Ex derivs trader here. If you want to understand this, here's a short explanation.
1. Options have asymmetric payoff. One way you win, the other way you don't lose, assuming you are the buyer. The opposite is true for the seller.
2. Because of the asymmetry, movement is positive for the owner. So the option is priced based on how much movement is expected. There's a distinction here between vega and gamma, which are both movement related. Vega is sensitivity to the appraised future volatility, which means the option has more of that if it's long dated. Gamma is (2nd deriv) sensitivity to movement in the current price, so you'll have more of that if you're near the expiry.
3. As an options market maker, you are not in it to guess which way the underlying stock goes, you are in it to gain an edge in pricing that movement risk. So you hedge it when you trade an option with someone. If you buy a call, you sell some underlying so that your delta is flat. (Delta is just the 1st order price sensitivity)
4. If you flattened your delta, you can imagine the gamma making your local PnL vs price a parabola. If you are long the option you'll make money when the price moves either way, and you'll be needing to buy low and sell high to get back to flat delta. (Incidentally this is also why we don't really care whether it was literally a put or a call, as in the right to buy or sell. The 2nd order is the same, so has the same effect on your PnL when the price moves.)
5. The opposite is a less nice position to be in. The more the price moves, the more wrong your position is. In fact, the more it runs away, the more you have to push it the wrong way, because you'll be selling into a falling market or buying a rising one.
6. Natural question is, why would anyone ever be short gamma then? Well, you're going to charge people for their gamma, and if things turn out right you'll have charged them more than it's worth in terms of potential movement that they can buy low and sell high on.
7. So what happens with squeezes? Well sometimes that gamma gets so out of hand, there's some player with a huge wrong position squirming to flatten his position, but he is pushing it further and further away with each trade. Note that near expiry, a lot of things get hairy. You can imagine the gamma gets really spiky near the strike, because the time value is coming out of the price, and so you have a discontinuity at that corner. Also with binaries, knock-outs, and other simple non-vanilla options, you get some weirdness that can momentarily cause some large sensitivities. I've seen more than a few days where someone is defending some level and the price action looks odd.
8. Isn't the pressure the opposite for someone else somewhere? Technically yes, but often incentives are not the same for different agents. Different players have different risk tolerances, and they have different business models. For instance if someone has taken a load of options and shoved them into a structured product, maybe they're not quite as worried about hedging it as someone who doesn't have that customer flow. Or someone just has some cash to take a punt on the option, and they don't care if it expires worthles. Remember it's not just the two people who traded the option, there's the market in the underlying as well. Interestingly that means you can trade an option with me and we can both make money.
1. Options have asymmetric payoff. One way you win, the other way you don't lose, assuming you are the buyer. The opposite is true for the seller.
2. Because of the asymmetry, movement is positive for the owner. So the option is priced based on how much movement is expected. There's a distinction here between vega and gamma, which are both movement related. Vega is sensitivity to the appraised future volatility, which means the option has more of that if it's long dated. Gamma is (2nd deriv) sensitivity to movement in the current price, so you'll have more of that if you're near the expiry.
3. As an options market maker, you are not in it to guess which way the underlying stock goes, you are in it to gain an edge in pricing that movement risk. So you hedge it when you trade an option with someone. If you buy a call, you sell some underlying so that your delta is flat. (Delta is just the 1st order price sensitivity)
4. If you flattened your delta, you can imagine the gamma making your local PnL vs price a parabola. If you are long the option you'll make money when the price moves either way, and you'll be needing to buy low and sell high to get back to flat delta. (Incidentally this is also why we don't really care whether it was literally a put or a call, as in the right to buy or sell. The 2nd order is the same, so has the same effect on your PnL when the price moves.)
5. The opposite is a less nice position to be in. The more the price moves, the more wrong your position is. In fact, the more it runs away, the more you have to push it the wrong way, because you'll be selling into a falling market or buying a rising one.
6. Natural question is, why would anyone ever be short gamma then? Well, you're going to charge people for their gamma, and if things turn out right you'll have charged them more than it's worth in terms of potential movement that they can buy low and sell high on.
7. So what happens with squeezes? Well sometimes that gamma gets so out of hand, there's some player with a huge wrong position squirming to flatten his position, but he is pushing it further and further away with each trade. Note that near expiry, a lot of things get hairy. You can imagine the gamma gets really spiky near the strike, because the time value is coming out of the price, and so you have a discontinuity at that corner. Also with binaries, knock-outs, and other simple non-vanilla options, you get some weirdness that can momentarily cause some large sensitivities. I've seen more than a few days where someone is defending some level and the price action looks odd.
8. Isn't the pressure the opposite for someone else somewhere? Technically yes, but often incentives are not the same for different agents. Different players have different risk tolerances, and they have different business models. For instance if someone has taken a load of options and shoved them into a structured product, maybe they're not quite as worried about hedging it as someone who doesn't have that customer flow. Or someone just has some cash to take a punt on the option, and they don't care if it expires worthles. Remember it's not just the two people who traded the option, there's the market in the underlying as well. Interestingly that means you can trade an option with me and we can both make money.