On Fri, Sep 09, 2011 at 12:50:39AM -0500, Venkat Inumella wrote:
> I'd say that many characterizations of individual decisions as irrational
> are precisely a result of narrowly defined utility functions. Saying that
> the rational decision in buying shoes should be based just on ROI as you
> calculated, for example, makes anybody with a strong preference for
> expensive purple shoes invariably seem irrational, even though that person
> was only maximizing his utility.  If you allow for the fact that everyone
> has a unique way of measuring utility, individual decisions wouldn't seem
> quite as irrational anymore.

The problem is that in experiments you can get people to express cyclic
preferences (they prefer A to B, B to C, and C to A), preferences that are
non-monotonic with time (preferring receiving something in three days to
receiving it in either two or four days), and preferences that vary enormously
depending on how things are worded.  And then people's actual choices often
vary from their stated preferences, and their self-reported satisfaction can
conflict with both of those.

These all conflict with the Folk Psychology model where people imagine how
satisfied they would be in various possible future states of affairs, then
undertake the actions that their beliefs predict would bring about the future
state with the highest satisfaction (or, in more sophisticated formulations,
the probability distribution of future states with the highest expected
satisfaction).

> > Perhaps I'm economically naive.  Why is consumption spending special?
> 
> It is in fact financial investments that are special: people make all
> decisions to maximize some (usually multivariate) utility function. For most
> financial investments, however, this utility comes almost entirely from ROI,
> so it's a valid simplification to maximize only ROI, thereby maximizing
> utility as well. As long as I'm maximizing returns, I don't quite care if
> I'm investing in a Vanguard fund or a Fidelity fund, in equities or bonds,
> locally or internationally. Not true for my shoes.

Okay.  But you still need to calculate NPV (or something similar, like ROI) to
compare the costs of alternative proposed consumption plans, don't you?  Or do
you think that it's meaningless to compare the costs of two possible
alternative purchases, simply because you might prefer the result of one
purchase to the other?

I'm not advocating that people should *always* buy the less expensive
alternative, only that they should take prices into account, and that ROI is a
perfectly good shorthand way to compare the price of buying a durable good to
buying a stream of non-durable goods.  But you seem to be taking an
unreasonably extremist position in the opposite direction, arguing that there
is no objective way to compare these prices.  (Does it therefore follow, in
your belief, that you should just buy whatever you like best, regardless of the
price?)

By the way, a second chappal broke today, doubling my data points on their
reliability and doubling my estimate of their failure rate, and thus the cost
per year: US$8 or so, probably slightly more than the Red Wings at 3%.

> > > > it *never* wears out or depreciates.  You can probably find
> investments with a
> > > > 3% ROI in almost any economic times and at almost any scale.
> > >
> > > Any number of hedge fund managers would beg to differ, I'm sure!
> >
> > Can you elaborate?
> 
> There are few short-to-medium term risk-free investment opportunities in the
> US that will fetch an absolute 3% return currently. Were such an investment
> available, especially "at almost any scale", US treasuries wouldn't be
> nearly as widely held, given their current near-zero yields (
> http://www.treasury.gov/resource-center/data-chart-center/interest-rates/Pages/TextView.aspx?data=yield
> ).

There are no risk-free investment opportunities anywhere, not even in the US,
however much hedge-fund managers would like to fake them.  But you can still
probably get 3% over the next 20 years investing in the stock market there.

Kragen

Reply via email to