RBI’s decision to keep its policy rate on hold last week and rather wait for more price data is appropriate and welcome. Though this surprised markets, the central bank focused upon what analysts chose to not: that November headline inflation rates were a spike driven by vegetable prices, which was expected to correct sharply in December post state elections. Though it remained worried about rising headline inflation, the central bank rightly focused on the stabilising non-food, non-fuel (core) WPI and CPI inflation, which it expects to further moderate from rupee stability, a negative output gap and past monetary tightening. For this interplay to transpire, the central bank saw it optimal to wait for another data round in what was a ‘close call’. Should these data disprove its hopes, RBI will promptly act, even ahead of the scheduled January review. The question then is if twenty-five days are sufficient for such restraining effects to feed through into December inflation, subduing it enough for RBI’s comfort? Or should it wait a bit more perhaps?
Consider its specific guidance on the three response parameters for this. First up, the expected softening of food inflation where RBI expects more contribution from central and state governments. Political economy makes it hard to predict movement on this front, especially in less than a month. But the declining trend in almost every item, including cereals, in the WPI and CPI food basket, for some months—vegetables being the sole exception and culprit in July-November—with some government action RBI took note of, there is reason to hope the past trend may sustain. Indeed, December’s sharp correction in vegetable prices in itself could bring down headline CPI inflation into a 9% to 9.5% range.
Next, RBI will look for declining momentum in non-food, non-fuel or core inflation. WPI-core inflation rose a bit to 2.66% in November, from 2.64% in October, and a revised 2.52% and 2.29% in September and August respectively (provisional, 2.11% and 1.94%). But as provisional measures are often revised at two-month lag, and 35-41bps revisions were made to August-September numbers, it is reasonable to add about