Re: Uncertainty and Liquidity Preference

[chip poirot <csp@xxxxxxx>]

Ted Winslow wrote:

> Why is holding money a rational way of dealing with the fact that, in the
> main, the long run yields of capital assets can't be known so the
> prospective yield of such assets is "fundamentally uncertain", i.e.
> unknowable?

Are you distinguishing between holding long term financial assets and cash, or
between physical capital? My understanding is that Keynes said that the
expected rate of return on additional capital investments was fundamentally
uncertain. Also, it is my impression (I am sure others can correct or confirm
my impression) that it is not always uncertain-just sometimes so. Thus, there
are periods when rates of return might be fairly predictable and people have
reasonable expectations of getting the rates of return they want (though
exogenous events could always derail these plans). On other occasions, the
state of the economy may make the realization of any rate of return on physical
investment seem plausible.

On the other hand, if you are talking about investment in financial assets as
opposed to cash, the long run rate of return has been fairly stable-stocks have
returned on average 11% per year over the long run since 1929. Financial
managers say an 80/20 split between equities and debt is an optimal personal
financial strategy that carries little additional risk. It seems the danger of
real loss comes from investing in particular companies or specific, short run
market conditions. A sudden downturn in the market can destroy my position,
requiring a move to cash-that I may not have.

>
>
> Uncertainty of this kind isn't capable of degree, is it?  It can't increase
> and decrease.  "We simply do not know."  Consequently, variations in it
> can't be invoked to explain variations in liquidity preference.

It seems that at the very least subjective personal (and collective)
assessments of the expected rate of return on real and financial assets will
change. As i recall from econometrics, there are ways of using Bayesian
approaches to econometrics to specify  this -but that is somewhat beyond me.

It certainly seems possible that at any time t, investors have a subjective
probability estimate (Ps)=E[Ds] of default. There also exists a real
probability of default, call it (Pr)=E[Dr]. At any point in time, investors
based their expectations of default on information. If the information is
assymetrically distributed, then E[Ds] departs from E[Dr]. However, investor
behavior changes the values of the variables determining the real risk of
default. Thus, investors expectations can depart farther and farther from the
reality.

I think we saw something very much like this with Long Term Capital Management.
They were relying on long term trends over a period of many years to develop
their positions. However, they did not take into account that investors could
fundamentally misunderstand the nature of the real probability of default, or
that sudden changes could destroy their position. In early 1998, Stanley
Fischer, deputy director of the IMF, was giving the all clear signal on buying
Russian bonds. Investors believed that Russia was stable and that the IMF would
bail them out. Insiders in the Russian economy knew how real the danger was.
However, investor behaviour itself fundamentally altered the level of risk in
the Russian economy.

So, all this is a very long winded way of saying that yes-real and perceived
levels of risk and uncertainty can change dramatically, and investor behavior
can itself change the nature of the environment.