Hacker News new | ask | show | jobs
by TheTank 2305 days ago
Comments about events "priced in" are absolutely right. With a little work, you can make it transparent and use it to your advantage.

In simple terms, the price of an option at a given time is equal to the product of what you can expect to gain from it (its payout) and the likelihood of each payout (the implied distribution of the asset at maturity). The interesting consequence is that you can reverse-engineer option quotes to derive the "market-implied probability distribution of a given asset at expiry". You can then compare this to your expectations to enter positions (for example if you think the market overpriced/underpriced a given event, trade against it with options).

You first need to calculate the implied volatility of options quotes (both calls/puts on both bid/ask) which requires you to correctly adjust your forward, i.e divs and rates to obtain put-call parity. If your forward is wrong, your implied volatility curves will look off (for example put bids above call asks) which means you have the wrong rates or dividends expectations. Once you computed the implied volatilities and are happy with your forward, you can fit a curve between your 4 series (call ask, call bid, put ask, put bit). This is your implied volatility mark. You can then use this volatility mark to derive an implied probability density. There is a simple example of how this is done here:

https://www.mathworks.com/company/newsletters/articles/estim...

This is actually really useful when you are trying to manage your risk for a given event. It also has interesting dynamics. Back in 2014 for example, we were worried about our risk on PBR US (a massive petro company with strong political links) ahead of Brazilian elections. By using this method, we found out that the implied distribution of the stock was bimodal, each mode corresponding to one outcome of the election. This gave us an idea of how much the stock could move either way and helped us cover the risk.

If you would like to see whether a given event is indeed priced in as you would expect, you can use this method, bearing in mind there is a timing element and you should seek the option expiry just after the event or horizon you are considering.

One last point that is important to consider is that this is “market-implied distribution” and does not imply a future behavior for the asset in question. It merely gives you an idea of the expectations of actors at this moment. Moreover, it is highly dependent on your inputs (dividends, rates and how you fit your volatility curve between bid/ask options quotes, particularly on the wings).