I've noticed that companies' stock prices don't always correlate to the performance of their past quarter. Wonder why

Change in stock price on earnings news tends to correlate to the performance relative to expectations. The estimated performance is already priced in, so the movement will be based on how much better or worse the actual results were than the estimated/anticipated ones. In this case, it was obviously a bad quarter for Tesla, but the bump suggests that traders expected it to be even worse.

> but the bump suggests that traders expected it to be even worse.

You're a bit too kind with your wording, it doesn't _suggests_ so much as it literally _means_ that. Plenty of websites show you what investors are expecting, if Tesla says more/less price fluctuates accordingly.

> Plenty of websites show you what investors are expecting

Yes and no. Plenty of websites make predictions. Most money in the market however is controlled by institutions that don't tell you shit about what they are expecting or anything else that they aren't legally required to tell you.

Normally, and apparently in this case, yes. It is possible for a stock to go up on earnings that don't beat expectations though, due to other concurrent factors, like unexpected positive news regarding future prospects, or even an unrelated broad market movement.

To add something to this, traders have their own expectations of how companies will perform relative to their published estimates. I used to work at a unicorn that would consistently exceed its published estimates, and the stock price took a non-trivial hit the first time it ever met them.

If the news is expected to be bad, but is actually less bad than anticipated, then the stock goes up, even if it's bad. There's been a lot of that lately.

Stock price is extremely sensitive to both guidance from the company and analysis. There's almost always an analyst consensus on the expected performance of the stock given the knowns about the companys performance and the market. The stock is usually pegged right at that consensus, so if they lose less money than than the consensus was expecting (even if they lost more money) then the price goes up.

You buy stocks based on what you think they’ll do next, not what they did previously.

In what sense do you mean they don't correlate to performance? Overperforming or underperforming expectations will correlate very highly to the way the stock moves on the news

Some combination of momentum with the market or sector, pricing in future growth, an earnings call that does a good job explaining the bad quarter.

Stock price is forward looking, not past.

Maybe the market is not so intelligent after all.

After-hours trading doesn't necessarily mean much due to low liquidity and volume. You can get high swing for reasonably low volume in a single after-market trade.