Buy-the-dip or buy-the-bottomless-pit?

So the U.S. stock market crashed and you’re significantly poorer than you were last month. What should you do? Is now the time for you to be a true hero and boldly buy the dip while lesser minds panic?

I don’t think so. While there are reasonable arguments for buying in April 2025, the quasi-religious belief that the U.S. stock market always goes up is not one of them.

Here’s my view:

1. Should you buy the dip? Probably not, except as part of an existing rebalancing policy.
2. How bad could it get? Very bad.
3. Is the market cheap now? No.
4. What’s the dumb money doing? Buying the dip.
5. Are there any good reasons to buy right now? Yes.
6. Are there any good reasons to sell right now? Yes.

Should you buy the dip?

Contrarian investing is often a good idea, especially when implemented in a rules-based investment program. For example, the typical target date fund has a policy of rebalancing to fixed portfolio weights. When the stock market falls, the fund buys stocks. This type of systematic dip-buying makes sense as it sells stocks when they get expensive, and buys them back when they get cheap.

However, it’s generally not a smart idea to buy an asset simply because you notice that it has fallen greatly in price. “Buying-the-dip,” as implemented by retail investors, often results in disaster. For example, as I’ve previously discussed, Korean retail investors bought Lehman in August 2008 and Silicon Valley Bank in March 2023, both of which fell to about zero subsequently.

Dip-buying the aggregate equity market is often unsuccessful. If you bought the dip after the 1929 crash – participating in the Sucker’s Rally of 1930 – things did not end well for you. Similarly, if you bought the dip following Warren Buffett’s New York Times op-ed of October 2008, you lost more than 25% by March 2009.[1] Of course, Buffett’s advice worked out fine in the long-run, but it’s not obvious that this episode supports “buy-the-dip” as opposed to just “buy-and-hold.”

Both Buffett and target date funds follow contrarian strategies that do not involve one-time impulse buying, but are implemented over long timeframes. Buffett came into 2025 holding high amounts of cash, ready to deploy his dry powder if stock prices got too cheap. Similarly, target date funds had been selling equity as the market rose in 2024. So for them, buying the dip makes sense. But if you weren’t selling the rip in 2024, you shouldn’t be buying the dip in 2025.

Are there any times when discretionary dip-buying makes sense? I say yes. If there’s an obvious technical screw-up and temporary shortage of liquidity, such as the flash crash of May 2010, stepping in to provide liquidity makes sense. You might argue the crash of October 1987 is in this category; when stock prices fall for no good reason, maybe you should buy.

However, April 2025 is nothing like May 2010 or October 1987. The April 2025 decline in stock prices was in response to obvious bad news, and there’s no particular reason to think prices will snap back.

How bad can it get?

It can get very bad. During the Depression, the U.S. stock market had drawdowns of more than 80%. We saw drawdowns of more than 50% in 2009 and more than 40% in both 1974 and 2002. I’m not predicting a drawdown of 50%, but it’s definitely in the realm of possibility. If you can’t survive an equity drawdown of 50%, you’re holding too much equity.

Is the market cheap now?

The U.S. stock market is not cheap, either relative to trailing fundamentals or to forward fundamentals.

Looking first at trailing earnings, in March 2025, Shiller’s CAPE was 35. That’s very expensive. In March 2009, CAPE was 13. That’s cheap. To get from 35 to 13, prices would need to fall by a total of 63% from their level in March 2025. We have a long way to go before U.S. equities look cheap relative to history.

Second, on a forward-looking basis, we can ask whether the price decline in April 2025 was accompanied by a decline in expected fundamentals. One clue comes from dividend futures on the S&P 500. Let’s look at the futures price of 2027 dividends, which fell by about 11% in the week ending April 8, about the same amount as the decline in the S&P 500. Thus, the fall in price seems totally justified by the fall in future fundamentals. While there are a number of reasons that dividend futures are not the same as dividend expectations, we can at least say that it does not appear that the market got cheaper in April 2025 on a forward-looking basis.

What’s the dumb money doing?

Buy-the-dip is an article of religious faith among retail investors, and in April 2025, they kept the faith:

Individual investors net bought a record $4.7 billion worth of stocks on Thursday as new tariffs pummeled markets.[2]

Retail investors often follow contrarian buying strategies, sometimes with good results and sometimes with bad. Around market peaks, we generally see the smart money selling to the dumb money, as discussed here.

Retail investors in U.S. equities appear to believe that short-term volatility is merely “noise” and in the long-term the stock market is virtually riskless. It’s a delusion, as I’ve discussed here and here. While it’s true that the aggregate U.S. stock market has a temporary component that could be described as “noise,” the market also has a permanent component reflecting expected fundamentals. If we’re about to enter a ferocious recession in 2025, the stock market decline is not “noise” and buying the dip is a mistake.

