Call credit spread set-up

W

whitemills

I've been working on a relatively simple screen/set-up for generating good call credit spread candidates. Based on a relatively simple set of parameters (avoiding curve fitting as much as possible) the distribution of returns for the system vs. the s&p (courtesy of portfolio123) is shown below. The average hold time reflected in this simulation is about 10 trading sessions/13 days.

Thoughts/criticism?
 

Attachments

Sorry - wasn't trying to get into too much detail and I agree that there really isn't much meat except for the graph... but i was trying to zero in on any high-level feedback there. Specifically, if you could reliably generate signals/candidates for a 10 day holding period that would give you that outcome of price movements, would you choose a call credit spread strategy (as the distribution would imply that volatility on the upside may be overpriced), or does another idea leap off the page at you?
 
Why limit it to calls if you're arguing for CLT? Why the bias? If calls are overpriced then you're arguing for an upside skew, which is most often not the case (see index skews).

I get it if you're arguing for a -correlation on share price and volatility, but you would still want to go long the natural condor in lieu of the outright short deltas in the call spread.
 
Not necessarily saying that calls are the answer.... that is my more fundamental question. I think the distribution I posted shows a pretty clear skew in the return distribution - more so that I usually see priced into the option IVs in my (albeit limited) experience.

BTW - the distribution posted displays the weekly returns of about 250 trades over a 2 year period.

My intuition based on this is to sell slightly OTM or ATM call spreads (based on the IV structure/strikes of the specific stock), but I'm wondering if there is a more efficient/superior way to play such an edge.
 
Quote from whitemills:

Not necessarily saying that calls are the answer.... that is my more fundamental question. I think the distribution I posted shows a pretty clear skew in the return distribution - more so that I usually see priced into the option IVs in my (albeit limited) experience.

BTW - the distribution posted displays the weekly returns of about 250 trades over a 2 year period.

My intuition based on this is to sell slightly OTM or ATM call spreads (based on the IV structure/strikes of the specific stock), but I'm wondering if there is a more efficient/superior way to play such an edge.
More...

You model correlates vol and price (on the upside). IOW, you win on vega and lose on gamma, but you're drawing the wrong conclusion.
 
Quote from whitemills:

So you'd prefer a condor strategy on this to stay closer to delta neutral?
More...

You trade index as the delta adjusted for symmetry will have you short initial deltas due to BSM. The SPY July 128/30/36/38 condor is short a couple deltas under BSM, but you're long due to symmetry and volcorr. You're short more $prem in the put spread than in the call spread. Simply, you're long to 133.70 due to smile. This excludes any (-)volcorr on a rally in SPY. Further, there is a sticky-delta effect on strike-vol. OTM vols converge to ATM, assuming some limited movement in strip vols (like the VIX calculation, a stable strip). Upside vols will tend to fatten due to stickiness, reducing the speed of gamma, MTM. It's self-correcting under a single sigma, but again, strips will decline.

IOW, sell the above condor (example), even if you expect a move to 134. Simply go neutral one strike above what BSM would suggest. You can solve for the strike-symmetry (prem-based neutrality), but obviously we need price, vol, tenor, etc.

Net-net you want to be long index price based upon prem-symmetry rather than BSM delta-neutral (if biased to CLT).
 
You must log in or register to reply here.
Back
Top