Re: poly: Re: Why interest rates may stay low

I wrote:
At 07:53 AM 3/13/98 -0800, you wrote:
>... the average ROI isn't much above the marginal ROI. The market value
>... is in this case burned up in the race to get there first.
>To the extent that the supply curve for investment funds is relatively
>flat (as it seems to me that the literature suggests), ... And to the
>extent that growth rates are tied more to the average ROI than to the
>total amount invested, growth rates ... may be relatively insensitive to
>the real ROI that technology allows.

Peter McCluskey writes:
> This reasoning probably works for most circumstances, but I'm hesitant
>about extending it to unusually large changes in ROI.
> This flat investment supply curve can only have been observed over a
>small range of real interest rates. While I can believe it's easy for
>many people to increase their savings rate from, say, 3% to 10% of their
>income, I doubt that there is an easy source of additional investment
>once savings rates start approaching 100% of income.

Yes, the empirical evidence about the investment supply curve is mainly
in the region of relatively small fractions of income, and one expects the
ROI required by investors to rise as savings approaches 100% of income.

> I expect that in the not-too-distant future, there will be at least
>one investment opportunity will test this upper limit of investment
>supply. I'm thinking primarily of the opportunity to launch self-reproducing
>probes at near lightspeed to colonize large parts of the galaxy, although
>it's also conceivable that a race to produce a molecular assembler or
>the first AI could do similar things.

It is possible that one might make a case that one of these techs is a big
exception to our historical experience, but I think a careful analysis would
be needed to make such a case persuasive. You have to persuade us that tech X
would create so many attractive investments that even when well over half
of world income is devoted to (prematurely) investing in tech X, the worst
investments would still be expected to have a persistently large real ROI
(say over 20%/yr continuing over a decade). This must be true even when one
considers the rise in price of any bottleneck property rights, such as
patents, LEO slots, or a limited number of relevantly skilled professionals.

I can't see how you could make such a case, especially regarding the case of
self-reproducing galaxy colonization (which I've recently been analyzing in
detail), but I'd love to hear a good try.

>>On average though, people should discount a factor of two per generation, ...
> This seems to imply a lower discount rate once most reproduction is done
>primarily by copying, ...

Yes, at least beyond the timescale when selection effects and the (unknown)
plasticitly of discount rates allows new time prefernces to arise.

Robin Hanson
hanson@econ.berkeley.edu http://hanson.berkeley.edu/
RWJF Health Policy Scholar, Sch. of Public Health 510-643-1884
140 Warren Hall, UC Berkeley, CA 94720-7360 FAX: 510-643-8614
Received on Mon Mar 16 18:32:49 1998