Does buying the dip work on Nvidia?

Rarely. In 110 backtests over about 28 years of Nvidia (NVDA) prices, buying the dip beat simply holding the shares in 3. On average it trailed holding by 12.0 percentage points a year.

Most trades made money, and it still trailed holding

The rule's trades usually worked. Of 401 closed trades at default settings, 329 (82.0%) made money. That is the figure most dip-buying claims quote, and it answers a different question. A win rate counts trades that closed in profit. It says nothing about the time spent in cash waiting for the next fall.

Nvidia mostly rose over the years tested: its dominant market regime is a bull market. Each time the rule sold on a recovery, it waited for another fall before buying back. On a share that kept climbing, the gains missed while waiting were larger than the gains the dips paid.

The falls were still deep

Buying after a fall did not avoid the big falls. The average backtest still had a worst fall of 74.1% from a peak.

Where buying the dip has held up

Across every ticker StrategyLens tests, the two dip-buying rules beat holding in 89.8% of 14,775 backtests run through crash periods. Nvidia's history is mostly a rising market, so that pattern had little chance to show here.

How this was tested

"Buying the dip" here is the pullback rule: buy after a fall, sell on recovery or at a stop. The trade figures use its default settings. All figures assume no trading costs, are before tax and carry survivorship bias. They are simulated past results, not a forecast or a recommendation. Figures from the NVDA card dated 29 October 2026; the full verdict is on the NVDA page and the method on the methodology page.