# Buying Stocks After a Crash — 50,000 Deep-Drawdown Purchases Tested
- Source: FoldAlpha Research (https://app.foldalpha.com/en/research/dip-buying) · Published: 2026-09-01 · Series: Myth Testing 8

"Buy when it drops," "buy when there's fear" — we tested this advice at the level of individual stocks.
The conclusion first: buying deep drawdowns in individual stocks underperformed the market, and the one exception showed up not at the stock level but only at the index level.

## Test results

| Signal (entry at observation, 20-day hold) | Sample | Excess vs. market | Beat-market rate |
|---|---|---|---|
| Buy stocks down ≥15% over 5 days | 49,694 | **-3.8%p** | 39.4% |
| Chase stocks that spiked after capitulation (a drop of -8% or more) | 4,482 | -7.4%p | 35% |
| Bet on a next-day rebound after a down day (conditioned on investor flow direction) | 4,141 days | all conditions p>0.1 | — |

More on the third row: we tested flow-conditioned rebound hypotheses such as "down days when foreign investors were buying rise the next day"
across every down day in 16 years (4,141 days), and no flow condition had statistically significant predictive power.
The correlation between the direction of foreign investors' net buying on a down day and the next day's index return was -0.01 — investor flows are a coincident indicator, not a leading one.

## What these numbers say

A large drawdown in an individual stock is often a signal of "bad information came out," not of "it got cheap."
The group of deeply drawn-down stocks kept underperforming the market on average — declines have reasons,
and buying based on the drawdown alone, without knowing the reason, means standing on the other side of an informed seller.

That said, **one exception was observed at the index level.** Buying down days in the extreme zone where KOSPI sat 5% or more below its 20-day moving average
had a 60% win rate over 16 years — this lone survivor is covered in
[What Survived](/en/research/small-survivors), along with the limitation that the opportunity arises only 2.4 days a year.

## Limitations

- The drawdown threshold (-15% over 5 days) is one definition. An exhaustive search over other drawdown/window combinations would certainly turn up combinations that "work," but those are likely artifacts of multiple testing, so we did not run it
- Drawdown buying combined with financial conditions (profitability, low debt, etc.) is outside the scope of this test

## Data sources

- Korea Exchange daily prices and investor-type trading activity. Daily investor flows are available in the [flow report](/flow)

---
This article documents tests on historical data for informational purposes only. It is not investment advice or a recommendation to buy or sell any security. Past test results do not guarantee future returns.