We ran 40 textbook strategies on US data twice — on today's surviving large-caps, and on the survivorship-free S&P 500, Nasdaq and Dow indices (back to 1927). The gap between the two answers is the whole lesson.
On a basket of today's surviving US giants, 32 of the 38 long strategies scored almost exactly the same — around 73/100. That's not real skill; it's survivorship bias. On a universe of pre-selected winners, almost any 'buy the trend' rule looks alike, and the honesty score docks them all for the same reason.
Run the same strategies on the actual S&P 500 / Nasdaq / Dow index series — survivorship-free by construction, nearly a century of data — and the scores spread out from 29 to 89/100. Now you can actually tell the strategies apart.
| Strategy | US large-caps (survivor-biased) |
US index (survivorship-free) |
India index (cross-check) |
|---|---|---|---|
| the Golden Cross | 73 | 85 | 75 |
| buying the RSI dip | 73 | 84 | 56 |
| SuperTrend | 73 | 88 | 78 |
| the MACD crossover | 73 | 89 | 75 |
| shorting the Death Cross | 16 | 29 | 23 |
Same strategies, same engine — only the data changed. The biased column flattens; the two survivorship-free columns rank them. That gap is exactly what a normal backtest quietly hides.
Long strategies score high on the US index partly because the index itself trended up for a century — a rising tide lifts them. A high honesty score means a strategy survives statistical scrutiny (it isn't curve-fit or lucky), not that it beats simply buying and holding the index. That's a different, equally important question — and the reason to test your own exact rules rather than trust a headline.
Test any strategy on the US indices — free
Survivorship-free S&P 500, Nasdaq and Dow — the honest US benchmark. No signup needed to see the score.
Run it on US indices →See every strategy ranked on survivorship-free data →
Get the monthly honest teardown
One popular strategy put on trial, in your inbox — once a month. No signals, no hype.
Educational, not financial advice. An honest backtest score measures whether a result survives statistical scrutiny — it does not predict future profit, and is not a buy or sell signal. · About · FAQ · All strategies