Stock Market Crash: What to Do (and What Not to Do)

Step 1 — Separate the shock from your position

The first question is not “how bad is this?” but “how much of this touches me?” Write down what you actually hold, what share of your net worth it represents, and when you need the money. An investor with a ten-year horizon and no leverage faces a completely different problem from someone who needs their capital next quarter — even though both are reading the same headline.

Step 2 — Check whether the reason to own has changed

Price is not evidence. Ask whether earnings power, balance-sheet strength or the competitive position of what you own has actually deteriorated. In broad crashes, most holdings fall because of correlation and liquidity, not because their businesses changed in a single week.

• Has the company or index reported anything new, or only the market reacted?
• Is the decline concentrated in one sector, or indiscriminate?
• Are credit markets confirming the equity move, or diverging from it?

Step 3 — Measure the gap between crowd fear and expert conviction

This is the core of the framework. Retail sentiment collapses far faster than analyst conviction does. When the two diverge sharply — crowd sentiment near the floor while accuracy-weighted expert consensus stays intact — the drawdown is behaviorally driven, and historically those are the worst moments to sell.

Step 4 — Choose an action from a pre-written list

• Do nothing. The default, and statistically the strongest option for diversified long-term holders.
• Rebalance. Restore your target weights mechanically, which naturally buys what fell most.
• Add on a schedule. Fixed amounts on fixed dates remove the need to call a bottom.
• Trim a genuinely broken thesis. Only when step 2 produced real evidence, not a feeling.

What not to do

• Sell everything to “wait for clarity” — clarity arrives after prices recover.
• Check your portfolio hourly; frequency of looking is strongly linked to frequency of selling.
• Add leverage to recover losses faster.
• Let a single headline replace your written plan.

The recovery context

Crashes look unique while you are inside them and remarkably similar afterwards. Comparing the current drawdown’s shape, speed and breadth with historical episodes of the same profile turns an emotional question into a statistical one — which is exactly what the Panic Score is built to do.