Was a rate hike required?


Business Standard, 19 April 2006


In an environment where he expects inflation to rise, RBI Governor Y V Reddy has chosen to wait and watch and not raise rates now. This may be popular, but it is destabilising. The Taylor Principle asserts that when inflationary expectations accelerate, as the governor indicates, there is a strong case to have higher interest rates.

A careful look at inflation shows that there has, indeed, been a small but clear rise in inflation in the last two years. Annual YoY growth in the WPI has gone from below 4 per cent in 2002 and 2003 to above 4 per cent in 2006. The WPI is a severely flawed measure of inflation. The CPI, which was below 4 per cent in 2004 and 2005, has inched towards 5 per cent in the last four months. More importantly, in the light of higher global oil prices, inflationary expectations have set in. As the credit policy notes, "It appears that underlying inflation conditions are perhaps not being appropriately reflected in prices" and the outlook on inflation remains "clouded".

The difference between the short-term interest rate (which is controlled by the central bank) and expected inflation is the short-term real rate. Sometimes, it is felt that the central bank should meekly accept small increases in inflation. Sometimes, it is felt that a rise in expected inflation should be matched by a rise in the short-term interest rate, so as to deliver a stable short-term real rate.

In the last decade, a powerful new idea has sprung up in the field of monetary economics, called "the Taylor Principle". This asserts that when there is a 100-basis-point increase in expected inflation, the central bank must increase interest rates by more than 100 basis points. If and only if this is done, then monetary policy is acting to stabilise GDP. If this is not done -- if monetary policy is not hawkish when there is a rise in expected inflation -- then monetary policy is destabilising.

The Taylor Principle was postulated by John B Taylor, the same person who invented the "Taylor Rule" and more recently was in the US Treasury. John B Taylor and Michael Woodford are expected to share the Nobel Prize for their path-breaking work in monetary economics in the post-1993 period. The theoretical and empirical work underlying the Taylor Principle is enormously complex, but the basic intuition is remarkably simple.

Suppose there is a 100-basis-point increase in expected inflation, and suppose the central bank responds with an increase in the interest rate of only 50 basis points. In this case, the real rate has gone down by 50 basis points. In other words, when expansionary forces are acting on the economy, monetary policy is (additionally) expansionary. When times are good, monetary policy acts to make them better.

In reverse, suppose there is a 100 bps point drop in expected inflation, and suppose the central bank responds with a drop in the interest rate of only 50 basis points. In this case, the real rate has gone up by 50 basis points. In other words, when contractionary forces are operating on the economy, monetary policy is (additionally) contractionary. When times are bad, monetary policy acts to make them worse.

These two examples show that when there is a 100 bps rise in expected inflation, if the central bank is not hawkish, and rates are raised by less than 100 bps, then monetary policy is destabilising; monetary policy exacerbates the volatility of GDP. The Taylor Principle asserts that the only stabilising monetary policy is one which responds to a shock in expected inflation by greater than one-for-one. If the RBI wants to help stabilise GDP growth rates -- which is the ultimate goal of monetary policy -- then it has to be hawkish in responding to expected inflation. Interest rates must go up by more than one-on-one when expected inflation goes up. Conversely, interest rates must go down by more than one-on-one when expected inflation goes down.

Empirical evidence in support of the Taylor Principle, across numerous countries, is extensive. A few fascinating fragments concern the US economy. In the 1960s and 1970s, the US Fed had an inflation coefficient of 0.8. That is, on average, when expected inflation went up by 100 bps, rates were only raised by 80 bps. This violated the Taylor Principle, and US GDP volatility was high. From 1979 onwards, the monetary policy of Paul Volcker and Alan Greenspan has been compatible with the Taylor Principle. The inflation coefficient went up to 2.1. There has been a remarkable drop in US GDP volatility in the post-1979 period.

While other factors, such as financial globalisation and improved information technology, have surely played a role in calming GDP volatility, there is a strong consensus that adhering to the Taylor Principle has been an important part of the story of reduced US GDP volatility. Similar characteristics have been found in numerous major economies, particularly those that have adopted inflation targeting as the legal foundation of their central bank.

Existing estimates show that in the past, India has violated the Taylor Principle. Indian monetary policy has not responded properly to changes in expected inflation. The policy has been destabilising; interest rates have been raised to choke the economy at the wrong time, and they have been dropped at a time when they exacerbated inflation.

One difficulty here is the lack of proper measurement of both inflation and expected inflation. Bad inflation measurement has long bedevilled Indian economics. In the absence of a well-traded yield curve or inflation-linked bonds that measure expected inflation, knowing expected inflation is hard. Another problem appears to be the RBI's use of a pegged exchange rate regime. Given that de facto the Indian capital account is open, chasing a currency target comes at the price of monetary policy. The RBI's levers of power are being used up to implement the pegged exchange rate, so there is no space left to react to fluctuations of expected inflation in the business cycle.

Given that inflation measurement in India is foggy, one has to be necessarily tentative about where we stand. However, it appears that there has been a pronounced move in the WPI and the CPI, and that there are expectations of higher inflation in the days to come. Under these conditions, a monetary policy which targets macro stability would involve higher rates.


Back up to Ajay Shah's 2006 media page
Back up to Ajay Shah's home page