At a base level, the cause of letting people go is volatility and unpredictability. Not just because of valuations popping or whatever. Good things or bad thi

At a base level, the cause of letting people go is volatility and unpredictability. Not just because of valuations popping or whatever. Good things or bad things happening is not, itself, the direct cause. If you can say "18 months from now, the world will look like this", you can plan for this, good or bad. Maybe you get to "we will run out of money" (or whatever) and have to let people go. Maybe you get to "everything will be great" (or whatever) and you decide to hire despite things sucking now. It is only partially relevant what the world looks like right this second. It depends what you can float, the risk of floating, etc. But with things relatively volatile, you have to plan for unpredictably bad scenarios, or repeatedly lay people off if reality gets worse. Even if you have some idea "the economy will get bad", that isn't enough when the error bars are so high. In these situations it is usually better to overshoot than undershoot - overshooting is painful but "understandable" . Undershooting causing repeated layoffs is usually not.