What to do when the market drops — a checklist for the actual moment

You already know this. Knowing it is not the problem.

Why advice fails in the moment

Every article about market declines is written in a calm market and read in a falling one. That mismatch is the whole difficulty.

During a real decline, three things are true simultaneously:

• The information environment inverts. Every headline is bad, every analyst is cautious, and each one sounds more credible than the last. It is genuinely hard to hold a constructive view when nothing constructive is being published.
• Doing nothing feels like negligence. Your money is visibly shrinking and you are visibly not acting. The urge to do something isn't stupidity — it's the same instinct that's correct in almost every other domain of life.
• Selling works immediately. It stops the discomfort the second the order fills. The cost shows up much later, quietly, as a recovery you weren't in. That's the trade structure humans are worst at refusing, and it's the mechanism behind panic selling.

So the goal isn't to have better opinions during a decline. It's to have fewer decisions to make.

The checklist

In order. Steps 1 and 2 cost nothing and prevent most of the damage.

1. Establish the timeline before you touch anything

One question: when do I need this specific money?

• Not for 10+ years — a decline is not an event in this money's life. It's a price change on something you weren't going to sell anyway.
• 3–10 years — you should still hold, but this is worth knowing about your allocation later, when things are calm.
• Under 3 years — this money shouldn't have been fully exposed, and now is the worst possible time to fix it. Fix it after a recovery, not during a decline.

Most panic comes from money whose timeline was never defined. Define it and a large fraction of the anxiety turns out to be about a horizon you don't actually have.

2. Keep the automatic contributions running

If you contribute monthly, the single most valuable thing you can do during a decline is nothing — specifically, not touching that transfer.

A scheduled contribution during a decline buys more shares at lower prices. That's the entire benefit of dollar-cost averaging, and it only exists if you don't pause it exactly when it's working. Pausing feels prudent ("why throw money into a falling market?") and it removes the one advantage a long-horizon investor structurally has. More on the tradeoff in dollar-cost averaging vs lump sum.

3. Check whether anything actually changed

Now do the analytical part, and be strict about what counts.

For each individual position, go back to what you wrote when you bought it and check the falsifying condition. Not the price — the condition. Did the thing you said would prove you wrong happen?

Usually it hasn't. Broad declines move nearly everything, including companies whose situation is unchanged. Occasionally it has, and then you have a real decision — but it's a thesis decision, not a panic decision, and it's the one case where selling during a decline is correct. (If you don't have a written thesis to check, that's the gap to close afterward — here's the ten-minute version.)

4. Rebalance only if you're meaningfully off target

If you set 80/20 and the decline has pushed you to 70/30, rebalancing back is mechanical maintenance, and it happens to mean buying what fell. That's the discipline working.

Two cautions. Rebalance to a rule, not to a feeling — a threshold you set in advance, like five percentage points of drift. And don't use rebalancing as cover for a de-risking you actually want emotionally; moving to a "safer" allocation in the middle of a decline is just selling low with better vocabulary.

5. Harvest tax losses, if that's relevant

In a taxable account, selling a position at a loss and buying a similar-but-not-identical exposure keeps you invested while banking a loss against future gains. Watch the wash-sale rules; tax specifics vary by country and situation, so this is the one step worth confirming with someone who knows yours.

6. Write down what you're feeling

Sixty seconds, timestamped. What you want to do right now and why.

This costs nothing and it's the highest-return step in the list, because it's the only one that makes you better next time. In eighteen months, that note will tell you something no market history can: what you're actually like during a decline. Most people find the note embarrassing, which is the point — you can't correct a reaction you don't remember having.

What not to do

Don't check the balance hourly. Volatility is drama at a five-minute interval and noise at a five-year one. The frequency you look at is a choice, and it directly sets how much pressure you're under.

Don't sell to "get back in lower." That's two decisions. The first is easy and the second is the one nobody executes — the recovery's best days cluster near the bottom, when getting back in feels most reckless. The full mechanism is here.

Don't go hunting for confirmation. During a decline you can find an authoritative case for anything. Reading more is not analysis at this point; it's anxiety management with a productive-looking interface.

Don't concentrate into the "obvious" bargain. Declines produce strong convictions about specific beaten-down names. Some of them are right. Sizing them like they're certain is how people turn a market decline into a personal one.

The part you can prepare in advance

Everything above is easier if you've done it before. The problem is that real declines are rare and expensive teachers — you get a handful in an investing lifetime, and each one charges tuition.

That's the specific gap historical crash scenarios fill. You go through a real decline — 2008, the dot-com bust, 2022 — day by day, with the real prices and the real headlines, without knowing where the bottom is. You make the actual decisions under something reasonably close to the actual pressure, and at the end you see what you did.

Most people find out they sell. Better to find that out in a simulator, where the tuition is an afternoon, than in the market, where it's a decade of compounding.

FAQ

What should I do when the stock market drops?

Do nothing irreversible first. Confirm when you actually need the money, keep automatic contributions running, and check whether anything you own has fundamentally changed rather than just fallen in price. Then, if applicable, rebalance to a preset threshold and harvest tax losses. The main risk in a decline is converting a temporary loss into a permanent one by selling.

Should I stop investing when the market is falling?

Usually the opposite — pausing contributions during a decline removes the main advantage of investing regularly, which is buying more shares when prices are lower. The exception is if you need that cash for living expenses or a near-term obligation, in which case the priority is stability, not returns.

Should I sell my stocks before the market crashes further?

Selling into a decline requires being right twice: about when to exit and about when to re-enter. The second is far harder, because the strongest recovery days cluster near the bottom when buying feels most irrational. Sell an individual position only if the specific thesis you bought it on has broken — not because the price is falling.

How long do market crashes last?

Historically it varies enormously — the 2020 COVID decline recovered within months, while the dot-com bust took years for the Nasdaq to reclaim its high. That range is exactly why timing the re-entry is unreliable, and why plans that depend on knowing the duration tend to fail.

How do I stop panicking when my portfolio is down?

Reduce the number of decisions rather than trying to feel calmer. Define in advance when you need the money, automate contributions, write the falsifying condition for each position at purchase, and check your balance less often. Rehearsal helps most: going through a realistic historical decline shows you your own reaction before it's expensive.

---

You will not think clearly during the next decline. Nobody does. The work is making that not matter. Rehearse one →