The underground lottery: Thailand's other financial market

There is a form of illegal lottery here called huay hun, in which the winning numbers are the last two digits of the Stock Exchange of Thailand index at the close of trading.

A national stock exchange, the thing that prices our listed companies and signals economic confidence, also functions, twice a day, as a random number generator for illegal gambling.

Nobody designed this. It emerged because the SET close is public, frequent and impossible for a bookmaker to fix. As a randomising mechanism it’s genuinely excellent, which is the sort of sentence I did not expect to write about organised crime.

And it’s the least interesting thing about a system that is far larger than the legal one it shadows.

The scale

Thailand’s underground lottery is estimated at four to five times the size of the national lottery. The estimate is attributed to researchers at Rangsit University’s College of Social Innovation and is the figure most Thai and English coverage traces back to.

Two things a reader should know before leaning on it. The hard numbers usually quoted underneath it — around ฿108 billion on the underground lottery against ฿36 billion on the government one — date from the mid-2000s, which is twenty years ago. And nobody audits an illegal market: published total-value estimates run from hundreds of millions to several billion dollars depending on method. Treat the ratio as the durable finding and every absolute figure as indicative.

The ratio is the point. The Government Lottery Office’s product — twice monthly, fixed prices, capped prizes — is the smaller sibling. The larger business is the illegal one running alongside it, and something close to a third of the population is reported to play, which is a reported participation figure rather than a measured one.

It is illegal, entirely tolerated, and about as hidden as a food cart.

Why the illegal one wins

Because it’s a better product. I don’t enjoy writing that, but it’s plainly true.

Any stake. The state sells fixed-price tickets. The neighbourhood operation takes whatever you have — ฿20, ฿500, whatever’s in your pocket.

Any number. This is the one that matters. The state lottery sells numbers that exist in its printed inventory. If the number you dreamed about isn’t in stock, you cannot have it.

Think about how badly that misunderstands the product. In a system where the number is the entire point (where the whole activity is about your number, received in your dream, meaning something specifically to you) being told to pick from what’s in the box is a fundamental design failure. It’s like a bakery that sells whatever’s left.

Better payouts, because nobody’s funding state programmes out of the margin.

And it’s local. Someone you know, in cash, settled the same evening.

The state’s product is worse on price, choice, payout and convenience. It survives on being legal, which is a weaker competitive advantage than the Government Lottery Office seems to believe. The office has noticed. Since November 2024 it has sold a three-digit digital product, N3, that takes any number for ฿20 — a direct copy of the underground product’s best feature — and by August 2026 it was selling more than six million tickets a draw. Whether that has taken anything from the underground market is unpublished; the article on N3 has the figures and the part the state still can’t copy.

The omen economy

Here’s the part that genuinely doesn’t exist in English, and it’s the interesting one.

The number isn’t chosen. It’s interpreted.

The day before a draw, coffee shops fill with people describing dreams to each other and arguing about what they mean numerically. That isn’t a quaint detail — it’s a routine, widespread, semi-public activity, and if you’ve spent any time in a Thai family you’ve sat through it. A relative describes a dream about a snake. Someone else disagrees about the derivation. Somebody’s aunt has a competing convention she’s confident about. It goes on for a while.

Beyond dreams, numbers are read from almost anything anomalous: road accidents — particularly registration plates, which is why you’ll see people photographing a crashed car with more interest than sympathy — unusual animals, the ages of people involved in notable events, markings on tree bark, objects found at shrines, and numbers given by monks and fortune-tellers.

There is a publishing genre for this: หนังสือใบ้เลขเด็ด, lucky-number hint books, containing established conventions for converting a dream about a fire, or a snake, or a dead relative, into two or three digits. People disagree about the conventions the way other countries disagree about football tactics, and with similar heat.

And wherever there’s a shrine, there are lottery vendors. That’s not opportunism, it’s logistics. The shrine is where numbers are received, so it’s where the market sets up.

