gets decided, barely moved.
Data centres alone were about 35% of all the money. I would love to know when we started counting data centres as startup funding. Video generation tools took another big slice on top.
Think about what a data centre is. Power, concrete, cooling, and rack after rack of servers. It is a real estate and energy play with a technology label on the front. Perfectly good business, and we need them. It is just not a Southeast Asian founder with an idea and a laptop. The single biggest category in our record year is, quite literally, buildings.
So the record is not founders getting funded. The record is a small number of very large, very late cheques going into physical infrastructure and a couple of foreign technology giants, sitting on top of an early stage market running on fumes. That is not a recovery. That is a costume. A big, shiny, $13 billion costume.
To understand the costume, you have to understand the waves that came before it, because each one taught us a bad habit. We are still paying for all of them at once.
Two. The hype money
Most of you lived through this one. Rates were on the floor, capital was effectively free, and it flooded into Southeast Asia from big global funds that had made a fortune in China and were hunting for the next one.
The instruction attached to that money was simple. Grow at all costs. Buy the users, buy the market share, buy the headline. Worry about a business model later, or frankly never.
For a while it worked the way a sugar high works. Valuations went vertical. Everyone was a soonicorn. The pitch stopped being “here is a company that makes money” and became “here is a rocket, get on before it leaves the ground.” Burn was no longer a warning sign. Burn was the strategy. You were supposed to lose money faster than the next guy, because that proved you were winning the land grab.
I sat in rooms where a founder losing $10 million a year was treated as more investable than one quietly making $1 million, because the one making money was clearly not thinking big enough.
Then the free money ended. Rates went up, the global funds pulled back, and reality walked in through the front door and turned the lights on.
You know the wreckage, because some of it made headlines around the world. eFishery in Indonesia, the poster child for aquaculture meets technology, hundreds of millions raised, a genuine darling until it was not. According to reports, a very large share of its revenue turned out not to be real, inflated to feed a valuation that had become the whole point. Zilingo raised enormous money, scaled hard across the region, and came apart on governance and burn before it could grow into the price tag it had been given.
Underneath those two famous names sits a long, quiet graveyard. Companies that raised a monster round at a monster number, then went silent. Shut two or three country offices. Laid off most of the team. Then did a down round so brutal, or a bridge on such ugly terms, that they simply stopped sending press releases. You only find out when you check the cap table a year later.
But here is what I think wave one really did, and it is worse than the billions it burned. It rewired a generation of founders. It taught them that the core skill of building a company was fundraising, not building. That the most important customer was the investor, not the human being using the product. It trained people to run their companies like they were preparing to sell them to a greater fool, because for about five years the greater fool kept showing up on schedule with a bigger cheque.
When the greater fool stopped showing up, a lot of those founders did not know what to do next, because nobody had ever asked them to. I have watched genuinely talented people spend three years becoming world class at exactly one thing, raising money. When the market turned, that skill was worth nothing, and there was nothing underneath it.
That is the quiet tragedy of wave one. Not the companies that died loudly. The founders who were patiently taught the wrong trade, and the LPs, angels and fund managers who walked away with a damaged view of the whole region. Disguise number one: money that came to inflate a story into a bigger story and flip it before the music stopped.
Three. The grant money
Wave two is gentler, and in some ways it did more lasting damage, precisely because it showed up wearing the nicest costume of all. This money came to help.
When the hype money left a crater, two kinds of capital stepped into the gap. Both arrived with genuinely good intentions and genuinely terrible incentives. Hold both ideas at once.
The first is grant money. To be clear, this is everywhere, not one country: the US, Europe, Seoul, Japan, right across Asia. Governments decided they could build a startup scene by writing cheques directly to startups. Grants, matching funds, innovation programmes, disbursements tied to milestones. On paper it sounds wonderful. Free, non-dilutive capital for founders. What could go wrong?
Here is what goes wrong. That money does not select for the best company. It selects for the best grant application.
