Thai retail investors are taught a comforting story.
Buy and hold. DCA — dollar-cost-average — into Thai blue chips. Time in the market beats timing the market. Reinvest your dividends and let compounding do the work.
We tested all of it. Across 180 Thai stocks (we excluded 18 that underwent recent corporate actions — more on that below), with 15 years of monthly DCA, real prices, dividends with 10% withholding tax counted.
Here is what the data says.
The result
Of 180 Thai stocks tested under 15-year DCA — buying 20,000 baht every month, dividends compounded — only 73 produced a positive total return.
107 stocks lost money over fifteen years of patient monthly buying. The median Thai stock returned negative 6.4% including dividends. The mean was a tepid +6.3% — barely above keeping the money under a mattress, and dragged up entirely by a small group of compounders.
| Metric | Value |
|---|---|
| Universe analyzed | 180 stocks |
| Winners (positive total return) | 73 (40.6%) |
| Losers (negative total return) | 107 (59.4%) |
| Strong winners (≥50% return over 15yr) | 29 (16.1%) |
| Strong losers (lost ≥50% of capital) | 18 (10.0%) |
| Severe losers (lost ≥75% of capital) | 3 |
| Mean return | +6.3% |
| Median return | −6.4% |
The "buy and hold" story most Thais hear assumes the average stock compounds. The data says the average Thai stock barely treads water, and the median actually destroys capital over fifteen years of disciplined DCA.
The honest disclosure: data is harder than it looks
Earlier drafts of this essay quoted a 71% loser rate and claimed several stocks had "gone to zero." That was based on a third-party data source we could not verify. When a reader pushed back, we audited the underlying data against four independent sources — TradingView, Investing.com, Settrade Streaming, and the Stock Exchange of Thailand directly. Several stocks we'd marked as "going to zero" were in fact still trading. Several others had recently undergone reverse stock splits that obscured their actual returns.
We rebuilt the analysis. The corrected numbers are above. They are still bleak — just less hyperbolic.
We are publishing this footnote because we'd rather lose a clickable headline than mislead our readers. If we get the easy stuff wrong, why would you trust the hard stuff?
What we discovered while auditing — the April 2026 reverse-split wave
In April 2026 alone, 17 Thai stocks underwent emergency reverse splits. April 16 had ten of them; April 29 had six more; April 30 added one. A few stocks went through the meat grinder twice — KKC consolidated 37-to-1 on April 16, then another 429-to-1 on April 29. Cumulative ratio: about 16,000 to 1.
This is what dying stocks actually look like. Not "going to zero" — going to a fraction so small that the issuing company is forced to consolidate shares to keep listing standards.
Examples from the wave:
| Stock | Date | Reverse split | Effect |
|---|---|---|---|
| NWR | 2026-04-29 | ~756-to-1 | After 345-day trading halt |
| ECF | 2026-04-30 | ~479-to-1 | After 59-day halt |
| KKC | 2026-04-16 + 4-29 | 37× then 429× | Cumulative ~16,000-to-1 |
| BLISS | 2026-04-29 | ~87-to-1 | 13-day halt |
| AKS | 2026-04-16 | ~100-to-1 | Halt 45 days, resumed at 3 baht |
A DCA buyer who patiently bought 20,000 baht of NWR every month from 2010 onward woke up on April 30, 2026 holding 1/756th the shares they thought they had — at a price 756 times higher per share. Same total economic value. Different optics. The shares didn't go to zero. The capital still did.
We excluded these 18 stocks from the headline numbers above because their post-split prices distort the math. Including them would have made the loser count worse, not better.
Where we stand on TISCO
The TISCO simulation we previewed earlier holds up under audit. TISCO did not undergo a corporate action. The numbers are clean.
If you bought 100,000 baht of TISCO on May 4, 2010 — one trading day before its first ex-dividend date — and reinvested every dividend on the day it paid out for the next sixteen years:
- Today's portfolio value: 1,123,818 baht
- Total return: +1,024% (10.24×)
- CAGR: 16.31%
- 20 dividend events, 19 reinvestments
- Total dividends gross: 579,633 baht (5.8× your initial)
- WHT paid (10%): 57,963 baht
- Final share count: 10,034 (vs initial 3,840 — DRIP added 161% more shares)
- Average cost basis: 9.97 baht per share vs current 112 — an 11.2× markup
If you had received the same dividends but kept the cash instead of reinvesting (no DRIP), your final value would have been only 695,000 baht — a 595% return. DRIP added 429 percentage points by recycling dividends into more compound shares.
