⚠️ Personal research and trading journal — not investment advice. The author does not provide licensed advisory services.
One of the cleanest results from my assumption audit was also one of the most surprising: a signal that genuinely improves performance in Thailand actively hurts performance in the US.
Same signal. Same logic. Two markets. Opposite outcomes.
Here's what I found, and why it probably makes sense once you understand how different these two markets actually are.
The signal: require a volume pop on the breakout day
My base method — the contracting-base breakout — doesn't require a specific volume level on the breakout day by default. You're watching the base tighten, the pivot form, the price break through. Volume is part of the read, but it's not a hard gate.
The assumption I tested: require volume on the breakout day to be at least 1.5× the prior day's volume. If the stock breaks out quietly, skip it. Only take the breakouts where real buying shows up.
Minervini teaches this. O'Neil teaches this. It looks obvious on a good chart. So I tested whether it actually holds up across decades of data — not just on the memorable examples.
What the Thai data says
I ran a walk-forward test across Thai stocks from 1990 to 2026, split into two sub-periods to check that the result wasn't era-specific.
| Baseline (no volume gate) | With volume pop required | |
|---|---|---|
| Out-of-sample net R | 0.278 | 0.449 |
| Result | Noise (CI includes 0) | Real (CI excludes 0) |
| Better in X of 20 years | — | 13 of 20 |
| Median year | 0.17R | 0.40R |
The improvement holds in both sub-periods. It's not a bull-year artifact — it shows up in the typical year, not just the great ones. This is the test that matters: not the mean, but the median year.
The by-year check is the difference between a real edge and a lucky sample. When most of the mean comes from two or three exceptional years, you don't have a system — you have a lottery ticket with extra steps. When the median year improves, you have something real.
Volume pop passes that test in Thailand. The improvement is consistent, not concentrated.
What the US data says
Same signal, same test, US stocks only.
| Baseline (no volume gate) | With volume pop required | |
|---|---|---|
| Out-of-sample net R | 0.095 | 0.052 |
| Result | Real (CI excludes 0) | Noise (CI includes 0) |
The volume gate doesn't just fail to help — it destroys the edge. Without it, the US baseline has a real (if small) signal. With it, the signal disappears. Requiring volume confirmation in the US makes the system worse.
Why the two markets behave differently
This is the question worth sitting with, because if you don't understand the mechanism, you can't judge when to apply the rule.
In Thailand: The market is shallower, less liquid, and institutional flows are more concentrated. When a stock breaks out of a base on expanding volume in Thailand, it tends to mean real institutional demand surfaced — a visible, confirmable event. The volume spike is a signal because it represents something that didn't exist the day before.
In the US: Large institutions build positions differently. They accumulate shares over days or weeks, splitting orders to minimize market impact. By the time the stock breaks the pivot, the institutional buying may already be partially complete — and the high-volume day might represent the end of the accumulation, not the beginning. US breakouts on lower volume (quiet breakouts out of tight consolidations) have historically been some of the strongest.
This is speculative — I can test the pattern but I can't observe the actual order flow. But the mechanism fits: volume pop as confirmation works when volume is diagnostic. When volume is noisy, requiring it gates out good setups alongside bad ones.
The lesson that applies everywhere
Don't assume edge transfers between markets without testing it.
This is a specific version of a general principle: a finding from one context (one market, one era, one regime) is a hypothesis in another context, not a proven rule. You have to earn the right to apply it somewhere new.
If I'd taken "volume pop improves breakouts" — which is true in Thailand — and wired it into my US system without checking, I would have degraded real US edge with a rule that felt obviously correct but was empirically wrong in that context.
The fact that the same setup has a real edge in both markets (+0.35R Thai, +0.12R US) but responds differently to the same filter is itself informative. Markets have personalities. The job is to learn each one from the data, not borrow rules from a book and assume they generalize.
Thai: require the volume pop. US: don't. Both: check year by year before trusting any filter.
Track. Study. Wait. Strike.
Personal research and trading journal — not investment advice. The author does not provide licensed advisory services. — MOEasymmetry