Think about who approves these grants. Usually a panel. Often very few of the people on it have ever built or sold a piece of technology. They sit in a room with a scoring rubric, and they are not really assessing whether the thing is viable, whether the commercial plan makes sense, or whether a single customer on earth will pay for it. They are ticking boxes. Does it hit the criteria? Does it use the right words from this year’s policy priority? Does it score high enough on the technology readiness scale?
So founders learn to build for the rubric. Privately, most people in this industry have a name for them: the grantrepreneur. Brilliant at the form. Fluent in whatever this quarter’s programme wants to hear. Deep tech when the deep tech window is open. Green when the green money is flowing. AI on everything the moment the AI fund opens, whether or not there is any AI in the building. They are not building a company. They are building a company shaped like an application, optimised for a committee that has never once had to find a paying customer itself.
It gets worse, because the money usually arrives in tranches tied to milestones. Now the founder is not building toward a customer. They are building toward the next disbursement. The company starts running on the rhythm of the grant calendar instead of the rhythm of the market. I have watched founders quietly delay shipping a real product because the paperwork for the next tranche was due that month, and the paperwork paid better than the product did.
Four. The corporate money
The second kind of help was the corporate venture arms. On paper, strategic investment. The big incumbent takes a stake in the promising young company and everyone talks about synergy, partnership and the future.
In practice, a lot of it was much narrower. A big, slow company looking for a captive supplier, or buying a cheap option on a technology it was too slow or too internally tangled to build itself. Come in, take a stake, take a board seat, and quietly insert a right of first refusal.
That last part is the killer. Once a corporate holds a right of first refusal and a board seat, the startup can no longer freely sell to that corporate’s competitors. In most markets here, that is most of the market.
They called it validation. The founder got a logo for the website and a nice press release. What they often actually got was a slow, polite smothering. They lost the one thing an early company needs more than money: the freedom to sell to anyone who will take their call. You cannot grow when half your potential customers see their competitor sitting on your board.
So that is wave two. Money that came to help, and taught founders to perform for a scoring panel and to hug a corporate parent instead of chasing a customer. Best intentions, exactly the wrong incentives. Of all the disguises, it is the one that looks most like kindness.
Five. The passport money
Which brings us to now. Wave three, the money making this year look like a record, and the one almost nobody is willing to name out loud. So let me.
Go back to the deals that made up more than half of Southeast Asian funding this year. Do not read the label on the box. Open it.
· DayOne, the data centre business. Billions of dollars. Spun out of GDS, a Chinese company, now sitting in Singapore and lining up to list.
· Kling, the video generation tool. Dutifully counted by the trackers as Southeast Asian funding. It is a Kuaishou product, and Kuaishou is Chinese.
· Ant International. A $1.2 billion round this year at a $10 billion valuation. It is the overseas arm of Ant Group, and when you look at who actually put money into the round, it was Ant Group and Alibaba. A Chinese parent wrote a billion dollar cheque to its own Singapore subsidiary, and we booked it as Southeast Asian venture funding and put it in the record.
· Sharpa, the robotics company. Founded by co-founders of Hesai, a Chinese LiDAR company. A Singapore address, a Shanghai lab, and a cap table with Alibaba, Tencent, Meituan and JD.
To be completely clear, because this is the easiest thing in the world to get wrong: I have no problem with any of these companies, and I think more of them should look at Singapore. If Singapore is where the best firms in the world want to set up, that is a compliment to Singapore, and it is good for the whole region that they choose to be here rather than somewhere else.
That is not my point. This is not about geopolitics, this side or that side. My point is about what we are counting, and what we quietly tell ourselves when we count it.
There is a name for this now: Singapore washing. A company that is Chinese in origin, capital and leadership puts a Singapore address on the door so it can reach global investors who are nervous about country of origin, and list somewhere that is not caught in the crossfire between Washington and Beijing. It is a real, well documented strategy. Not a conspiracy, not a secret. And it is already starting to come undone, as Meta’s acquisition of Manus showed. Both governments have figured out the trick, and both are tightening their grip.
