how to backtest this option idea?

cesfx · Sep 9, 2023

Sergio123 said:
I have a hypotheses that high volatility in the mornings ends up suppressing after 2pm-3pm. I think it's just animal spirits and that won't change.

Except on FOMC days when it is the opposite.

I think a mechanical strategy would work on most but then you would need to account for exceptions.

I think 2-3pm might be late to get an acceptable range and risk/reward for this type of 0dte neutral play. Lunchtime maybe.

earth_imperator · Sep 9, 2023

M.W. said:
After you pissed on the GBM model just a day ago? That's rich. Do you even realize that stock price return terminal (as in around option expiries) distributions are not log normally distributed most of the time? Complete nonsense to base back-tests on data that was derived through GBM

You are not qualified to judge about my work on GBM etc.
Stock prices are not lognormal, you say? What a dummy you are.
I'm doing statistics, not your idiotic outlier shit.

clockwork-dev · Sep 9, 2023

elite1974 said:
Simple option idea.

Write 1 0DTE SPX straddle closest to the money, at market open and then every 30 minutes thereafter until 30 minutes before market close. For a total of 13 straddles.

Then take assignment or expiration of all 13 straddles.

How do I see the daily results of how this performed over the last year?
Or how much to pay someone to generate this data correctly?

That'd be a cool test. I could probably backtest the value of the straddles only, e.g. writing an ATM straddle every 30 minutes. But not the assignment and I only have SPY data for 2022, PM me if interested though.

M.W. · Sep 9, 2023

Statistics @ community college? All your other posts on this or related topics show you have no idea what you are talking about. And you don't deserve to be corrected. Go on.

earth_imperator said:
You are not qualified to judge about my work on GBM etc.
Stock prices are not lognormal, you say? What a dummy you are.
I'm doing statistics, not your idiotic outlier shit.

earth_imperator · Sep 9, 2023

M.W. said:
Statistics @ community college? All your other posts on this or related topics show you have no idea what you are talking about. And you don't deserve to be corrected. Go on.

Up into my ignore list, silly woman! Bye bye!

kalapur73 · Sep 10, 2023

I think you can model this with MesoSim.
I will try to put this into a backtest once I get home.

Sergio123 · Sep 10, 2023

cesfx said:
I think 2-3pm might be late to get an acceptable range and risk/reward for this type of 0dte neutral play. Lunchtime maybe.

I agree. After 2pm would be the time to start covering.

The strategy is essentially SVOL and could be replicated by selling VIX calls on Tuesdays and selling SVXY puts on Fridays. Or you could setup an algo to sell a /VX contract every 30 minutes and cover sometime in the late afternoon.

If you do it this way then you won't have directional risk even though a rising VIX is more correlated with a falling SPX and vice versa, but there are exceptions.

I am not doing any of these strategies. I am focusing on oil and equities. The indexes and volatility is flat right now.

kalapur73 · Sep 11, 2023

elite1974 said:
Simple option idea.

Write 1 0DTE SPX straddle closest to the money, at market open and then every 30 minutes thereafter until 30 minutes before market close. For a total of 13 straddles.

Then take assignment or expiration of all 13 straddles.

How do I see the daily results of how this performed over the last year?
Or how much to pay someone to generate this data correctly?

I created this backtest in MesoSim, with one modification (12 vs 13 trades per day):
- ATM Straddle opened on SPX every 30 min, starting at 10AM
- 12 trades per day run into expiration
- Backtest period: 2022. 01. 01. - 2023. 05. 01.
- Fills: At Mid Price (no slippage)
- Commission: 1.5 per contract
- Initial allocation (cash): $100k

Result:
- CAGR: -1.18%
- MaxDD: -392%

I wouldn't trade this.. Not even with SL as the drawdowns are so huge.

You can check the results and the rules in detail here:
https://portal.deltaray.io/backtests/4f90f0a8-1ff0-4103-9021-645fb1d4cfc0

I'll make some expirements with ICs, maybe they perform better.

elite1974 · Sep 11, 2023

thanks for those results!

I wonder how much the 4% down day on May 18,2022 skewed the results.

also, it looks like the idea did quite bad in 2022 when the market was trending down, and quite good in 2023 when the market was chopping higher.

cesfx · Sep 11, 2023

kalapur73 said:
I created this backtest in MesoSim, with one modification (12 vs 13 trades per day):
- ATM Straddle opened on SPX every 30 min, starting at 10AM
- 12 trades per day run into expiration
- Backtest period: 2022. 01. 01. - 2023. 05. 01.
- Fills: At Mid Price (no slippage)
- Commission: 1.5 per contract
- Initial allocation (cash): $100k

Result:
- CAGR: -1.18%
- MaxDD: -392%

I wouldn't trade this.. Not even with SL as the drawdowns are so huge.

You can check the results and the rules in detail here:
https://portal.deltaray.io/backtests/4f90f0a8-1ff0-4103-9021-645fb1d4cfc0

I'll make some expirements with ICs, maybe they perform better.

Just took a look at mesosim website.

Did you run the backtest on the free basic plan or on one of the premium plan?