STOCKS · STRATEGY TEST
Does buying the dip work on NVDA?
Buy when RSI drops below 30 — the textbook “oversold” signal — and hold for 20 trading days.
Holding came out ahead
23.22%
buying the dip per year
56.15%
Buy & hold per year
-32.93%
Difference
rule vs held, over 9.2 years
43%
Worst drawdown
holding fell 66%
Holding NVDA came out ahead of buying the dip over 9.2 years. Across the last 9.2 years of real prices, simply buying and holding Nvidia made 32.93 percentage points a year more than buying the dip (56.15% versus 23.22%). The rule did give a gentler ride: its worst fall was 43.5% against 66.3% for holding — a smoother path, paid for with some of the return.
Every rule on NVDA, ranked
Buy and hold · benchmark56.15%
5923% total · worst drawdown 66.3% · in market 100%
200-day moving average timing55.48%
5690% total · worst drawdown 41.8% · in market 79%
Golden cross (50/200)55.05%
5547% total · worst drawdown 40.1% · in market 80%
Buy the dip (RSI < 30, hold 20d)23.22%
582% total · worst drawdown 43.5% · in market 25%
Sell in May (hold Nov–Apr)16.71%
314% total · worst drawdown 45.9% · in market 48%
the last 9.2 years · 2,312 trading days · $3.523 → $212.17
How to read this
One asset over one window is a single data point, not a law. A rule that comes out ahead on NVDA may trail holding on the next ticker you try — which is exactly why the free tester exists: change the asset and see whether the edge survives. When it evaporates on a different market, it was fitted to this one.
Note the drawdown column as much as the return. Most timing rules earn less than holding but fall less in a crash, and whether that trade is worth it depends entirely on whether you would actually have held through the worst of it.
Method
All strategies are measured from the same starting bar so the comparison is like-for-like. Signals are read from the previous close and the next day's return is applied — no lookahead. Prices are dividend-adjusted so buy-and-hold is not handicapped. 0.05% cost per switch. No leverage, no shorting.
Test the same rule on something else
Get the next autopsy
We test one strategy people swear by and publish whatever the data says — including when it says the strategy doesn't work. Free, no pitch.
Vorrik Research publishes tests of publicly discussed trading strategies using public market data. Where a test was also run with our own money, the article says so. It is factual reporting of what happened in historical and past live tests — not investment advice, and not a recommendation to buy or sell anything. Past results never predict future results.