Speculative Demand for Money

Robert Waldmann

One of Keynes’ concepts which never made any sense to me was the idea that there is speculative demand for money. As far as I could tell, it was a very fance name for the liquidity trap — if the safe short term interest rate is zero, money is a close substitute for t-bills so money demand becomes enormous — effectively infinite.

This, of course, is not what he meant at all. But what could he have had in mind ? I think one thing is clear, he was, in effect, speaking in the first person. Keynes was an active investor, and he thought investors sometimes hold money for speculative reasons, so I’m pretty sure he recalled doing so from time to time.

My sense was that it is obvious if safe short term interest rates are positive, it can’t possibly be rational to hold money as a store of value, that is to hold money for speculative reasons. I conclude that I was unconciously making an assumption or two. Now I think I know what assumption I made. Assumption after the jump.

Of course I did not assume that money pays zero interest (this assumption is made for simplicity when presenting models to undergraduates). I assume that there is interest on money (I’ve never had a non interest paying checking account). Second I never assume efficient markets.

However, I think I assumed that there is a continuous stream of maturiing bonds, so there is always a bond 5 minutes from maturity. Well at least one day from maturity.
This is not true, at least, of treasury securities (as in safe you know).

This means that, most nights, the expected overnight yield on the safe security which is closest to maturity is and must be a prediction of its price tomorrow. In contrast reported yields are the hold to maturity nominal interest rate.

So it is clearly possible that someone might decide to hold an asset which will mature tomorrow (money) even if the reported yield on a safe short term asset is higher. Obvious.

I think another barrier to understanding what Keynes meant by speculative demand for money is that, when I was taught, the first step in getting micro foundations for macro models was assuming a representative consumer. This does not (yet) imply efficient markets, but it does rule out speculative demand for money. If the representative consumer thinks that the overnight return on money is higher than the expected return on a bond, she will hold only money so financial markets will not clear. If she expects the return on money is lower, she will only hold money to the extent that she might be forced to sell the bond before it matures — this is precautionary demand for money not speculative demand for money. If there is a representative investor, there can’t be specualative demand for money.

A big if. If there are different investors and each is convinced that he can beat the market, some will hold money. Others will want to short money. If they can, they are called banks. The money shorted by banks is called money supply (not negative money demand). That means that speculative long positions in money have to be counted as part of money demand.