For more than a month now I have enjoyed highly analytical and yet somehow also very intuitive analysis from Kris at Vtrender. As part of his post, Kris also publishes a link to FII data in futures and options. I used to maintain a similar sheet myself but stopped doing that once Kris started publishing his much more useful post. One thing that has always bothered me about FII option data however is that there is no separate data for puts and calls – both are clubbed together. Somehow I never looked at it carefully therefore, assuming that option data cannot be of much use because there is no way to find out how much of it is puts and how much is calls.
Those of you who have already read today’s post from Kris, would have noticed that he has not drawn much conclusion from today’s action – the prime reason being not much net addition in future OI and hardly any difference between yesterday’s and today’s close in Nifty. After reading his post I again looked carefully at his option data sheet and saw something today that was always there but I was intellectually too lazy to try making sense of it. Pay attention now because not only this is mathematical in nature it might make sense to only those who understand options already.
Given below is option data – buy and sell – of last 2 days only. Apologies for pasting picture of data. I wasn’t sure how to paste my sheet here without loosing formatting.
Data in black above is already in Kris’ sheet. Data in Red is calculated data once I tried to convert option buy and sell data into an average “Strike price” for that day’s buy and sell in options. Change in option OI is easier but what stands out there is clear difference between how change in OI in number of contracts markedly differs from change in OI in total value and that may have some additional insight to offer apart from calculated “theoretical strike price” at which FIIs were active during that day.
Look at 12 September data. It was a highly bearish day and managed to close higher at 4940 only after a late decent pullback. Option data suggests FIIs were buying options at average strike price of 4858 and selling at average strike price of 4898 – a difference of 40 points. On top of that one can also see that while 3000 contracts got added, the value of option OI went down. This probably suggests that on monday, FIIs were buying puts/ selling calls but overall bought back more puts written previously compared to the number of new calls written. If one is writing lesser number of calls at higher strike price compared to number of puts written at lower strike price, one can end up with a higher OI in number of contracts and yet overall higher OI in value. That means monday FIIs were trying to bring down the market and simultaneously trying to benefit by writing calls and covering puts. Clubbed with their net sell in futures and cash, this points to uniform bearish view among FIIs on 12th.
Look at 13 September data. Here Kris’ conclusion about lack of direction is clear from option buy/sell as average buy strike is 4925 and sell is 4933 – just 8 point difference over previous day’s 40. That means overall FIIs ended up cancelling whatever they did today. Probably sold puts in morning and covered calls, covered those puts in afternoon and wrote calls afresh. Confusing! However relatively “high number of contracts addition in OI + value in OI also being positive but lesser than yesterday” points to net addition of puts if one combines the average strike price (4925-33). Still not sure if those puts were written or bought. One can only look at net sell in cash and futures and guess this is more likely to be put buying (number of contracts added much higher compared to value of contracts added). So FIIs are still bearish on tuesday despite market closing not having changed much. Market now probably has a bigger chance of breaking 4900 going forward this week.
I am not sure if this analysis is correct but I invite my friends at this forum to put their heads to this problem and see if additional number crunching and analysis of option data does yield any useful insights! Who knows, we might be onto something useful here.
Regards
Chandrashekhar