One careful initial purchase. One discipline (reinvest the dividends). 16 years of doing nothing else.
The two errors most Thai retail investors make
Error #1: They DCA without selection. They buy "the SET" or "Thai blue chips" indiscriminately. They land in the 59% that lose money — or worse, in the penny stocks that get reverse-split into oblivion.
Error #2: They don't reinvest dividends. They take the cash and use it for cars, travel, lifestyle. The dividends never compound. The 1024% becomes 595%.
The TISCO example shows the path. But it only works if you picked TISCO — not the AE at 7 satang, not the CIG at 3 satang, not the CGD at 11 satang. The Thai market has roughly 730 listed stocks. Roughly 100 are "compounders" in any meaningful sense. The rest are noise.
So the question becomes: how do you systematically identify the 100 instead of being trapped in the 600?
The 10 cleanest 15-year compounders we found
Excluding the corporate-action mess, the top ten in our 180-stock dataset:
| # | Sym | TotRet 15yr | PriceRet | Notes |
|---|---|---|---|---|
| 1 | CMR | +435.6% | −18.0% | Heavy dividend payer despite price decline |
| 2 | CCET | +195.1% | +164.2% | Tech / electronics |
| 3 | AOT | +185.6% | +127.5% | Airports of Thailand |
| 4 | AMARC | +181.6% | +170.9% | Commercial services |
| 5 | AI | +163.8% | +46.5% | Industrial — dividend story |
| 6 | PTT | +146.3% | +3.8% | Energy — almost all return from dividends |
| 7 | ASIAN | +142.4% | +88.0% | Consumer goods |
| 8 | TISCO | +136.1% | +79.3% | Bank — our DRIP example |
| 9 | BOL | +130.6% | +88.0% | Tech services |
| 10 | BANPU | +129.4% | −47.4% | Energy — pure dividend compounding |
Notice how many of these had flat or even negative price returns over 15 years (CMR, PTT, AI, BANPU). Most of the return came from dividends reinvested. Without DRIP discipline, half this list would have been forgettable.
This is exactly the methodology we publish — Track, Study, Wait, Strike — applied to long-horizon compounding rather than swing trading.
What this implies for you
If you're DCA-ing into Thai stocks today without active selection, the math is brutally against you. 59% of stocks will hurt you over 15 years. And about 10% will hurt you so badly the issuer has to reverse-split your shares to maintain listing.
The Thai market is not the S&P 500. The headline 30-year SET return that retail brochures cite obscures a long tail of capital destroyers — and an April 2026 reverse-split wave is the rule, not the exception.
We're not suggesting you stop DCA. We're suggesting you DCA into the 16% strong winners (AOT, PTT, AMARC, TISCO, ADVANC kind of names), with dividend reinvestment, and patience measured in decades.
The methodology we publish — Track, Study, Wait, Strike — exists for exactly this filtering job. Find the survivors before they're consensus winners. Hover. Compound.
Methodology footnote
This analysis used: - 182 Thai stocks tracked by aio.panphol.com (15-year DCA backtest) - Re-validated using set100_prices.db (our authoritative price store, ~1990-present) - 18 stocks excluded for documented corporate actions in April 2026 (reverse splits / forward splits) - TISCO DRIP: real prices from set100_prices.db, real dividends from SET filings - 10% Thai dividend WHT, 0.157% commission per trade, pay-date reinvestment for TISCO simulation - DCA stats use path-independent dividend approximation (dividends × average shares) — small bias vs full path-dependent calculation but consistent across stocks - Cross-validation against TradingView Scanner API (S&P Capital IQ data) for current prices
Research, not advice. Past performance does not guarantee future results. We publish realistic-mode backtest failures including System W (research candidate, not deployed live).
How we work
Built by [Founder Name], with Claude (Anthropic) as a research and development partner. All trade decisions and methodology choices are made by humans; AI accelerates code, backtests, and data analysis. Every published trade is a real position in a real broker account.
When a reader caught us quoting unverified third-party numbers, we audited the data, rebuilt the analysis, and published the corrected numbers along with this footnote. That is the standard we hold ourselves to.
— MOEasymmetry · Track. Study. Wait. Strike.