Are there any good reasons to buy right now?

In April 2025, VIX spiked to an extremely high level over 50. While we don’t have a lot of observations of historical periods when VIX is elevated, those observations that we do have suggest that high VIX leads to high future returns. In other words, when fear is high and there’s blood in the streets, that’s a good time to buy equities if you are risk neutral.

Now, before explaining this result further, let me caution: just because the expected return on U.S. stocks has risen, that doesn’t mean you should be buying stocks, because risk has also risen. If, for example, the annualized expected return doubled from 10% to 20%, but the volatility tripled from 20% to 60%, a risk averse investor should be selling stocks, not buying.

Martin (2017) develops a measure of expected stock returns implied by options prices. Now, his measure is not exactly the same as VIX (he weights implied volatility of different options differently) but it more or less implies that when VIX is high today, we predict high stock returns over the next month. His measure suggests that annualized expected returns were over 20% at the height of the GFC. Martin’s findings have been replicated by Lou, Polk, and Skouras (2024). The predictive power of VIX had an out-of-sample confirmation in March 2020: VIX incredibly high, good time to buy stocks. VIX also spiked in August 2024, and although you could argue that this episode was a meaningless blip, it still turned out to be a good time to buy stocks.

So, we have respectable and evidence-based arguments for buying in April 2025.

Are there any good reasons to sell right now?

The main reason you might want to sell equities in April 2025 is prudent risk reduction.

A standard result from one-period mean-variance math is that your portfolio allocation to equities should be proportional to the ratio of expected return to variance. For example, let’s say that prior to April 2025, you were 60% in equities, and equities had volatility of 20%. Then in April 2025, volatility triples to 60% while expected return remains unchanged.

The standard formula says in this scenario, you should cut your equity allocation from 60% to 7%. Pretty dramatic! You should not follow this advice, because it is based on static optimization and we live in a dynamic, multi-period world. However, I think there’s wisdom here: selling in response to higher volatility is not necessarily irrational panic. It could be a perfectly reasoned and prudent response to changed market conditions. One thing we’re absolutely sure about is that volatility is clustered over time, so that if realized volatility is high this week, it will also be high next week.

Moving to more realistic dynamic strategies, Moreira and Muir (2017, 2019) advocate decreasing equity allocations when volatility goes up. “Volatility scaling” or “volatility managed portfolios” would make sense in a world where higher volatility is not accompanied by higher expected return. The idea is similar to low volatility investing: why take risk if that risk is not compensated? While volatility scaling may not be tremendously valuable in practice (as argued by Cederburg et al (2020)), it makes sense as a general principle.

Conclusion

A basic property of optimal decision-making is that you should act more conservatively when you have less accurate information. As of April 2025, we’re confronted with a situation where we have very little historical precedent and the outcomes we observe are extremely volatile. As a logical consequence, we should be cautious in responding to events.

So, whatever it is you do in response to the stock market decline of April 2025, it should probably be a modest tweak to your existing portfolio. If you were 60/40 before April 2025, a reasonable response might be to switch to 50/50 or 70/30, but not 100/0 or 0/100. Swashbuckling macro calls make little sense.

Let me conclude with a comment on language. A “dip” sounds like a temporary and mild decrease, so “buy-the-dip” sounds smart. The truth is that we don’t know if April 2025 is just a dip in the road leading to the sunlit uplands of stock market Utopia, or instead is the initial phase of Great Depression II, leading to a bottomless pit of investor despair. “Buy-the-bottomless-pit” sounds a lot less appealing than “buy-the-dip.”

Endnotes

[1] Buffett, Warren E. "Buy American. I am." The New York Times, October 16, 2008.

[2] "Individual investors net bought a record $4.7 billion worth of stocks on Thursday as new tariffs pummeled markets," Dow Jones, April 4, 2025.

References

Cederburg, Scott, Michael S. O’Doherty, Feifei Wang, and Xuemin Sterling Yan. "On the performance of volatility-managed portfolios." Journal of financial Economics 138, no. 1 (2020): 95-117.

Lou, Dong, Christopher Polk, and Spyros Skouras. "The day destroys the night, night extends the day: A clientele perspective on equity premium variation." London School of Economics Working Paper (2024).

Martin, Ian. "What is the Expected Return on the Market?." The Quarterly Journal of Economics 132, no. 1 (2017): 367-433.

Moreira, Alan, and Tyler Muir. "Volatility‐managed portfolios."The Journal of Finance 72, no. 4 (2017): 1611-1644.

Moreira, Alan, and Tyler Muir. "Should long-term investors time volatility?." Journal of Financial Economics 131, no. 3 (2019): 507-527.