What this amounts to is a nationwide, continuously running system for converting random events into numerical predictions — with a literature, a vocabulary, professional interpreters and physical distribution. It simply isn’t recognised as a system, because we file it under superstition rather than under the substantial informal financial apparatus it actually is.

Why it’s less irrational than it looks

The comfortable foreign reading is that this is silly. I’d push back on that, and not defensively.

It is not a savings product, and the risk goes first because the shorthand is tempting and wrong. A savings account returns your capital. A lottery ticket has an expected value well below its price — that gap is how the operator is paid, and over any length of time the household loses money. Most people who play lose. That is the design working, not failing. An underground operator adds counterparty risk on top: there is no regulator, no deposit guarantee and no court.

With that said, here is why “irrational” is still the wrong criticism. Household debt here is 85.9% of GDP, and much of it is non-productive — personal loans and credit cards rather than mortgages. The often-quoted “over half” appears to describe the share of accounts rather than the share of value, so this article states the direction and not a proportion. For a household with no realistic route to accumulating wealth, a small regular outlay at a very small chance of a transforming return is a different kind of decision from the one a saver makes — not a better-returning one, but one aimed at an outcome that saving ฿20 a week cannot reach at all.

That is not an endorsement. Lotteries move money from the poor to operators, here as everywhere, and I’ve watched it do real damage to people I know. But “irrational” is the wrong criticism. It’s a rational response to a shortage of good options, which is a considerably more uncomfortable finding.

It’s also social, which foreigners consistently miss. The dream discussion, the shared numbers, the collective analysis of an accident — it’s participatory. The Government Lottery Office sells a ticket. The man on the corner sells membership of an ongoing conversation, and that’s most of what people are buying.

And it’s a genuine parallel economy, with credit, trust and enforcement operating outside the formal financial system — in a country where, as article 14 Opening a Thai bank account in 2026: what actually works covered, formal financial access still isn’t universal.

What Thai sources say and English ones don’t

Thai coverage of this is substantial and takes it seriously. The Department of Mental Health has published on lottery-playing behaviour. Actuarial trade publications have analysed the underground lottery’s mathematics against insurance principles — which is a wonderfully Thai thing to do, treating an illegal numbers game as a legitimate object of professional study. Thairath has investigated the business structure.

English coverage is mostly tourist explainers on how to buy a legal ticket.

The whole interpretive culture (the derivation conventions, the hint-book industry, the shrine economics, the coffee-shop analysis) is documented at length in Thai and essentially absent in English. Which is a strange gap for something a third of the country participates in, and it exists for the obvious reason: the people writing about Thailand in English mostly can’t read about Thailand in Thai.

What people get wrong

“It’s just gambling.” It’s gambling with an interpretive apparatus attached, and the apparatus is the culturally distinctive part.

“The legal lottery is the main one.” The underground market is estimated four to five times larger, on a Rangsit University figure whose underlying numbers date from the mid-2000s.

“It’s hidden.” It’s illegal and about as concealed as a noodle stall.

“Only poor people play.” Participation spans income levels. The harm falls disproportionately on lower incomes, which is a different claim and a true one.

“Thai people are unusually superstitious.” Every country has a lottery and every lottery has a folk numerology. Ours is unusually well-developed, systematised and publicly discussed, which makes it visible, not exceptional. I’d gently point out that people who buy their birthday every week are running a derivation system too. It’s just a worse one.

Final thoughts

The takeaway isn’t that Thailand has an illegal lottery. Most countries do.

It’s that we’ve built, without anyone designing it, a distributed national system for turning the random events of daily life into numbers — with reference literature, professional interpreters, distribution at shrines and a settlement layer running on trust. And that it outcompetes the state’s version four or five times over, on product quality.

Somewhere in there is a lesson about what happens when the official option is worse than the unofficial one, and about how much financial ingenuity gets classified as superstition purely because it’s practised by people without bank accounts.

Also, if you dream about a snake, there are conventions. Several. My aunt is certain about hers and I have learned not to argue.