3. Buying and selling volatility as a position

Derivatives 7: Volatility as an asset

So far volatility has always shown up as a side effect: something that happens to your option while you were paying attention to something else. This module changes the point of view. Volatility can be bought and sold just as a share is bought and sold, with its own thesis, its own horizon and its own way of failing.

Buying a call is two bets, not one

A stock trades at 100 and you buy an at-the-money call at 30 days with an implied of 60, paying around 6.9. The next day it reports earnings: the stock rises 4% and the implied collapses from 60 to 35, as tends to happen when uncertainty is resolved. The option is now worth about 6.3. You got the direction right and you lost money.

The breakdown explains why. Delta was around 0.5, so the four points of rise contributed close to 2 in your favor. Vega was about 0.11 per point of implied, and twenty-five points of decline took away around 2.8. Gamma gave part of it back and the day that passed took away another part. The net is a loss of some 0.6, around 8% of what you put in, with the underlying rising.

Now compare with the alternative scenario: if the implied had stayed at 60, that same option would be worth around 9, 30% more than you paid. The difference between making 30% and losing 8% does not lie in direction, which you got right in both cases. It lies in volatility, on which you had probably formed no opinion at all.

What it means to be long or short volatility

Being long volatility means holding a position that gains if volatility rises, whether because the implied rises or because the underlying moves more than was priced in. The long straddle, that is, buying a call and a put of the same strike and expiry at the same time, is the cleanest form: it is born with a delta close to zero and only amplitude matters to it.

Being short volatility is the opposite: selling those same structures and being paid for nothing happening. Level 6 already described its payoff profile and why it kills. What is new here is that it is a thesis, not an accident: whoever sells a straddle is asserting that the market is overpaying for movement, and that assertion can be right or wrong like any other.

Vega and gamma are two different ways of being long

This is the distinction that separates someone who understands the matter from someone who parrots it:

• Vega pays you when the market's opinion about the future changes, even if the underlying does not move a cent. It is exposure to the price of volatility.
• Gamma pays you when the underlying actually moves, and it is only collected by readjusting the hedge. It is exposure to realized volatility.
• Long expiries are almost all vega and little gamma. Short ones are the other way round: a one-week straddle is a gamma machine with an almost irrelevant vega.

The consequence is harsh and few people see it coming: you can be right that the next four weeks will be wildly volatile and lose money because you bought a two-year expiry, since realized movement is collected through gamma and there is none there. And the reverse: you can be right that the implied is going to rise and make nothing with a weekly expiry that dies before the rise arrives.

How the two bets are separated

• Neutralize the delta. You buy the option and sell the underlying in the proportion the delta indicates. What is left no longer depends on direction. But as the price moves the delta changes, which is what gamma measures, and it has to be readjusted. That readjustment is the mechanism by which movement is turned into money: it forces you to sell high and buy low mechanically.
• Start from a neutral structure. A straddle is born with a delta close to zero and saves you the first adjustment, though not the ones after it.
• Isolate the curve from module 1. Buying one expiry and selling another leaves a position whose outcome depends on how the term structure moves, not on the level of the underlying.
• And the reverse separation, which is the one most needed. If your thesis is purely directional, the clean instrument is the share or the future. Buying a call because you think something is going up is paying for insurance you never wanted to take out.

The price of being right with the wrong instrument

Being long volatility carries a daily bill. A one-month straddle on an asset at 100 with an implied of 20 costs of the order of 4.6, and in the first weeks it loses something like 1.5% of its value every day purely through the passage of time, considerably more in the last one. Buying volatility because it looks cheap without having decided how many weeks you are willing to pay that bill for is not a thesis: it is a subscription.

Real case: January 2021

The GameStop episode left the most extreme example of this lesson. During the last week of January 2021, the implieds of the short-dated options on that stock reached several hundred percent, when in a large, normal company the figure is around 25 or 30 points. On January 28 the stock touched 483 dollars intraday and closed that same session at around 190. By mid-February it was trading at around 40.

Anyone who bought calls in those days paying three-figure implieds was not only betting that the stock would rise. They were betting, at the same time and without knowing it, that it would keep moving with that violence until expiry. The first bet came good for a few hours and the second did not come good at all: when the implied returned to double digits over the following weeks, there were calls that lost value on sessions in which the stock rose.

None of this is a recommendation. It is education about what exactly is being bought when an option is bought.