Gamma scalping:what volatility are you trading?

B

bologeorge

I've read some threads about gamma scalping, and I find those who familiar with this strategy believe that, when you long straddle and scalp the underlying, the low implied volatility is the key to profit. Because you are trading the realized volatility against the implied volatility. This is correct and obvious.

However, my question is what realized volatility are you actually trading? As I know, when you do gamma scalping you can rebalance the delta frequently, like in every 30 minutes or in every 0.5% of the price move. In that case you are trading the intraday realized volatility, and the implied volatility of the option is based on the time to expiration, which can be few days or few months.

In most time, the intraday realized volatility, which means you calculate the volatility on hourly or minute basis, is much higher than the daily realized volatility. For example, from the price of options we know the current implied annual volatility for SPY is 14%, but if you calculate the realized volatility on hourly samples, the annual volatility will be 30%!Does that mean we can have a huge theoretic edge to do the intradat gamma scalping?
 
Let me say first that I strongly suggest you to get a copy of "Volatility trading" by E. Sinclair where you can find answers for all these questions. However it is essential to spend a lot of time to get a good undesrstanding of the different forms of volatility.

I'm probably in the same level as you but I'll try to comment on the last paragraph but by no means take it as an expert opinion.

You don't know the realised volatility until maybe on the expiry date. You can estimate or forecast the realised volatility on the expiry date by looking at today's and past data. There are numerous ways to do this but obviously none of them is accurate. Looking at intraday data is a good approach but it is advised generally not to go below 5 mins to avoid microstructure effects (again check the book I mentioned). I don't think though that you can rely on only 1 day's data.

A quoted 1 month implied ATM volatility of 15% roughly means that the market believes that realised volatility in 1 month will be 15% (that's not 100% right but a good approximation). If today's realised volatility is 30% it doesn't mean that it is going to be 30% in 1 month's time, but there is a good chance that it will be (or at least greater than 15% which will make a profit for gamma scalping).For example because of autocorrelation in volatility or because there is a good chance that the market has misestimated how fast the volatility will revert to its mean.

Also don't forget that vanilla P&L is path dependent since your gamma is not always stable during the life of the option, same with volatility (unlike what the Black-Scholes would suggest)
 
Thank you kapw7. I will try to find the book you recommended.

Well, I definitely understand the difference between implied and realized volatility. My point is that, the realized volatility will be different if you measure it in different time scale. In most time the realized volatility on hourly basis is higher than that on daily basis. If the price movement is a pure random walk, they should be the same.
 
Sorry for repeating myself but again the book I mentioned has a detailed discussion on this. btw I'm not affiliated with the author or the publisher and there might be other books/sources that are better.

One thing to try is an XL sheet of a simulation of a lognormal random walk where you can define exactly the volatility value and then apply on this your methods of estimating volatility and see what results you will get
 
Quote from bologeorge:

Thank you kapw7. I will try to find the book you recommended.

Well, I definitely understand the difference between implied and realized volatility. My point is that, the realized volatility will be different if you measure it in different time scale. In most time the realized volatility on hourly basis is higher than that on daily basis. If the price movement is a pure random walk, they should be the same.
More...
not a math guy but if a 100 stock is up or down 1 at the end of the day,thats a 1% move,but if it trades from 99 to 101 on the opening and then back and forth using those as the high lows,the movement within that 2 point spread is going to be much larger than 1% so the volatility would be higher
 
couldn't find the curve for the ratio that volatility changes say from weekly to intraday..or from 4 hours to 1 tick.. and find a sweet spot to gamma scalp.. your saying that theres a hidden edge because implied volatility levels are calculated on the daily close and intra day realized is much higher then implied.. such that you could gamma scalp at a higher frequency then daily adjustments and arbitrage the difference in daily implied calculations to the actual intraday realized... i personally think that the true players are scalping options against options in this respect or any others... maybe across the term structure..
 
Quote from kapw7:

Let me say first that I strongly suggest you to get a copy of "Volatility trading" by E. Sinclair where you can find answers for all these questions.
More...

i have this book.. yet didn't think their was anymore then i had read in others.. obviously their is.. i'm gonna crack it wide open..
 
Quote from cdcaveman:

i have this book.. yet didn't think their was anymore then i had read in others.. obviously their is.. i'm gonna crack it wide open..
More...

"Volatility Trading" is a great resource. If you need a primer, check out Sinclair's "Option Trading," much better than Natenberg. I really enjoy the utility of his work; more directed to the practitioner than the academic.
 
You must log in or register to reply here.
Back
Top