## Tuesday, June 03, 2008

### Monetary policy must respond to changes in future inflation

Writing in Mint today, Vivek Moorthy argues that there is a conflict between inflation targeting and pre-emptive attacks on inflation.

Because monetary policy impacts upon aggregate demand and hence inflation with a lag, monetary policy formulation must necessarily be concerned with forecasts of inflation. Inflation targeting inherently means : watching future inflation and modifying the short rate so as to influence future inflation.

If Vivek is saying: It's silly to target last month's inflation, I could not agree more. The job of monetary policy is to anticipate future shocks to inflation and respond to them well ahead of time. Last month's inflation is an input into forecasting future inflation, and is only one input into this task. A smart inflation targeting central bank must forecast inflation, and act on what it sees in the forecasts.

As an example, when RBI setup an INR/USD depreciation of 9.7%, from 39.12 on 1 Feb to 42.93 on 22 May, it was easy to forecast that this was going to exacerbate inflation over a horizon of roughly 4 months. We know that a shock to the currency of 1% induces a shock of between 0.2 and 0.3 to WPI over a horizon of roughly four months. In other words, a shock to WPI of 2 to 3 percentage points is in the offing for WPI data from June to September, purely owing to the movement of the currency. Such empirical strategies can yield a lot of insight into future inflation.

I think a key strength of an inflation targeting central bank is that it is not in a fog of conflicting goals. There is only one goal. The entire organisation is focused on one thing. This generates better work on forecasting inflation and acting on these forecasts.

While on this subject, do read this piece about the `Taylor principle'. This is a necessary condition for a sensible monetary policy, and it's entirely phrased in terms of shocks to inflation that is expected at future dates.

#### 1 comment:

1. Good to see your emprical analysis - 10% of exchange rate change leads to 2-3% increase in inflation getting wider acceptance - it would have been better with attribution.

But on the inflation subject - Mr. Aiyer's column in ET on Rajan committee recommendation of inflation targeting seems way off. (As an aside, I thought Dr. Mistry was peeved at Dr. Acharya, now Dr.Aiyer is going p-off Dr.Rajan) He starts off with in India you can not have RBI focus on inflation alone (he is mixing "core" CPI with general inflation) because food inflation, hurts the poor disproportionately. But then he goes on to argue that RBI can not control food inflation anyways - if they can not control food inflation then what is the problem of letting RBI control "core" inflation - general inflation minus energy and food?

Please note: LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.