Six Billion Dollars And A Whole Philosophy: Southeast Asian Tech Closes Out 2025

Southeast Asian tech raised six billion dollars in 2025, according to industry researcher Duke — a thirty percent fall from the year before and the worst year for capital raising in the region since 2016. ^{[1] [2] [3]} More than a quarter of that capital came in the form of debt rather than equity, ^[4] a marked departure from the long-favored playbook of growth-at-all-costs. Investors, having spent the better part of a decade handing out cash to companies that promised the moon, have apparently decided to ask to be paid back. ^[5]

This was supposed to be the year of the rebound. Instead it was the year the rebound was “deferred to 2026,” which is industry shorthand for “we will figure this out after the holidays.” The composition of the money changed in a way that should not be surprising and, yet, in the regional tech press, has been treated as a mild surprise.

For a region that long sold itself on the promise of patient capital and the discipline of the long-term founder, the migration to debt feels a little like a five-star restaurant quietly switching to a fast-food menu. The food is still there. The vibe is different. Lenders, as a class, are less patient than equity tourists, not to mention better at tracking outstanding balances.

Indonesia, of course, was the regional exhibit A for the growth-at-all-costs approach — the place where the strategy was pursued most enthusiastically, and where, personally, I have been waiting for the bill to arrive. The fact that investors are, at long last, asking what things cost is, in fairness, the most adult development the regional tech sector has had in some time.

Why now? Because, as Duke notes, macroeconomic conditions have not exactly been friendly, with the U.S. Federal Reserve — the American central bank — keeping its interest rates above pre-pandemic levels. ^[6] The price of money is high, the appetite for risk is low, and the safe assumption is that any company whose pitch deck features a slide titled “TAM, allegedly” is going to have a rough quarter. The Fed has, however, indicated that it may cut rates in 2026, ^[7] which is the financial equivalent of a parent saying “we’ll see,” and is being received with appropriate caution.

This is, of course, also the part where I am meant to blame the Americans, because that is the global pastime, and the U.S. Federal Reserve makes a very convenient villain. The Fed did, after all, set the cost of capital for the entire world, including the parts of the world that have never sent it a Christmas card. So: thank you, Federal Reserve, for making everyone else’s spreadsheets harder to balance. You are, as always, very welcome to cut now.

In any case, the year was not a total loss. Southeast Asia minted five new unicorns in 2025, up from two in 2024. ^[8] That is “more than double,” and the doubling is real, in the same way that going from two to five is a real statement about doubling. It is, I would argue, the kind of math that sounds excellent in a press release and considerably less excellent when you actually do the math. Two to five is a hundred and fifty percent increase. It is also, in absolute terms, three more unicorns than last year. The press release does not usually include the second number.

The new cohort, per Duke, is tilted toward artificial intelligence and digitalization ^[9] — a sensible bet, given that AI is currently the only sector in which companies can lose money at speed and still be considered a good investment. The market has decided that the next generation of Southeast Asian tech will be built on models, agents, and the patient expectation that someone, somewhere, will figure out how to monetize all of it. That, too, is a tradition with a long history in this region.

So: less money, more debt, twice as many unicorns, a Federal Reserve that may or may not cut, and a region entering 2026 with the kind of cautious optimism usually reserved for people who have just checked their brokerage account and are choosing not to check it again until February.

The growth-at-all-costs era is over. The new era, judging by the data, will be something closer to growth-with-the-lights-on, funded largely by people who would like their money back at some point, and betting heavily on the small but real possibility that the next big thing is currently being prototyped in someone’s garage, in someone’s coworking space, or, increasingly, in someone’s large language model.

I, for the record, would like to remind the audience that ‘growth at all costs’ was always a strange slogan for an industry that, in practice, kept discovering what the costs were. But that’s just one man’s opinion.