Re: flow or stock?

["Dr. Bruce McFarling" <ecbm@xxxxxxxxxxxxxxxxxxx>]

> Just to push this point (which, like Christian, I must confess to be

> less than sure of), I am intrigued by your use of the word
> actuarially" in this passage.  Would you accept that an actuary
> is able to make statistically reliable projections of the cash
> flows of an insurance company going out into the future at least
> as long as the average term of a life assurance policy?

There are two parts to the answer ... the first is, to a
certain extent they can, since the average term of a life
assurance policy is not that bad in terms of the basic
demographics ... few multi-generation policies, for example.
Obviously three generations out would be a bad bet to take.
Second is, of course, like all good bookmakers, life assurance
companies can lay off the bet, at least on average, by paying
annuities.  If you are betting with one slice of a cohort that
they are going to die quickly, and with another slice of the
same cohort that they are going to take a long time to die,
then being a bit off target will not kill you, as long as there
is no systematic bias between the two groups.  And if they
are, well, that is why life insurance companies make the money
in ordinary times.

But to the extent that the risk exposure is based on a system
that runs with a more rapid heartbeat than demographic changes
... obviously you cannot make an prediction with even that
partial degree of confidence.  Rather, you are accepting an
income in order to bear a truly uncertain risk, and as
re-insurers found after 9/11, sometimes there are systemic
shifts that undermine the valuations of companies that do that.

As far as "supernormal" trading, isn't there a question of
whether you identify the "supernormal" trader ex post or
ex ante?  Obviously those that are supernormal _ex post_
would make substantially better than normal returns ...
but those that try to be supernormal and fail? Say, only to
buy a quarter way down the decline during a temporary
correction, and to wait until well after the following peak,
believing that the next decline is only a temporary correction
when it turns out to be the start of a long bear market.
If you throw in everyone who ex ante intends to be supernormal
traders, wouldn't that be too large a share of the total market
to depart too far from normal returns in the market?