The turn, the tail, and a published edge that has thinned
The failure is not gradual. Seventy small wins accumulate quietly, and the loss that matters arrives all at once at a trend reversal.
Key takeaway
- At a genuine trend turn the strategy buys repeatedly while the 200-day filter is still permitting it
- Signals cluster, so the bad trades arrive together rather than spread out
- Independent testing since publication has found the edge markedly weaker than the original results
The reversal, where the filter has not caught up
A bull market ends. The index falls, RSI(2) hits 5, and price is still above its 200-day average — the average is a year of data and takes months to be crossed. The system buys. It falls again, buys again, and again.
Every one of those signals is a correct application of the rules. The filter is doing what a 200-day average does, which is to be slow, and slowness is the property that makes it useful in every other situation.
By the time the filter finally turns the strategy off, it may hold several losing positions taken during the descent — each one bought because it was cheaper than the last.
Why the losses arrive together
The strategy's signals are triggered by market-wide declines, so they fire simultaneously across many instruments. In calm markets that means several positions; in a sharp decline it means as many as your rules permit.
This is why a cap on concurrent positions matters more here than a per-trade stop. The tail event is not one position going badly wrong — it is all of them going wrong at once, which is a portfolio failure rather than a trade failure.
The published edge has thinned
Short Term Trading Strategies That Work was published in 2008, on data ending shortly before. Independent testing on subsequent data has generally found the strategy's returns markedly lower than the original figures, and in some tests close to flat after costs.
- Some of it is publication. A widely known, easily coded, purely mechanical signal is exactly the kind that gets arbitraged.
- Some of it is market structure. Short-term reversal effects have weakened as electronic market making has become faster and cheaper.
- Some of it was always sample-specific. The original tests covered a period unusually favourable to buying dips.
The specific psychological trap of a high win rate
Winning three trades in four produces a strong and false sense of skill. The natural response to a run of wins is to increase position size — and the increase almost always happens before the loss rather than after it.
Fixing position size in advance, in writing, and not revising it in response to a winning streak is the single most valuable discipline for this particular strategy.
Common questions
- Is RSI(2) still worth trading?
- Short-term mean reversion in equity indices remains a documented effect, and the strategy is a clean expression of it. What has changed is the size of the edge — treat this as a small effect requiring low costs and disciplined sizing, not as the results in the book.
- Would a shorter trend filter help at the turn?
- A faster filter — a 100-day or 50-day average — turns the strategy off sooner in a decline, and also turns it off during ordinary pullbacks within a bull market, which is when most of the good trades occur. The published testing found the 200-day a reasonable compromise, and the trade-off is genuine rather than an unfound optimum.
These are documented methods described for study. Nothing here is investment advice, a recommendation, or a claim about future returns — every system on this page has losing periods, and the pages say where.
Reading about a system is not having one.
Plutux is where you write your own rules down, test them against real data, and keep the record your memory would otherwise rewrite. Join the waitlist for early access.