May 24, 2008

End to Price Controls by Central Banks

It is becoming to be quite evident that control of short term interest rates by central banks is prone to be a cause of market disturbances, including speculative manias and crashes. Financial bubbles have of course existed much before the existence of central banks. The current problem, however, is a common problem in situations where prices are set by a third party that is not fully in control of supply. Such price controls cause both rampant overproduction or overconsumption and shortages, depending on whether the prices are set too low or too high.

Central banks used to be in control of the supply of money, but since the early eighties, they have given this power away by drastically reducing reserve requirements of commercial banks. This means that commercial banks are now primarily in charge of the development of the money supply.

This divergence between the power to control prices and the power to control supply is inherently unstable, so we have to go back to the old days of strict control of bank lending, or we have to go forward to release short interest rates from the clumsy control of central monetary authorities.

Central banks seem to be aware of their loss of control, but there does not seem to be much thought about what this implies. Price controls by decree without control of supply have been found to be a bad solution to economic problems.

Discussion of the possibility of US Federal Reserve paying interest on bank reserve balances has been a hot topic earlier this month. The Fed seems to be working towards implementation of an interest rate corridor policy, which has already been adopted in Canada, UK, and New Zealand.

In this system, the central bank sets both a lower and an upper bound for the interest rates of over-night inter-bank loans. This is accomplished by paying a minimum interest on cash balances at the central bank. There is no reason for banks to offer loans to other banks on interest any lower than that. The upper limit is the discount rate, at which banks can borrow directly from the central bank.

A 2006 paper titled "Monetary Policy Implementation Without Averaging or Rate Corridors" by William Whitesell discusses a third alternative for monetary policy application, but the focus of this paper is hopelessly fixated on how the central bank could fix the inter-bank interest rates as tightly as possible. What is missing is any discussion of the sense in trying to achieve such a rigid goal in the first place.

My intuition would tell me that a fixed interest rate for short term lending is inherently a bad thing. Lending rates should accurately reflect the existing earning opportunities and costs of investment activity. A misalignment between opportunities and policy rates create bubbles of hyper-active speculation, because of seemingly risk free opportunities for arbitrage between short term borrowing and lucrative investment opportunities. On the other hand, a misalignment between costs and policy rates create unnecessary slow-downs in investment and savings.

An ideal would be a market-determined interest rate that would dynamically reflect changes in actual opportunities and costs. When a dislocation would suddenly create an investment opportunity, there is always a risk of a speculative bubble. Short term interest rates should react swiftly to sudden increases in demand for credit. If the opportunities are real, it is only just that the savings customers of banks get their fare share of the investment income. If the opportunities are a mirage, the sudden rise in interest rates would quickly deflate the emerging speculative bubble without letting it gather too much steam.

I have been thinking of a mechanism by which central banks could fulfill the function that they were created for, ie. liquidity support, without actually engaging in direct control of interest rates.

First let's be clear about what liquidity actually is. Liquidity is a multi-faceted concept. It has something to do with the volume of trade. It has something to do with transaction costs. Liquidity is not a synonym for lack of volatility, which is possible in perfectly liquid assets.

Liquidity can probably be best described by insensitivity of market prices to a unit volume of trade transactions. It is also related to the amount of losses incurred by simultaneous buying and selling in a single market.

In a trading system there is a certain volume of bidding and asking at different prices. In a real time system, there is a gap between these. This bid-ask spread can be used as a quite reliable measure of liquidity.

I have been thinking of a liquidity provision system through a dynamic interest rate corridor. Instead of a policy rate, set by a committee, the central rate of the corridor would be based on a market average.

In such a system, all interbank lending would pass through a transaction system that is managed by the central bank. The central bank gathers pricing information from this system, and determines an average market price. A suitable margin is then put around the central rate to determine an interest rate corridor. This corridor guarantees a maximum bid-ask spread, ie. a minimum liquidity, at with banks can directly transact with the central bank.

The central rate could be calculated by taking a significant amount, say $100 million, of both the highest asking rates and the lowest bids. By taking a weighed average of these, a neutral market rate would be determined. A meaningful bidding amount should be used for the determination to prevent market manipulation with insignificantly small bids.

Banks would be free to transact with the central bank at the corridor rates, but they could still get better rates by bidding inside the corridor. The corridor should be wide enough for everyday trading to occur between the market participants. But in times of uncertainty, when the spread would otherwise be stretched uncomfortably wide, there would still be a guarantee of minimum liquidity from the central bank.

In times of stress, such a market-based interest rate would be equally uncomfortable to both sides of the market, which is as close as one can get to neutrality. A strictly enforced policy rate, on the other hand, can create highly biased situations, in which either the lender or the borrower would benefit at the expense of the other.

I am quite completely ignorant of most issues of monetary policy, but I would hazard a guess that the level of activity in such a dynamic interest rate corridor could be used as a measure of monetary conditions. Persistent borrowing from the central bank could be a sign of the need for increasing reserves. Persistent deposits would signal a need for reduction.

No comments: