Correct me if I'm wrong. While doing some math and live trades with long/short front/back month futures options, that was the way of my thinking and trading path: 1. Let's short naked strangles (yeah, want to cut both sides, and don't like directions in trading). Open 1 month DTE strangle, wait for its expiration, collected premium turned to realized PL, cool. Till the moment of sharp move, where both increased unrealized PL and margin requirements hit your account to death. 2. OK, let's add some protection. Don't like verticals in the same month, so try diagonals: short front month, long back month, then after some research long back-back month i.e. short Sep strangle, long Nov strangle. Important note: I choose strikes of different months only by 'same premium' basis (no 'same delta' or 'same vega' or 'same strike' rules). It's zero total PL at the initial, and the main goal to let front month expired wortless, and back month close with some profit after theta decay. This model plays well, history says it can make some profits on expirations. But... sharp moves happen inside the trading time window (1 month DTE, remember), and though they're not that hard as naked (protection works), still they make some pain of unrealized PL and margin. So why not to turn this on my side? 3. Let's reverse the logic of #2 and long front month and short back month strangles. Let's go closer to ATM, in order to grab volatile moves. Is there any edge? History says that there're moves in UL pretty often that would rocket your front month options while back ones grow less. Usually the difference is big enough and (tu-dum!) there's almost no margin requirements accorging to SPAN. I mean margin is mere $50-100 for this setups. There're some month of total standstill, and yes, I have to drop all long side and grab remains of short side, that would be net loss, but profits of other months with the moves are much higher. So any pros and cons of that approach? Any input would be appreciated.
Unless you have an edge on either/both leg, you have no edge on the spread. There's no magic alchemy that turns a collection of fair-valued options into a good bet.
Dael, without an edge you will get killed doing both strategies. I mean really killed. In equities you would simply get killed. In the commodity world I have to add the "really" in there because the forward curve is going to eat you alive from both ends. Good luck.
Deuteronomy_24_7, Maverick74 Thanks for the input. Yeah, I understand options price quoting is highly effective. But let's assume I'm dumb in options. I have EOD prices for commodities options saved manually from Barchart. I run trades on paper doing long front month strangle, short back month strangle. While there's no action in UL, front month expires completely worthless, but back month prices go down too, so I take not so big loss. But every time any signable move in any direction happen in UL, front month grows much harder then back month and generates profit due to difference in deltas and so difference in option price movements. Example: crude oil EOD price on 08/02 was 39.51, on 08/16 was 46.58. Positions (opened on 08/02 and closed on 08/16) and PL are: Long Oct16 32.50 P went 0.31 to 0.02 = -290 Long Oct16 48.50 C went 0.28 to 1.23 = +950 Short Nov16 30.00 P went 0.32 to 0.03 = +290 Short Nov16 54.00 C went 0.29 to 0.70 = -410 The key difference is bolded. Total +540 net profit. So where are the flaws? The picture seems to be too good to be true.
What you are profiting from is the front spread correlation. In other words, the front month on CL is the delivery month and therefore is more volatile then the back month. When you buy the front month straddle and sell the back month you are synthetically buying a CSO option (calendar spread option). CSO's are highly sensitive to correlation. When the correlation goes to one, vol goes to zero and you lose. When correlations drop below one, vol increases and that spread will profit. You have to consistently model the correlation accurately to profit long term from this. What you have had so far is purely blind luck.
Not too sound anal, but you probably mean covariance and not correlation, right? Cause in a spread, you can have correlation go to unity and the spread will still be volatile due to the difference in volatility of each component ( vol[A-B]^2 = volA^2 + volB^2 - 2*volA*volB*corrAB)
No. I'm referring to the fact that the underlying's spot price "is" the correlation. I'm not referring to the variance of the actual option but the fact that the option is a bet "on" the correlation. Then there is variance itself around the correlation. CSO's blow out when the correlations break away from one.
Underlying spot for the CSO is the spread, right? Volatility of the spread is the sum of the variances of the two legs minus twice the covariance, correct (*)? So, even if correlation goes to unity, the difference between the two volatilities will keep the spread volatility at non-zero. In fact, it's a fairly common situation where correlation increases, but so do volatilities of the two legs and in the end volatility of the spread actually increases (**). Anyway, the CSO argument is not right, IMHO, since if he was doing a straddle swap on two correlated assets which is in no way equivalent to buying or selling a spread option. In a spread option, you are long volatility on both legs and carry correlation exposure (***). * most probably, you want to look at spead vol in normal space, so that adjustment comes first ** it is especially possible in spreads where correlation is already pretty high *** I am not saying you are wrong in saying it was pure luck, just merely am bored and anal today
He is "not" trading the volatility of the options, he is holding into expiration. His "outcome" will be completely determined by the relationship "at expiration" of the two contracts. Let's say he bought a straddle on WTI front month at 49 and sold a straddle in WTI at 50 in second month. If at expiration, front month is 50 and back is 51, there is zero change in the spread. The prompt month spread was -1.00 when he put on his position and it's -1.00 at expiration. He will get lit up on that as would the CSO buy of the -1.00 straddle. On the other hand, if front month expires at 55 and back month at 52, then the prompt month spread went from -1.00 to +3.00. THAT will do very well on. The -1.00 straddle buyer is selling his CSO with spot at +3. Yes, I'm aware that vol can increase as corr increases and vice versa, but he is not trading vol. He is holding till expiration. Just one man's opinion of course.
Indeed, I missed that he's holding one the front leg to expiration. It's tricky, since he's exposed to the terminal distribution of the calendar spread, but then he's just as well exposed to the implied vol on the remaining leg. Ok, then... Dael, A. what is the view that you are trying to express with your strategy? B. if it is a persistent bias, why does it exist in the first place?