We Tested "Buy the Dip" for 25 Years. Here's the Result.

ELI5/TLDR

Zerodha ran 25 years of Nifty data to settle an old argument: is it smarter to invest a fixed amount every month (a boring SIP), or to hoard cash and pounce when the market dips? They gave the dip-buyer a cheat code — perfect hindsight, so it always bought at the exact bottom. The dip-buyer still barely won: about 2.6 crore versus the SIP’s 2.5 crore over 25 years. The effort was enormous; the payoff was a rounding error.

The Full Story

The setup

The test pits two investors against each other, both putting aside 10,000 rupees a month from 2000 to 2026. The SIP investor dumps it straight into the Nifty 50 the moment it arrives. The dip-buyer parks it in a savings account earning 6%, waiting for the market to fall by some threshold before deploying.

A nice touch: they used the Nifty 50 total return index, not the price index you see on TV. The total return version assumes dividends get reinvested.

The dividends reinvested can add a meaningful chunk to your final corpus. 1% or more every year.

Over 20 years that 1% compounds into a real gap, so it’s the honest benchmark.

The dip-buyer ran four flavors of patience: buy on a 2–5% fall, a 6–10% fall, a 10–20% fall, or only on a full-blown 20%+ bear market. And crucially, the model used perfect hindsight — it always knew exactly when the recovery would arrive, so it spread its cash perfectly across the down days and never bought too early.

In our calculation, we have assumed that there’s perfect hindsight, which means we are giving the buy the dip the cheat code to win against SIPs.

The result

The boring SIP ended with about 2.5 crore, on roughly 31 lakh of actual money invested. The dip-buyers landed at 2.57, 2.59, 2.60, and 2.57 crore across the four scenarios. The best case beat the SIP by maybe 2%.

After all that tracking, waiting, and that excitement of timing the market perfectly, the best-case outcome is 2.6 crore versus the SIP’s 2.54 crores.

Why hoarding cash loses

Markets spend far more time rising than falling, and a big slice of the total return hides in a handful of days each year — sometimes ten or fifteen. If your money is sitting in a savings account waiting for a dip, it tends to miss exactly those days. Add the opportunity cost: every month the cash earns 6% instead of the market’s 12–14%, and across hundreds of waiting months that gap compounds into real losses.

The patient bear-market hunter (scenario four) has the worst version of this problem. Big crashes are rare. You might wait 18, 24, even 36 months while the Nifty quietly climbs from 15,000 to 22,000 — a rally you missed because it never fell 20% first.

The honest caveat

Yes, if you had bought at the literal bottom of 2008 or March 2020, returns would have been extraordinary. But the model already assumed you could — perfect hindsight — and it still barely moved the needle. In real life, you’d have been frozen by panic headlines, and even the brave usually deploy a timid amount and never put in what they meant to.

The takeaway framework

Automate the SIP with your regular income and forget it. Use bonuses or windfalls — money outside your budget — to scratch the dip-buying itch.

Stay boring.

Key Takeaways

Over 25 years (2000–2026), a 10k/month Nifty SIP grew ~31 lakh invested into ~2.5 crore; perfect-hindsight dip-buying topped out at ~2.6 crore — a ~2% edge for vastly more effort.
The test used the Nifty 50 Total Return Index (dividends reinvested), which adds ~1%+ per year over the price index — the correct long-term benchmark.
Markets spend far more time going up than down, so cash on the sidelines structurally underperforms by being absent.
A disproportionate share of annual returns comes from ~10–15 trading days; cash waiting for a dip routinely misses them.
Opportunity cost is the silent killer: idle cash at 6% versus the market’s 12–14% compounds into real loss over hundreds of waiting months.
Waiting for a 20%+ crash is the worst strategy — crashes are rare, and you can sit out multi-year rallies that never dipped enough to trigger a buy.
Even perfect timing of 2008/2020 bottoms barely helped in the model — and in reality people freeze during panics or deploy too little.
Results held across rolling 5-year and 10-year windows, not just the full 25-year run.
Practical rule: SIP your salary on autopilot; reserve only windfalls/bonuses for occasional dip-buying to satisfy the thrill without derailing the plan.

Claude’s Take

This is a well-constructed piece of propaganda for index discipline, and I mean that as a compliment. The clever move is handicapping the experiment in the dip-buyer’s favor — perfect hindsight is an absurd advantage, and the strategy still only wins by a whisker. That’s a much stronger argument than a naive backtest, because it pre-empts the “but I’ll time it right” objection. The dip-buyer’s tiny win is real, though, and they’re honest about it: the strategy does edge out the SIP in most windows, just not by enough to justify the effort.

Two things to keep in mind. First, the conclusion is specific to a relentlessly upward market over this particular 25-year stretch; a flatter or longer-sideways market would shift the math toward the patient buyer. Second, “perfect hindsight” cuts both ways as a framing — it inflates the dip-buyer’s plausible result, so the real-world gap is probably negative, which is arguably the more useful headline. The 6% savings-account assumption is also reasonable but load-bearing; a higher idle yield would narrow the gap.

Knocking a point off because it’s an explainer with a foregone conclusion (Zerodha sells SIPs) rather than a neutral investigation, and the underlying analysis isn’t shown — we take the numbers on trust. But the logic is sound and the framing is genuinely smart. A 7.