Qanot / Thesis
The constraint is not talent. It is the first cheque.
Central Asia has 80 million people, a median age under 30, tens of thousands of engineering graduates a year, and two companies that have already crossed a billion dollars. What it does not have is anybody whose job is to fund a prototype. This is the argument for why that gap is an opportunity rather than a warning — including the part where we set out what would prove us wrong.
Stated plainly, so it can be argued with.
Formation-stage capital, not founder quality, is the binding constraint on company creation in Central Asia — and it is the cheapest constraint to relieve.
Every emerging market pitch sounds the same: large young population, rising smartphone penetration, underserved consumers. Those things are true here and they are also true of twenty other places. They are not a thesis; they are a backdrop.
The specific claim is narrower and testable. In Central Asia, the number of companies that reach a Series A is limited almost entirely by how many survive the eighteen months between prototype and first revenue — and that survival is determined by whether anyone will write $150,000 into them. There is no domestic institution whose mandate is to do that.
If the claim is right, adding a small amount of capital at exactly that point produces a disproportionate increase in the number of companies reaching later stages. If it is wrong — if the real constraint is founder capability, market size, or exit paths — then adding capital produces a larger number of well-funded failures and nothing else.
We think it is right, and we have set out in SEC.06 exactly what we would accept as evidence that it is not.
A young market that is getting richer, not older.
Uzbekistan's median age is roughly 28. For comparison, India is about 29, China about 40, Germany about 46. The region is adding working-age population at a time when almost every developed market is losing it.
What makes this commercially interesting rather than merely demographic is that the same cohort is the first to be fully banked and fully online. Uzbekistan went from a predominantly cash economy to majority digital payments inside a decade. That is not a gradual adoption curve; it is a category being created from zero, with no incumbent behaviour to dislodge.
The comparison people reach for is China 2005 or India 2013. The more useful comparison is that in both of those cases, the companies that won were domestic, founded by people from the market, solving a problem the market already had — not localised versions of Western products.
Engineers are not the scarce input.
Uzbekistan produces tens of thousands of technical graduates a year, and a substantial outsourcing sector has grown up to absorb them — which is, in a sense, the problem. An outsourcing economy converts engineering capacity into billable hours at a fixed margin. It does not convert it into equity.
Every hour a strong engineer in Tashkent spends building someone else's product at $25 an hour is an hour not spent building something they own. The outsourcing sector is not a competitor to the startup ecosystem; it is the default outcome when there is no alternative, and it is what the first cheque is competing against.
Approximate figures for illustration. Source and date-stamp each before launch.
Venture capital per head, compared.
The chart is the thesis in one image. Israel and the United States deploy hundreds of dollars of venture capital per person per year. India, a decade into building domestic venture infrastructure, deploys single digits. Uzbekistan deploys a rounding error.
Two readings are possible. The pessimistic one: capital goes where returns are, and its absence is the market's verdict. The optimistic one: capital is slow, geographically sticky, and follows proof — and the proof arrives before the capital does, not after.
Kaspi and Uzum are the proof arriving first. Both were built without a domestic venture ecosystem to speak of. Both are now worth more than the entire annual venture deployment of the region by a wide margin. The capital has not yet responded.
Why now and not five years ago.
- Currency convertibility
- Uzbekistan floated the som in 2017. Before that, a foreign investor could not reliably get money out, which made institutional venture investment structurally impossible rather than merely unattractive.
- Digital payments
- Card and wallet penetration moved from marginal to majority inside a decade, creating the rails that every fintech and marketplace company depends on.
- Legal structuring
- ADGM and DIFC in the Gulf now give regional companies a credible, English-law holding structure that a US or European fund can invest into without a bespoke legal opinion.
- Proof of exit
- Kaspi's Nasdaq listing established that a Central Asian company can be priced by global public markets. Before it, the exit path was theoretical.
- Cost of building
- A functioning product that required a team of eight in 2018 requires two or three now. The minimum capital to reach evidence has fallen by roughly an order of magnitude, which is what makes a $150K first cheque a serious instrument rather than a gesture.
India in 2013 is the closest analogue.
In 2013, India had a handful of billion-dollar technology companies, a thin domestic seed ecosystem, and a widespread view among global investors that the market was too poor and too fragmented to support venture returns. A decade of formation-stage capital later, it has roughly eighty-five.
We are not claiming Central Asia becomes India — the population is a twentieth of the size and the domestic market cannot support the same absolute outcomes. The point is the mechanism, not the magnitude: the constraint that looked structural turned out to be a capital-formation problem, and it resolved in under ten years once someone funded the bottom of the pipeline.
The honest version of our ambition is proportionate. Eighty million people with a functioning first-cheque ecosystem should produce, over a decade, something in the range of five to fifteen companies worth more than a billion dollars. That is not a spectacular claim. It is enough to make a $25M fund work several times over.
What would prove this thesis wrong.
Stated in advanceA thesis that cannot be wrong is a marketing position. These are the four observations that would tell us the constraint is something other than capital — written down now, before the data arrives, so we cannot reinterpret them later.
- Funded companies stall anyway
- If companies that receive a first cheque fail to reach Series A at a materially higher rate than those that do not, the constraint was never capital. Measured against a control group of applicants we declined at the margin.
- The best founders emigrate regardless
- If the strongest teams relocate to Dubai, Berlin or the US within eighteen months of funding, we are subsidising an export pipeline rather than building a domestic ecosystem. Tracked explicitly.
- No Series A leads appear
- Our model depends on somebody pricing the next round. If regional and Gulf funds do not follow into companies with genuine metrics, the pipeline terminates at our stage and the whole structure fails.
- Currency or policy reversal
- Convertibility is recent and politically reversible. A return to capital controls would make the entire asset class uninvestable irrespective of company quality — this is the risk we can neither hedge nor diversify away.
What follows from all of this.
- Write first, not second
- The gap is at formation. Being the third cheque into a company that already raised is not addressing the constraint we claim exists.
- Be physically present
- A studio floor in Tashkent, not a fund that visits twice a year. The information advantage that justifies the position is only available in person.
- Publish the rubric
- In a market with no established venture norms, an opaque process selects for people who already know an investor. A written rubric selects for evidence.
- Build the Series-A bridge
- Our own falsification list names this as the failure mode most within our control, so it gets partner time rather than a paragraph.
- Keep the fund small
- As the fund-math model shows, a large fund requires an outcome this market may not produce for a decade. A $25M fund works on outcomes it can plausibly produce now.