Hi, I'm having difficulty reconciling long dated vol of ETFs vs its benchmark index. For example Dec13/14 SPY vol seems systematically cheaper than SPX vol by around 30bps. Is it possible the screens are just biased low on the etf since the smaller contract size makes it easier to 'manipulate' or am I missing the consequences of a real effect here? Since SPY goes ex the day before expiry, the forwards should indeed be in line vs SPX, so I don't think its just the case that I'm not comparing the relevant options with one another. Moreover the effect seems present in both inverted and normalized term structures, so I also don't think its simply the case that the volatility at the shorter stopping time of the SPY (American vs SPX European) is simply lower. The effect is noticeable on SPY, IWM and DIA, but not on QQQ, so its very likely a dividend effect thats causing the discrepancy, but I can't put my finger on exactly what it is. Any light that could be shed on this topic would be much appreciated. Thanks
30bp is insignificant on 2-3y durations. Are the steps discrete, or are you seeing a hump Dec13? It's not easy to directly compare as SPX is very wide out there, even though it's 10-20x the notional in OI. Microstructure alone could account for it, and the SPX ATM combo is probably 8-10 wide.
Thanks Atticus. I'm not entirely sure I agree that 30bps is insignficant. Its about $4 on the Dec13 straddle, which is almost the entire bid/ask spread the floor would quote. I suppose it depends on your application - if you're taking an outright vol position, sure its nothing, but if one were market making SPY and hedging with SPX (for some reason), you could be paying away your entire edge. Further, while I agree that SPX is very wide and its difficult to compare, my experience is that lately Dec13 has been better bid on SPX and I'm pretty sure I could sell SPX (my interest, so not just getting lifted) at a vol level basically comensurate with where I see the equivalent SPY offer. Its not just a hump at Dec13 - I see Dec12 SPY slightly better offered than SPX, with this trend increasing at Dec13 and further at Dec14 (though I don't think Dec14 screens for either SPX or SPY are very reliable at this point). This increasing nature further makes me think the issue could be dividend/forward related, but I'm at a bit of a loss as to what exactly I'm missing here.
From 24.85% to 25.15% (guess, haven't looked at the vol-line there) is 4.00 on the Dec13 straddle? I'd guess it's more like 2.00. I'll take a look.
Back of the envelope calculation: ATFM vega ~ 40bps x sqrt(T) so 0.4% x 1.414 x $1250 x 0.3vol x 2 (put and call) = $4.2
Bloomberg shows the vols are 12bp off. Anyway, it's academic as you're not going to arb 12bp over that duration and beat bill-rates. I imagine that financing would result in a modest -return.
Agree - I'm certainly not proposing there is an arb here, nor would it be sensible to engage in it if so. Rather, my interest in this ostensible vol mismatch stems from extrapolating it out further in time and considering if using SPX vol to price say a 7y SPY option is the correct thing to do. Back of the envelope calculation given ATFM vega ~ 40bps x sqrt(T): 0.4% x 1.414 x 1240 x 0.3 vol x 2 (put and call) = 4.21
A bit of it is balance sheet costs, actually. If you look, Qs and IWM vols are a bit higher then their respective indices.