Thursday, December 01, 2016

The Bankers Bubble or the Derivatives Bubble - A mathematical certainty?

A bubble usually develops when people artificially jack up the price of an asset based on the speculation that it will continue to be massively profitable in the foreseeable future. There is some truth to the basic notion, however the amount of interest and desperation in purchasing such an asset is far in excess of the amount of real growth the asset is able to support. 

An asset can be anything - it could be a unit of productivity, a physical object or a piece of paper that legal binds two parties to a transaction. Each asset has a "price" associated with it.

One peculiar financial asset is a "derivative". A "derivative" is a financial instrument that derives value from the rate of change of another asset - usually a stock. As the derivative is computed from financial data - like *all* mathematical operations - it is vulnerable to computational noise. If the underlying asset price itself is noisy then the derivative can end up being even more noisy. 

Generally speaking governments in the world heavily regulate trade in derivatives. One particular point of agreement is to avoid creating assets that are derivatives of derivatives (or Greeks as the financial people call them) - although you are allowed to use Greeks to model the behavior of the derivative. 

Given the way that noise pervades all transactions, it can be extremely challenging to estimate the true derivative of an asset price. 

Perhaps the most secular way to look at a bubble is to realize that speculation takes hold when the true growth rate of asset value cannot be efficiently computed. As the computational crisis continues speculation dominates pricing and we see mass psychology effects kick in. The bubble builds and then bursts. 

In more advanced modeling, historical data on asset value is fit to a model and the derivative is computed from this artificially smooth function. This sounds find in theory - but what happens in actual practice is that the excursions from the mean value of an asset can rise rapidly and this makes the fit (or the model) pretty worthless. A derivative of a worthless fit leave the field wide open to speculation. 

A bubble currently exists in the derivatives market. And major financial institutions are exposed to the effects of this bubble. 

It seems from the above logic that this is a mathematical certainty. 

Another weird thing that few seem to realize is that "secure"/"safe" government bonds and annuities are actually *implicitly based on derivatives*. The whole idea of a bond is based on the assumption that economic productivity will grow.  

You can think of this as taking a derivative based on a very long average where the randomness is somehow evened out. But this is the proverbial Golden Calf. There is no clarity on what the randomization timescale associated with market fluctuations is. We cannot claim that productivity fluctuations will necessarily average to zero over a long enough timescale and a clear trend will actually be visible to a trivial mathematical operator. 

Given the high exposure of major banking institutions to explicit and implicit derivatives - the implications of such a crash are likely to be catastrophic. 

Deregulation of banking sector (especially of the kind anticipated by Trump supporters) will allow financial institutions to shift the visible bad debt into less visible forms. While that might temporarily improve the optics - it will ultimately set the market up for an even bigger fall. 


Post a Comment

<< Home