Monetary policy: RBI should specify its inflation metric

Updated: December 2, 2014 5:14:57 AM

Though the economy is weak, underlying inflation is still too high to warrant a rate cut

Today is the much awaited credit policy day. With continuing weakness in the economy, and a steep drop in CPI and WPI inflation in recent months, the chorus for a rate cut has intensified. Recent statements by the finance minister about a rate cut can only add pressure on the Reserve Bank of India (RBI) to heed this chorus.

Against this backdrop, the question is whether RBI will cut the repo rate from 8%? Going by a poll of money market experts in a newspaper last Monday, the unanimous response was ‘no’. However, one can never tell, and I am not in the business of forecasting what RBI will do.

The Urjit Patel Committee inflation-targeting recommendations in January 2014 were that RBI should wait until inflation falls along its ‘glide path’ of 8% by January 2015 and 6% by January 2016 before formalising its inflation target of 4% to be then adopted by Parliament. Only then should it ease. The umpteen and vociferous critics arguing for a rate cut legitimately point out that if inflation is already below the January 2016 target 6% (5.5% for October 2014), what is RBI’s justification for not cutting rates? Especially when the economy is so weak.


The decision would have been made up by now. Nevertheless, in my opinion, RBI should not cut rates now. But to justify this decision, RBI governor Raghuram Rajan should clearly state that his focus is on a much lower metric of inflation—such as the three-year average. The term metric, as used here, refers to the frequency at which inflation is measured and/or how exactly it is calculated. The term measure, as used here, refers to the variable being measured; for example, CPI, WPI etc.

In general, most statements about high frequency metrics of inflation (not just intra-year but even year-on-year) can be characterised as “noise about noise” (Overheating and Underheating, Moorthy and Kolhar, Business Standard, August 12, 2007). Indeed, that article pointed to the inflation risk based on a rising three-year metric of CPI. It was written as a rejoinder to Surjit Bhalla (When will they ever learn?, Business Standard, June 23, 2007). In his article, he concluded that “look out for GDP growth above 9.4% to even begin thinking about an overheating India.” In recent years, we have seen the damaging inflationary consequences of such growth fuelled by, among other factors, RBI liquidity (as Honey Singh might say, Char botal vodka, 9% growth ka!).

The accompanying chart plots the standard year-on-year metric of inflation and three-year average. This is calculated as an average of year-on-year inflation of last three years for that month; for example, the value for October 2014 is an average of year-on-year rates for October 2012, 2013 and 2014, respectively. The inflation measure chosen is from the Urjit Patel report. This is the new CPI available from 2011 onwards and ‘backcasted’ to 2001 based on the prevailing CPI (industrial worker) series.

If we instead use the CPI (industrial worker) for the full period, there is not much difference: The standard year-on-year measure is subject to huge swings but the three-year measure is smooth. The choice of metric is a vastly ignored issue in monetary policy, and matters much more than the measure. (For evidence on these issues, see Inflation Measurement for Policy doctoral thesis of my student and co-author, Shrikant Kolhar, IIM Bangalore, 2013, available from

What does the chart show? For the decade, the huge swings in the year-on-year measure are evident while the three-year average moves far more smoothly. Within three-year averages, there are other metrics. The huge swings in the former are evident while the latter moves much more smoothly. For occasional shocks and spikes in the price level that transmit to the inflation rate, RBI or any other central bank can invoke the ‘base effect’ to avoid going by its inflation target and/or follow a distant target date as RBI is now doing: 6% in January 2016. But it is ad hoc to always keep invoking the base effect.

It is far better and simpler to choose a low frequency metric with a near target date. Otherwise the glide path is a wide and uncertain path, and RBI loses credibility. I request governor Rajan to consider announcing to the financial markets and public which metric RBI will consistently follow.

What does the proposed metric show? The headline metric dropped sharply from about 11% last November to 5.5% latest. But the three-year average has fallen about 200 basis points from last November. But, at 8.5%, it is still above the 8% repo rate. Whatever the drop in inflation, cutting the repo rate further below may lead to monetary instability and a flight to real assets etc. Unless credit demand weakens so much that market rates tumble and the repo window goes into huge surplus mode, RBI should not ease. In that case, it has to ease anyway.

I should clarify that I am not in favour of (direct) inflation-targeting but in favour of a Taylor strategy. Starting at the desired inflation rate, the policy rate should respond to both inflation and the strength or weakness in the economy. However, though the economy is weak, at present in India, underlying inflation is still too high to warrant a rate cut.

(This article draws upon the author’s presentation made at an RBI organised seminar on October 20, 2014)

By Vivek Moorthy
The author is professor, IIM Bangalore

Get live Stock Prices from BSE and NSE and latest NAV, portfolio of Mutual Funds, calculate your tax by Income Tax Calculator, know market’s Top Gainers, Top Losers & Best Equity Funds. Like us on Facebook and follow us on Twitter.

Switch to Hindi Edition