Re: poly: Why interest rates may stay low

From: Hal Finney <>
Date: Fri Feb 27 1998 - 14:16:15 PST

I should know better than to ask basic economic questions, but I *almost*
understood Robin's article, so...

Here's how I picture the situation Robin describes:

  M | D
        | DD
  O | DD
           | DDDD SSS
            1% 2% 3% 4% 5%

               I N T E R E S T R A T E S

This shows the supply (S) and demand (D) for investment money as a
function of interest rate (= the cost of money). The money is
supplied by investors, and the higher the rates, the more they are
willing to part with it, so the supply curve slopes upwards. This
represents people's willingness to give up money today in return for
getting more in the future. Robin suggests that the curve is roughly
flat in the range around where historical interest rates have been,
1.5% to 3.5% or so, so I have drawn it that way.

The demand curve slants the other way; with lower interest rates, more
people want to borrow. We are focussing on industrial demand, fueled
by business growth. A company borrows because it can invest in new
production equipment using the borrowed money, grow, and pay off the
loan using their higher profits.

Supply meets demand at the point marked **, which will be the current
market interest rate.

The naive analysis, which Robin contradicts, says that as technology
improves, companies can increase their productivity more for a given
investment. This means they can afford to pay more for money due to
their higher growth. This should shift the demand curve to the right:
a company which used to be willing to pay 3% is now willing to pay 4%.
The equilibrium point will shift as well, causing interest rates to
increase and giving more profits to investors.

That's a simplistic model, but there are some aspects of Robin's analysis
which I don't understand:

> First, one the supply side, there are good reasons to
> think the supply curve is basically flat at about 1-2%.
(Later Robin corrected this to 1.5-3.5%.)
> This is the rate predicted by evolutionary analsysis,
> being about a factor of two per generation. And I think
> historical risk-free interest rates have stayed about
> this number for many centuries. This is also the average
> return on stocks worldwide over the last century, once
> you account for selection effects.

I don't see why these various facts imply that the supply curve is
flat. A flat supply curve means that you won't get many more people
willing to invest at 3.5% than at 1.5%. How does this relate to the
fact that historical returns have been in this range? Is the point
that historically, people haven't invested any more when returns were
at the high end than at the low end? I would have thought you'd see a
lot more money offered for investment when returns get twice as high.

It's also not clear what the significance of the flat supply curve is
in Robin's description. It seems to relate to what happens to the
demand curve, which is also unclear to me:

> On the demand side you have to realize that over the long
> term there aren't really property rights over most of the
> main technological investments. Most any company or nation
> could try to be the future providers of telecom, software,
> banking, electronics, toys, or whatever. The game is
> wide open. But that means they all entrants should expect
> to make about the same return on investment (though of
> course actual returns can vary widely).

This sounds plausible. If I enhance my plant with better robots to build
shoes faster, my competitors can easily do the same thing.

> How does this square with an image of a downward-sloping
> demand curve given by the available investments, with
> the average return higher than the marginal return?
> It squares by realizing that while there might be an
> optimal time to make any one investment, say trying to
> create an internet directory, the investment will happen
> at the first time that it's expected return reaches the
> going interest rate. That is, without property rights,
> there is a race to be first, a race which burns up all
> the value of the investment over the marginal rate.

So, what ends up happening to the demand curve? Does it stay the same?
If so, what did it matter what the shape of the supply curve was?

I've enhanced my show factory with robots and it can produce three
times more shoes than before. But so did everyone else, plus a bunch
of other people opened robot shoe factories, and the price has fallen
so low that I'm not making any more profits than I was. So I can't
afford to pay any higher interest than I was before, and my demand for
money has stayed the same. This scenario corresponds to a stationary
demand curve.

But society as a whole benefits because shoes are really cheap. This
means that everyone has more shoes, and they also have more money to
spend on other goods. If some of this money becomes available for
investment, raising the supply curve, that will actually lower interest

> So regardless of how much technology makes investments
> intrinsically profitable, without property rights the
> average return on investment stays at the marginal rate,
> which is determined by our evolutionary heritiage on
> discounting time, at 1-2%/year. And to the extend that
> growth rates are tied to the average return on investment,
> growth rates will also say moderate.

With my shoe example, growth in terms of dollars' of goods produced
was slow. But growth in terms of shoes produced was very high. We
had a drop in shoe prices, and measuring growth in dollars ignores
this. Could it be that the moderate growth rates are in deflating
dollars, with the growth in goods happening much faster?

Received on Fri Feb 27 22:46:09 1998

This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:45:30 PST