Contagion, correlation, fragility, feedback loop, liquidity crashes, pricewatchers, cross market arbitrageur.  These are the terms that a new paper titled “Illiquidity Contagion and Liquidity Crashes” written by Cespa and Foucault use to describe what happened during the Flash Crash and how it could happen again.  The authors believe that the evaporation of liquidity was the cause of the Flash Crash and not the large e-mini S&P futures seller that many have blamed.  The authors propose that:

“Markets for different assets have become more interconnected as market makers in one asset increasingly rely on the information contained in the prices of other securities to set their quotes. The reason is that prices contain information and progress in information technology has considerably increased liquidity providers ability to access and process this information in real-time. Using a rational expectations model of trading, we show that this evolution can make securities markets more liquid but also more fragile. For instance, a small exogenous increase in the illiquidity of one asset can trigger, through a multiplier mechanism, a large drop in the liquidity of other assets.”

 Ever wonder why small cap stocks track the Euro?  Or why when one giant stock like Apple moves, an entire subset of stocks move instantaneously?  Gone are the days when stocks traded based on supply and demand of natural buyers and sellers.  Now, stocks move because some signal was given in another asset class that caused a chain reaction of events that occurred in milliseconds.  Good luck human.

The authors give this example of a feedback loop which could cause liquidity to suddenly evaporate without any apparent reason:

“Now suppose that a shock specific to security Y increases the cost of liquidity provision for dealers in this security (e.g., dealers’s risk appetite in security Y declines). Thus, security Y becomes less liquid and, for this reason, the price of security Y becomes less informative for dealers in security X. As a result, inventory risk for dealers in security X is higher and the cost of liquidity provision for these dealers increases as well. In this way, the drop in liquidity for security Y propagates to security X. This spillover makes the price of security X less informative for dealers in security Y , sparking a chain reaction amplifying the initial shock.”

The authors then note that if this “liquidity crash” were to “precede the arrival of a large sell order in one security then large price drops can happen in all securities. Thus, the model can generate a simultaneous drop in liquidity and prices of multiple securities, very much as in the Flash Crash.”

We highly recommend reading this paper as there is much more details regarding cross market arbitrageurs and why ETF’s can experience sudden meltdown’s in liquidity.  But the bottom line take away from this paper is this statement from the authors:

“This is consistent with our model in which the liquidity of a security drops when prices of other securities become less reliable as a source of information.”

Our markets have become one giant, correlated, fragile feedback loop which are susceptible to shocks at any time.  In the past, the shock absorbers were the market makers who could cushion these moves with their inventory.  Today, market makers hold no inventory and withdraw when there is a sign of trouble.  Economics 101 taught us when supply crossed demand then price was set.  Too bad that our markets have now rendered this equation useless.  Good luck human.