The Market is Long or Short Vol? Not So Fast....

There is an inherent need for market participants to ascribe cause and effect with respect to asset price movement. “Bonds rose on the Fed’s dovish tilt”. ”Stocks fell as trade war fears intensified”. When repeated often enough, these narratives can serve as an important shorthand to describe market psychology. If we believe trade or QE or stock/bond correlation or the latest tweet on North Korea matter for risk, then (in self-fulfilling fashion) they can matter. The same might be said of psychological, technical levels on macro assets like rates, the dollar and the SPX.

But we should be mindful that our need for markets to make sense leads to inherent narrative fallacy and confirmation bias. We are naturally inclined to explain why things happened in markets. With this in mind, we should be skeptical of calculations of the “reaction function” of investors that results from financial products and trading strategies. For example, consider a graph of the market’s purported SPX gamma profile by spot. We are often told, ex-post, that the market’s volatility (or lack of it) was the result of short or long gamma positioning. The precision implied by this analysis is simply inconsistent with the reality that there are far too many unknowns to be anywhere close to precise. To reasonably quantify this exposure would take solid estimates of: 1) the amount of volatility actually bought or sold 2) the amount of this vol that is being delta-hedged as vol and the manner in which it is traded. 

In the aggregate, we know little about either of these. There is a laundry list of unknowns that I don’t see anyone trying to solve for.  Here are 10 to start. 

1. Bought Versus Sold. Can we measure and properly classify the trades as buys versus sells? Are put spreads and put “stupids” classified as the same trade? How about combos versus straddles? Adding to this, reported open interest is considered unreliable.
2. Underlyings. Most of the analysis focuses on the SPX. While this is surely the largest option complex, there are other important underlyings that leave investors/dealers with meaningful exposures including listed options traded on the SX5E, NKY, Kospi. These markets are correlated to the SPX on both a spot and greek basis.
3. OTC Listed Look-a-likes.  While some of these are reported to the SDR, the lion’s share are not. 
4. Long dated OTC trades that have aged. Remember that giant 5 year collar done with a pension fund 4 and a half years ago? It might now be live with regard to gamma exposure.
5. The complex of variance swaps (and the vast number of permutations) along with light exotics such as down and out puts and rate contingent puts. These are all significant products with meaningful greeks. No one has any real idea how much is out there and where the trades are struck.
6. Listed VIX Options and Volatility ETPs.  There is a huge market here that often goes unrecognized in the analysis of the market’s gamma profile. Even when included, there is no effort to distinguish between a VIX 1x2 call spread or a VIX 2x1 call spread, both of which trade in the market.
7. Cross Asset Vol Exposures. We must appreciate that events that originate outside of the SPX can cause movement in the SPX. The spike in credit spreads and credit vol at the end of 2015 are an example. The market’s long or short exposure to vol in related asset classes matters.
8. The huge structured products market that is the cornerstone of vol, skew and correlation exposure in dealer books across Europe and Asia. The market’s “vanna” profile that results from structures like auto-callables is substantial. It is left out of the analysis.
9. Relative value trades. If the Street facilitated a long SX5E / short SPX vol trade for a client, how should we think about the vol exposure that results and the hedging strategy pursued?  
10. The exceedingly large and growing universe of alternative risk premia products. These days, one can flip a switch to pay or receive carry in various formats (variance risk premium, currency risk premium, VIX systematic strategies). Add to this the substantial amount of capital in “short vol lite” strategies like iron condors.

All of this is to say that while we certainly shouldn’t ignore the data that is available, we should appreciate how many aspects of positioning, exposures and hedging protocols we simply cannot see or calculate.