Comment on Calendar Spreads

Amey commented on 14 Dec 2018, 01:56 PM

Yes the play is with mean and SD. Actually I made a program which gives me a signal considering bid-ask and ask-bid like you mentioned. But when I want to unwind and look at ask-bid, it’s far from mean mostly, so even though the reversion to mean has happened on the basis of LTP, on the basis of unwinding of the trade at then ask-bid, it has not.

Consider this data for TCS (only as an example, not factual numbers):
For Dec-Jan contracts
Mean: 8
SD: 2.5

So when the spread b/w Jan and Dec goes below 5.5, I will buy Jan, sell Dec (considering 1SD for ease).

Now at the time of unwinding, I have to Sell Jan and Buy Dec, and based on LTP, the difference between Jan and Dec has now moved to mean and is at 8. But the spread of Bid for Jan and Ask for Dec is only 4.5, which means that if I unwind at market based on bid-ask, I’ll make a loss of 5.5-4.5=1 per lot instead of profit of 8-5.5=2.5 per lot.

This has been the problem with all the stocks including liquid stocks. I’m unable to get around this because of this execution problem. Now to unwind, I have no choice but to go for Limit orders, but that carries at risk when prices move fast. Only one of the contract might get executed and other may not. Even a small change in spread can make the profitable trade into a loss making one.

So I wanted to know how you go about it? Does this ask-big (unwinding) spread also normalize and revert to mean near expiry? DO you go with limit orders or market orders?

Appreciate your reply.

View the full comment thread »