Here's a thread about short-selling. When someone sells a stock short, they borrow a share from someone else and immediately sell it. Then later they have to buy a share and return it to its owner. If the stock's price falls in between, the short seller makes a profit.

A key thing that a lot of people miss (including me until a few days ago) is that every act of short-selling also creates an additional long position. The guy the short borrowed the stock from still has a long position, and so does whoever you sold the share to.

So when you take a big short position on a stock you're effectively flooding the market with newly created shares. Expanding the supply of anything pushes down its price.

To make this concrete, suppose a particular company has 100 shares in circulation (known as the "float"), and someone takes a 50-share short position. That will create an addition 50 long positions, for a total of 150.

Looking at things this way made a couple of things clearer for me. First, it helped me understand what's going on in a short squeeze. This is the phenomenon where short sellers closing their position pushes up the share price.

The reason is that when someone closes a short position, they're effectively taking a share of the stock out of circulation. Just as expanding the supply of the stock pushes its value down, so contracting the supply of the stock pushes its value up.

But crucially this only happens if the total number of short positions declines. If somebody closes a short position but then immediately opens a new one—or someone else comes in and takes an equal-sized short position—then there's no upward pressure on the price.

This is important because the premise of Redditors' GME gambit was that if they can push the share price up high enough, a short squeeze will push the stock still higher.

But even if the Redditors manage to bankrupt the initial short sellers and force them out, this won't cause a squeeze unless the total number of shorts falls. If an equal number of others sell short (perhaps attracted by the new, even more ludicrous price), no squeeze occurs.

The second important implication here is that there's nothing special about a stock with more than 100 percent short interest. There's been a lot made of the fact that the number of short positions in GME exceeded the number of shares in circulation.

That sounds impossible if you assume that you can't sell more than 100 percent of outstanding shares. But it makes more sense when you remember that every short position creates a corresponding long position.

Suppose a company has 100 shares in circulation and people have shorted it for 150 shares. In this case, there are 250 people holding long positions—100 people holding actual shares and 150 people holding promises by shorts to get shares back later.

Part of the Reddit theory is that if Redditors buy enough shares of GME, the shorts will be forced to pay them arbitrarily high amounts to close out their short positions.

But even if reddit buys 100 percent of the GME float (outstanding shares), they still hold less than half the outstanding long positions if the short interest is >100 percent.

This brings me to the most important mistake the Redditors seem to be making: they assume that since shorts hold >100 percent short interest, that when "the short squeeze" happens, shorts will need to buy every share and every redditor will be able to cash out at a high price.

This is wrong in two ways. First "the short squeeze" is likely to be a gradual process, not a single event. There won't be a single time when everyone gets to cash out.

Second, this ignores that >100 percent short positions necessarily means >200 percent long positions. So even if every short closes out their position, there will still be a lot of people holding overpriced GME stock—likely including a lot of gullible redditors.