Option Premium and backtests
So these days, I see a lot of backtest reports going around on #fintwit and majority of these tweets are reporting strategy stats in absolute INR terms.
I have always been little confused on how to interpret these stats since margin requirements have changed over time.
- higher margin assumption tend to show lower drawdowns
- lower margin assumption tend to show higher returns
whats the right balance?
let’s go down the rabbit hole and crunch some data, we may not answer all or any questions but since we call ourselves quants, so why not !!
below are few things we will look at:
- change in lotsize
- change in margin requirement
- change in ATR
- change in 1st OTM strangle combined premium (snapshot at 9.20)
- change in 1st OTM strangle combined theta (snapshot at 9.20)
before we dig in:
- All the analysis is done for banknifty
- Date range is from May 2016 to June 2022.
Lotsize
no comments, below chart is all we need.
Margin Requirement per lot at ATM
I am not 100% sure if margins were around 40k before 2019, so if anybody has better data, please correct me.
but around Oct 2020, due to new sebi rules, leverage available was reduced in a phased manner.
ATR
below chart shows banknifty daily ATR, current ATR is around 600 odd points.
Combined premium for OTM 1 strangle
why are we looking at premiums, well the idea is to check
- are premiums reflecting the changes in margin requirements?
- are we (option sellers) getting paid for putting up extra margin?
For this, we essentially look at the future price at 9.20 am and fetch premiums for 1st OTM call and put.
Also since lotsize has changed over time, we would like to adjust the combined premium as per today’s lotsize (25) so that there is only one moving part which is margin requirement.
combined_premium_adj = (combined_prem * lotsize at the time) / 25
ok, so quarter wise box plot looks nice, it was the first visualisation I thought of and it does contains a lot of information, but something simple needs to be done.
line charts are great but a bit noisy, lets take a 12 week rolling median
pretty neat !!
y-scale seems to be creating a bit of hindrance. let’s breakup the chart by each dte.
notice the shift in premiums around Q2/Q3 2020 which happens to be also the starting phase of change in margin requirements.
average of second regime is roughly 2x of first regime for all the charts
Lot of people may also attribute these data points to vix which peaked in Mar 2020 but then vix has gone through its ups and downs in last 2 years, so it’s up to you how you want to interpret the data.
Combined Theta for OTM 1 strangle
Looking at theta is probably going to show the same picture as option premium over long time but since we can and it does sound a bit more cool, so why not.
Format stays the same:
For this, we essentially look at the future price at 9.20 am and fetch theta values for 1st OTM call and put.
Also since lotsize has changed over time, we would like to adjust the combined theta as per today’s lotsize (25) so that there is only one moving part which is margin requirement.
combined_theta_adj = (combined_theta * lotsize at the time) / 25
box plots
line charts
median line charts
I think, all charts are essentially showing a regime change post 2020 Q2 and I don’t think the second regime is entirely due to high vix during 2020. It seems to be compensating for higher margin requirements.
