Has anyone on this forum tried vol arb with success? For example the apparel industry is showing high implied vol while footlocker has extremely low implied vol. Does this dispersion happen often? Any other examples of doing vol arb?
Sure. In 2000-02 I arbed a lot of inter-exchange crossed markets in COMS prior/after the PALM spinoff. PSE would quote a COMS ATM call at 3.25 bid while PHLX was 3.00 offered for the same. You could trade up to 20 on a ticket as retail. After a year or so the OCC locked any option that showed a crossed-mkt to give the MMers time to adjust. A month later the exchanges were all linked. I received a C&D letter stating they would disgorge the gains. IBKR's attorney handled it for me. I never heard from the OCC attorneys again. ~2006: betsfortraders.com was a retail-arm of Trinitas Capital. Trinitas traded from the buy-side in OTC lookbacks (short). The offered retail lookbacks on their retail front-end, but long-only, in an effort to cross some of their institutional trades at edge. That would not be legal in the US. Their retail end was offering Double No Touch markets on US stocks in addition to the long lookbacks. They were making a market in WMT with the shares at $50, and the 1M 45/55 DNT was priced at 32/100. $32 pays $100 if WMT never touched $45 or $55/share over 30 days. The DNT was worth 65/100. They had $10K limits, but they had 15 os so issues, all had this absurd pricing. ~2007: Oanda was quoting "box options" which were a variety of fwd-start digitals and touches. They changed models and began quoting touches as digitals (50% off sale!) and we destroyed them. I posted it to Oanda's forums (and here) and a day later Oanda stopped quoting all options. They stated that they were concerned about "future regulatory issues." Edge in listed equity arbs is a function of rates. The lower the rates the tighter the markets. You can make an explicit bet on rates as well, but it's usually far less costly to simply spread Treasuries. I've done a lot of rev/conv; boxes, etc., but got a portion of a box unwound by IBKR a while back, so I would only do rates stuff on a platform with discretionary risk (human intervention). Most of what I've wrote is moot as it was OTC and not rates. Unlikely that those arb opportunities will occur again at retail. There is a thread on boxes running in the forum.
Dispersion is done with baskets, but if you're going to do pairs, then it's best to always be long the share vol ("street") and short the index.
If I look at WMT for example, their beta is .5 but there 1M ATM vol is higher then the SPX. I understand that there is company risk which is why the vol is higher. But it seems to me I would rather be short a basket of low Beta stocks and long the SPX vol. Could you elaborate on why I would do it the opposite? (short SPX, long equity). hahaha thats insane!! why didnt you just suck oanda dry?!!!!
have you seen opportunity in todays markets, where the auto-makers index 3 month implied vol is trading near all time highs but F or GM 3 month vol is trading near lows, or Crude Oil Term structure shows High IV in August but the 6 month vol on XLE is quoting normal levels?
Index vol is generally overpriced because of regulatory pressure, structured products and as a general use of hedging. Single stocks do not have the same issues and is considered as generally under-priced compared to its blow up risks. Dispersion is usually structured as vega-notional equivalent (which takes into account beta) or dollar-notional equivalent. I agree with @destriero it is better to do in baskets. If it is single stock, I would rather do direct pairs trading
Beta is corr*(share SD/index SD). A step removed from simply trading IV against HV. Beta isn’t going to be hedged through basket/diversification (corr). I realize that you’re talking trading a single name vs. index, but I wouldn’t use beta as a input. You can structure the index gamma in short puts with long puts on street vol but there are a lot of moving parts. I used to trade a “dirty” form of dispersion where I would short -10D index combos for the slight skew; long street combos at -10D and cover the street combos with gains and redeploy in a DN combo in the same ticker.
[QUOTE="I used to trade a “dirty” form of dispersion where I would short -10D index combos for the slight skew; long street combos at -10D and cover the street combos with gains and redeploy in a DN combo in the same ticker.[/QUOTE] What is the philosophy by redeploying the money into a delta neutral position? I know you like to trade with time to expiration < 30. Wouldn't the 10 D options have super negative charm? (Would think you would rarely make money on those meaning rarely deploying the capital into the DN strategy) Can you explain the reasoning for doing 10D
Booking delta gains as a implicit rebalancing. Implicit as you are looking to cover in lieu of discrete hedging those street gains. A straddle (local:10D) isn’t going to have measurable charm. Charm is only an issue in deep OTM. Say that you’re doing an INDU dispersion and you have three street names in your basket up an average of 20%. You realize gains in those three and rebalance ATM in those three, assuming you’re flat against your index delta. A member of this forum no longer posting (IV_Trader) implemented this method for years. As I recall he did really well with it.