The weird and wonderful world of options markets and models can sometimes provide useful insights on a reflexive/contrarian basis if we know where to look. Everyone is used to reading/hearing about VIX (Pisani's Fear Index) which tracks a near-the-money relatively short-dated implied volatility (note upside and downside volatility not just downside - though volatility and price do tend to co-depend quite highly). There are many other 'implied' distributional measures one can glean from the broad array of liquid options prices.
By comparing the normal (or log-normal) assumptions from simple options models with traded market prices for options, sophisticated investors can get a grasp for how different the market's view of the forward-looking distribution of prices looks relative to a normal distribution.
Two of these are Skewness - how biased to the left (downside) or right (upside) the distribution is expected to be and Kurtosis - how fat the tails of the expected distribution are (how much more likely extreme non-normal moves are).
Short- and long-term skewness implied by the options markets has begun to drop (rise in the chart) which implies a skew more to the right and less to the left - or a lower chance of downside risk. This shift reflects options participants reducing hedges for downside.
Long-term kurtosis has dropped notably (as the risk of extreme moves or fat-tails shifted from a longer-term to shorter-term stressor. While short-term kurtosis remains relatively high (fat-tails are still a concern), it has started to drop very recently as investors pull back on their hgedges of extreme moves.
The actual skew in options implied volatility has also dropped as we crashed in August - makes perfect sense as those that were betting on a nasty downside surprise or hedging, have taken profits or reduced their hedge as a degree of comfort comes back into the market.
While VIX has dropped, implied correlation (another options implied measure) remains very elevated. The implied correlation measures the difference between what the index options imply and what the aggregation of the underlying components of the index imply - if correlation was 1 between every pair of names in the S&P 500 then S&P index implied volatility would be the same as a weighted average of all the underlying options implied vols. We know that correlations rise in crash scenarios considerably more than in rising scenarios and so tracking implied correlation can offer insight into how professionals view crash risk.
This measure is extremely hard for any retail or non-hedge-fun/non-market-making desk to trade due to the number of tickets and simultaneity of the trades required (as well as the hedging program) and so when this moves, we believe it reflects professional opinion in a non-contrarian manner (i.e. retail is not involved).
As VIX has dropped from crisis levels recently, Implied Correlation has continued to rise as professionals adjust their prices for index options vs underlying single-name options to reflect this systemic risk concern.
The bottom line is that options skews have dropped (the cost of protecting significant downside has reduced relative to upside: e.g. 90%/110% SPY skew has dropped from 1.8x to 1.5x in the last 2 months), and long and short-term kurtosis (fat-tail concerns - high risk of extreme moves) has dropped recently, distributional skewness (downside tails relative to upside tails) is rising - indicating less concern over downside tail moves.
Put all of these together and it is clear that investors have swung from very anxious to more neutral and add the non-contrarian implied correlation signaling concerns over crash risk (or a high correlation event) and we are concerned.
When all of these indications are at extremes, there is little chance of an extended downside move since broad swathes of investors are hedged and hence not feeling all the pain - however, with current levels having normalized modestly, any downside shock (no QE3 for example) could easily be exaggerated by unhedged forced selling. It is somewhat contrarian but in the same way as when everyone is short we can get a short squeeze, when everyone is hedged a market will not drop as aggressively - our indications show investors are less hedged (even though Pisani will tell us the Fear Index remains elevated) and downside risk potential is high.