Is it possible to backtest an gamma scalping strategy?

[Rudolf13100]

Quote from hlpsg:

You can bactest it in a software called OptionVue. You'll need the Backtrader module if you want to backtest.

It's data moves forward in 15 min blocks, but it might be the best option you have. It'll also include IV charts so that'll help in picking the vehicles you want to do it on.

Thanks for the reference. The Backtrader module is for the historical data?
Are you using option vue yourself? Is it a hard task to backtest a gamma scalping strategy on option vue (in regard to programing skills needed)?
Are there any other software can do it?

 

[Rudolf13100]

Quote from IV_Trader:

One can “improvise” scalping back testing without historical price action data (tick by tick). You need RV and historical IV.

How do you calculate the scalping gains then, without historical action price data?
Could you illustrate your statement by an example?

 

[cvds16]

I think I understand what you are trying to do, but I doubt is this is very meaningfull. are you going to scalp at fixed points or ... ?

 

[Rudolf13100]

Quote from cvds16:

I think I understand what you are trying to do, but I doubt is this is very meaningfull. are you going to scalp at fixed points or ... ?

I used to build trading systems on stocks, as a hobby. I have never found any interesting system so I stopped.
Recently, I started to get interested at options strategies and I read a thread about gamma scalping. I remembered that I had found a trading system on stocks which was working well as long as there was so strong trend. As soon as there was a strong trend the system was loosing.
When I read about gamma scalping I naturally asked myself: what would happen if I added the purchase of a straddle to that trading system?
That's why I was asking questions regarding the possibility to backtest a gamma scalping strategy.

Please let me know if I explain myself correctly (English not being my native language...)

 

[dmo]

Rudolf - read the attached article on gamma scalping then come back with any questions. This should clarify it for you.

 

[Rudolf13100]

Thanks a lot for the article. I will read it and be back with questions tomorrow.

 

[Rudolf13100]

The article you posted is very good. Still, I am confused about few things.

1) To make money gamma scalping you have to buy underpriced options.
Thus far, is my readings I saw that an option would be considered underpriced either if its IV was historically low or if its IV was lower than its HV.
Now, you seem to define underpriced options in a different maner: "[an option] trading at an implied volatility significantly below the true volatility"
What is the true volatility? How is it calculated? Is it the standard deviation of the daily closes over a specific time period?

2) "But if you can find options that are underpriced then scalping
gammas will allow you to profit from the discrepancy. It also will let you
profit from correctly guessing that true volatility will increase."
Now, is that assuming that we hold the expiration until expiration?
Also, from your statement should I understand that if IV<"true volatility" then it is given that "true volatility" will increase?

3) "It doesn’t even matter if the implied volatility never rises along
with the actual volatility."
Then, can i say that :the main source of gains would be an increase in actual volatility [please define actual volatility] and a potential, less important, source of gains would be an increase in IV?

4) In your article, why do you choose the purchase of a straddle with 45 days left to expiration? Why not a straddle with more time to expiration (to have less time decay to race with)?

5) What is the potential maximum loss in your position? Is it equal to the time decay (plus commissions)?

6) More generally speaking, how does the esperance of gains/losses of a gamma scalping strategy compare with a simple straddle purchase? What elements should be taken into consideration when choosing one strategy over the other?

7) "If the futures continue going higher, you’ll continue to make more money on your options than you’ll lose on your short futures."
Could you please explain this phenomena?

 

[cvds16]

rudolf, I understand your questions, but I am afraid, from reading them, that this is way over your head at this moment. Like someone allready said: gamma-scalping is not really for beginners in options. First you got to read up on options and how dynamic delta neutral hedging really works. Read Natenberg and Baird as a starter. Things will gradually become more clear. It's not possible to explain everything here quickly in a thread.
First it's important to make a clear disctintion between different sorts of volatility: historical, implied, realised ...
What is described as actual volatility: is the volatility of the underlying taking place, you could call it realised future volatility too. This is what you are trying to guess by gammascalping: you think implied vol in the options is too low and you think realised vol of the underlying will be higher.
Also there is no such thing as one implied vol for an uderlying: it differs across months and even across strikes, part of this is due to different timeframes, part of this is due to the reasoning behind statistical distribution of prices in the option model.
Furthermore you really need a good view on future volatility, which is an art in itself.

 

[hlpsg]

Quote from Rudolf13100:

Thanks for the reference. The Backtrader module is for the historical data?
Are you using option vue yourself? Is it a hard task to backtest a gamma scalping strategy on option vue (in regard to programing skills needed)?
Are there any other software can do it?

Yes the Backtrader is for the historical option quotes. It even gives the bid-ask spread at that time. It's not perfect by any means, but I think it's the best option for now.

I'm not using it now as I'm on TOS, but I do miss the Backtesting feature.

It's not hard to backtest. There's no programming needed. You just click on an instrument you want to backtest in, go into the option matrix, activate the backtrader module and type in the date you want to go to, then you just trade it like you would on your brokerage platform, buying and selling stuff, and looking at the P/L curve. You can click forwards and backwards in 15min increments.

I heard that Optionetics platinum also has a backtesting feature, but it's a web based program and is much slower, and I have never personally used it before, but it's worth checking out.

 

[Rudolf13100]

Hi cvds,

I think I got it now. Let me tell you what I understood, and please correct me if I am wrong.

We purchase a straddle at t=0. We want the IV to be "low" in order to spend "little" money.
From t>0, we want the realized volatility on the underlying to be "high", so we make a lot of scalping trades (the more scalping trades we make, the more money we make).
So want is meant in the article is that: if realised volatility (from t>0) is greater than IV (at t=0) we make money. Otherwise we loose money.

Did I understand right?