« Signs of population aging | Main | Should new editions of intermediate microeconomics textbooks be banned? »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Isn't the main reason why most textbooks use exogenous money supply (i.e. a vertical Ms Curve) because it avoids the issue of there essentially being two equilibria given the standard hicksian Md curve (i.e. one intersection at the classical cambriddge k and another intersection at the liquidity preference "section" of the Md curve)? Antoher reason I could imagine is that it presents the historic case where the government wants to retain control of money (my gripe with "MMT" often is, that their theory seems to be less of a soft/fiat money theory, but rather a "the gov't is really liberal with banking licenses and terrible at regulating theory" - but that just might be "political economics: political science guys bullshitting about economics").

But this leads me to an issue i have with your reconciliations, because the way i understand it it would imply that loanable funds is indeterminate in a centrally planned economy. But how do we then decide in war time between producing tanks now or producing more tanks in the futre by upgrading factories? I was under the impression only loanable funds could explain this, because for every machine you want to build, you have to take workers/machine parts/whatever factor away from the tanks or soldiers away from the front. A non real theory of the real interest in that scenario wouldnt make sense, but if the real theory were indeterminate, wouldntthat mean that we could never decide how many tanks to produce and thus wars shouldnt happen? Maybe I am completely missunderstanding everything though (in fact I am sure I am)

For what it's worth, I tend to agree with ritwik that the conception of the supply of loanable funds as "saving" may be a source of problems. The loanable funds market is the market in which money (typically deposit instruments) is traded for bonds. The interest rate is just a way of stating the price of a bond. Indeed, bonds are quoted by either price or yield; they are identical expressions, with every yield corresponding to a unique bond price and vice versa. In effect, borrowers (issuers of bonds) are renting purchasing power, and lenders are "renting out" purchasing power.

How does this relate to "saving"? Suppose my employer pays me at the end of the month in the typical fashion: by assigning me a (small) part of its deposit instrument. Suppose I retain the deposit instrument and do not trade it for a bond. Have I "saved" it? I would say yes. But I have not added to the supply of loanable funds. (Note that the balance sheet of the bank is not changed when my employer assigns me part of its deposit instrument.) I could trade the deposit instrument for bonds if I wanted to -- in which case I *would* be augmenting the supply of loanable funds. But the supply of loanable funds does not seem to be identical with "saving."

This gets to a more general point. Let's call the market for loanable funds the "financing market" -- the rental market for purchasing power. I would argue that the interest rate (the rental price of purchasing power) -- which is just a way of quoting the price of a bond -- is determined in the financing market, and nowhere else.

Here is an analogy. There is a rental market for apartments. Suppose a bunch of apartments fall into a sinkhole. The rental price of apartments goes up. You would depict this development in the "rental market for apartments" as a leftward shift in the supply curve for "apartments for rent." You would not say that the rental price for apartments is *also*, separately determined in a "market for apartments," and that your two theories then need to be "reconciled" in a separate model. What would be the point of that?

I think The Fisher diagram is the most informative. You said:

"The tangency with the PPF tells you how much firms want to divert resources away from producing output for current consumption towards investment for output of future consumption."

There are 2 major problems here.

1) Note that we're diverting resources, we're NOT magically exchanging current consumption for future consumption. The relationship between the current consumption sacrificed and the future consumption obtained is complicated and subtle, and calling it intertemporal substitution is very misleading and the source of a great deal confusion.

At the micro level, I attempt to transform the utility of x dollars of current consumption into an amount of future utility consumption u, where u is a function of my NPV utility discount and a weighting for risk and uncertainty. To do this I have to estimate the price of that utility integrated over my risk and uncertainty weighted estimate of the probability of consumption, which in turn requires an estimate of a personal consumption price deflator function over the time period.

This is obviously an insoluble problem, and can only be approximated by a heuristic. Different people choose different heuristics, based on intellectual preference, economic situation, social influences and personal psychology. A large part of the economics of saving is determining a simple way to approximate the distribution of heuristic preferences as a function of simple economic observables.

2) There is also a problem of available resources. The consumption that a person saves is a share of the production of the time of saving. The consumption of dis-saving is a share of the production of the time of dis-saving.

The amount of this production will depend on the quality of the investment of the sacrificed consumption (this is true both in micro and in macro), but also on current and estimated future economic trends (which estimates may not be correct and may even have perverse effects), unpredictable future events and relatively predictable demographics.

When these 2 factors interact with feedback through expectation channels, life gets very interesting. Obviously the saver in factor 1 is updating expected returns based on observations of factor 2, while the economic circumstances depend on actions taken based on these observations and the estimates of trends are influenced by changes in investor beliefs.

This is a very difficult system to model over the long term. It is also, as with any nonlinear feedback system with large delays, potentially unstable.

If savers believe the future pie will be smaller than they anticipated, they will want an increased share such that their expected real return is unaffected. This would be at the expense economy of future labor, which has a number of ways of resisting (default, voting with the feet, diversion to the underground economy, social and political unrest...).

The prospect of such instability is, itself, something which can have effects through expectation channels. Investment in such an economy is unattractive, and investors would expect a large risk and uncertainty premium.


Exactly! The theory of capital is joint at the hip with the theory of interest but the theory of capital itself is most contentious. If you treat capital as the result of inter-temporal optimisation of consumption and production, you will believe that loanable funds and savings are one and the same. But consider the expansion of gross leverage, or firms accumulating/ disgorging retained earnings on aggregate - it is easy to see that financing and saving are not the same. National income accounts are important, but so is the flow of funds account.

And then, if you really want to mess things up more, you can add liquidity preference, properly defined as the refusal of the cost of capital to come down in response to monetary interventions.

Which is why the one way I like to rescue inter-temporal optimization even in the loanable funds = financing view is to take a radical full neo-classical general equilibrium position position (but allowing for uncertainty). Financial wealth is real wealth, wealth is aggregate supply, and that the aggregate supply and interest rate are both stochastic, so that financing decisions boost aggregate wealth and loanable funds itself reduces to the increase in capital net of consumption. S=I, horizontal IS curve. It's my interpretation of Fischer Black's model. Debates about business cycles and equilibrium/ disequilibrium then reduce to debates about the market model (inter-temporal CAPM) and decision making under risk and uncertainty. It is a non-monetary model but you can still rescue monetary policy by replacing the AS curve with an LM curve (both are 'wealth' curves, fundamentally and you can't have both - the system will be overdetermined) and saying that the aim of monetary policy is to minimize any misperceptions of wealth.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad