[lit-ideas] Re: Rational decisions?

Simon Ward writes:

"Much
depends on how many decisions you're required to take. Given, say, twenty
decisions, it makes sense to court the risky options perhaps four or five
times, on the assumption that those risky decisions, if they work out, will
provide the greater rewards. For the remaining fifteen or sixteen decisions,
choose those options that present less risk for less reward.  On the other 
hand, if there's only one
decision to make, then choose the less risky option; you'd get less reward but
the chances of receiving no reward would be slim. 

In portfolio analysis, profit is a reward to risk, but risk needs to be evened
out across the portfolio."

In a way, what I intended to be questioning in my first post about this was the 
underpinning of portfolio analysis.  The idea that I can, today, rationally 
determine what course of action to take by assessing the risk/reward profile of 
various alternatives before me is only useful if there is a meaningful sense in 
which I can actually assess the risk part of that profile.

But when it comes to almost every real decision which managers make, the idea 
that the manager is in a position to assess the risk is at best a call to 
attentiveness and at worst a delusion.  

The future only happens once, and risk is only definable against the backdrop 
of situations that can happen repeatedly.  The reason that you can define 
finite odds for a particular outcome on the roulette wheel (and hence the risk 
of its not happening), even though at the time the wheel spins there are 
infinitely many possible futures, is that the definition of spinning the 
roulette wheel implicitly rules out all but a finite set of possible futures.  
If the ball falls out off the table, the spin doesn't count.  If the roof falls 
in on the roulette table, the spin doesn't count.  But it is the act of will -- 
tacit though it may be -- under which we all agree to ignore such disturbances 
to the game that makes it possible to say anything certain about what the odds 
are.

Any assessment of the risk/reward profile of a portfolio contains within it the 
tacit assumption that the conditions will persist which make it possible to 
define the risk for each asset in the portfolio.  Such assumptions often seem 
screamingly obvious -- until circumstances occur that violate them.  Very few 
people would have taken seriously the need to take into account the possibility 
that someone would fly two commercial jets into the World Trade Center before 
September 11.  Moreover, the risk assessment of the portfolios that contained 
Wall Street stocks on September 10, 2001 almost certainly would not have 
contained a factor discounting for what happened to stock values the next time 
the market opened after September 11.

The notion that rational decisions can be made by analysis of risk/reward 
profiles of asset portfolios differs from the notion that rational decisions 
can be made by considering the configuration of chicken entrails only in the 
elaboration of the processes that prepare for the consideration of the oracle 
and for the enactment of its advice.

Moreover, when, as is the case in the current financial meltdown, the assessors 
of risk are essentially claiming to be able to assess the odds of their own 
future behavior, put a price on those odds and then sell off the obligation to 
fulfill that forecast, the dangers of the delusion are made painfully concrete. 
 

Regards to all,
Eric Dean
Washington DC

Other related posts: