OK, let's see what happens to volatility in back month/front month. My first guess would be that front and back would adjust to around 40%, which would not be good for the trade as front month would lose more, but let's see. I was looking at put backratio because I have a bearish bias. Problem with those is that they will suffer from IV drop. Maybe back ratio + RC...
you're right. i'm recalculating this with my other black scholes excel model and the opposite spread would work better as you suggest. this TOS modeler must be flawed. at 40% IV I would need a move to either 48 or 63 to breakeven. On my model it would take a drop to 20% IV to breakeven at 55. What explains TOS's flawed IV modeler?
I don't know why TOS is flawed but this is very frustrating. Maybe someone here knows. My guess is that they use Vega for their calculation but just before earnings this does not work. I don't know. Tried to start a thread about it here but no one responded. Maybe you should send TOS a message with your calculations and theirs and ask them the question. They are usually REALLY good at answering.
You know, I'm leaving my straddle spread on .. since it is in the red equivilent to what I calculated would be my worst loss with the expected drop in IV. But this all boils down to a fundamental understanding of vega.. The 55 April strike has a vega of .0877 and 55 Mar is at .0566. If I am short Apr, then I am net short .03111 vega per spread. So with IV alone considered (i am not factoring in theta here), I should get a .62 profit (20 x .0311) per spread with a drop in IV of 20% in each month (to avg 40%) if price does not change... correct? perhaps my options calc is wrong. If I get an inequitable drop of vega per month, ie 20% drop front month, 15% drop rear month, here we go: (-20 * .0566) + (-15 * -.0877) = -1.132 + 1.3155 = net loss of .00045 * 100 per spread. (near breakeven)
I agree with you, except that I don't think Vega is accurate just before earnings. Remember: Vega does not know that today is earnings. Yes, vega is always higher for back months because premium is higher. But just before earnings, there is a lot of "demand" for front month options, so premiums are inflated on those. And even if vega is low, IV usually gets crushed on those faster than on back months. This is what I think, from what I have seen so far. Would be great if someone could confirm this.
Which is exactly the reasoning I intuitively came to when i started the thread shorting bidu front month straddles and attempting to hedge price movement with rear month ... When overviewing these pricing models, its becoming clear that the relationship between front month and rear month vols and expected inequitible IV collapse is already priced in to the options markets. I used another options calc and entered a starting IV of 60% (avg between the two months), and collapsed it 20%. The result looked the same as TOS, so its no bug. Then I went to my excel black scholes plugin and started the two months at same IV, and got the same result. So I imagine TOS is modeling an equitable drop in IV amongst front and back months, and that is where precisely the problem lies. It doesn't model for IVs to normalize across the calendar, as they actually do. So short rear straddle, long front, is essentially a way to be long a straddle without paying as much for IV drop. Of course you pay dearly as your reward is capped as well. (this strategy) .... hmmm... makes you realize you have to be careful using these options modelers since they normalize changes in IV.... I'll upload my excel plugin. It'll give precise results.
On TOS, you can change IV as you wish for each month. Click on the wrench at the bottom of the screen, where you see the positions, then click on "More".
that works well, actually. (the more and tool button) i think this will be slightly profitable. did it on 5 spreads... happy i stayed neutral in regards to direction.