So here is the uncomfortable truth about the record. When three or four companies that are Chinese in everything but their postcode make up more than half of Southeast Asian funding, the league table is not measuring Southeast Asia building anything. It is measuring spin-outs and re-domiciling. The money showed up wearing our postcode. The region is the venue, not the destination.
I want to be fair about the data centre money in particular, because it will keep getting counted. Those builds are real. They bring power, some jobs, and capacity this region genuinely needs. But a global hyperscaler pouring billions into concrete and cooling is an infrastructure decision. Counting it as Southeast Asian venture funding makes no sense to me.
The people most fooled by that number are not the journalists who print it. They are the founders and fund managers on the ground who read “record year for Southeast Asia” and feel like the party has restarted for them. It has not. Strip off all the disguises, the hype, the help and the passport, and what is left for a real Southeast Asian founder is the sliver from the top of this piece. $270 million. That is the real market. And unlike the first two disguises, this one is happening right now, in the number you will read this week.
Six. Why I am still bullish
That is the rant. Now the other half, the part I believe most of all.
The constraint in this region was never talent. I sit across the table from founders every week, and the quality of the people building right now, after the hype washed out and the tide went all the way down, is the best I have seen in a decade of doing this. They are not chasing a soonicorn headline, because they watched that game die with their own eyes. They are building with less money, more discipline, and actual paying customers, because that is the only option the market left them. That is not a weakness. That is exactly the founder you want to back with real money.
The opportunity was never the problem either. The problems worth solving here are bigger than ever. Hundreds of millions of people coming online, formalising, moving money for the first time, buying, selling, borrowing and building, in markets the global giants still do not understand from the inside. The opportunity is not shrinking. It is growing while everyone is distracted by the headline number.
The only thing that has ever been missing is the right kind of money. Look at the four waves again. Hype money that came to flip a story. Grant money that came to feed a committee. Corporate money that came to capture a supplier. Passport money that came to borrow an address. Four waves, and not one came here to build a Southeast Asian company and hold it long enough to matter.
That is the wave we are waiting for. Patient, honest capital. Money that prices risk instead of chasing a headline. Money that backs a founder to go and find a real customer, and is perfectly happy to be boring and early and local while it does.
Seven. What the right money asks
Here is the good news, and the reason I do this at all. That money is starting to show up. Some of it, honestly, is me and the people I work alongside, writing seed cheques into exactly this moment on purpose. The valuations are finally sane, the founders are finally serious, and almost nobody else is bothering to look.
What that kind of money does differently is not complicated.
· It asks who is paying you. Not who else is on your cap table.
· It is comfortable with a company growing at a sane speed instead of a reckless one.
· It does not need a logo for a press release. It needs a business still standing in ten years.
It is a little boring. After the decade this region just lived through, boring is the most exciting thing I can imagine.
The bow
The talent is already here. The opportunity is already here. The missing piece was the money, and it is starting to turn up. When the right capital finally meets the founders who survived all of this, this region is going to surprise a lot of people who only ever read the top line of the report.
So the next time someone forwards you a breathless headline about a record funding year, do two things. Ask how many companies, not how much money. And ask where the backer actually flew in from.
If you are a founder sitting in that seed sliver right now, building the unglamorous thing at a sane valuation with one real customer who actually pays you, and you feel like you somehow missed the party: you did not miss it. You are early for the one that is actually going to count.
If this changed how you read the funding numbers, forward it to one founder who needs to hear the last section. Then reply and tell me where I am wrong. The pushback is usually where the next episode comes from.
This piece accompanies this week’s episode of SEA of Startups. Real. Raw. Relatable. Listen on Spotify, Apple Podcasts, or YouTube, and subscribe here so the next one lands in your inbox.
