How far can we defer the next asset price crisis? Depends on how fat the tail of the distribution is!

I have blogged on fundamental liquidity issues recently and one point that I want to bring out is that the greater the divergence between asset values and GDP and the greater the divergence between broad MS growth and GDP growth, especially in slower growth frames, the “fatter the tail of the distribution”.  

Volatility at one level is a measure of the sensitivity of an asset’s price to new information, shocks to the system/de facto changes in the energy of the system.   It reflects changes in demand flows for assets which can reflect changes in risk preferences and risk/return expectations.   In a general equilibrium volatility is meant to be a static physical characteristic reflecting the fundamental nature of the asset and its relationships, but we do not currently have general equilibrium relationships and volatility is not a stable measure of anything.

Essentially when we have excess asset focussed money supply growth (EAFMS) amidst a slowing growth frame the “accumulated liquidity in” decisions exceed the “present value of future liquidity out” (PVLO) decisions.  In a sense liquidity (at its heart a function of the relationship between asset allocation decisions and C/S/I/P decisions) becomes more sensitive to short term  changes in demand flows and risk/return expectations, risk preferences and other factors.   As the ratio of EAFMS to PVLO rises so does the natural volatility of the system.

Why the tail?   Why not volatility at 1 standard deviation?  During periods of excess monetary flows demand changes are not in totality covariance issues (ie. relative attractiveness of one asset to another) but absolute flows that suppress relative price reaction.   In other words we see a fall in volatility throughout most of the distribution.   All the while the system due to EAFMS/PVLO imbalances becomes more sensitive to changes in flows, preferences, expectations and shocks.  

Given that the system because of its imbalances becomes more sensitive to small changes in any one factor, the bigger the divergence noted in paragraph A the greater the probability of an extreme risk event.   The greater the accumulated liquidity in to PVLO the larger the tail: the risk event and its probability increase. 

In reality, from a given point on, we can effectively discount the rest of the distribution in any analysis as a dynamically widening tail is merely a statistical constraint on the way we should be viewing risk.  We are only exposed to the wider risk distribution if forces suppressing risk remain influential.  

Leave a Reply