With its forecast coming so horribly wrong, the central bank will need to see what is wrong with its model
Though consensus opinion among economists seems to be that the central bank will not cut repo rates today, but that it will soften its stance right now—and then cut rates in August—what is more important is what RBI chooses to say about its inflation-forecasting methodology. Everyone can, and does, get forecasts wrong, but that it should happen so often, and for even short intervals, is surely worrying. In October 2016, when headline CPI inflation was at 4.2%, RBI projected March 2017 inflation at 5.3%—inflation had been 6.1% just a few months ago in July 2016. By December, when inflation had continued to fall and was at 3.4%, RBI lowered its March 2017 inflation outlook to 5%, albeit with an upside risk. Inflation continued to remain benign and though it had risen marginally in February 2017, RBI’s March 2017 forecast was under 5% because, as the central bank said, “persistence of inflation excluding food and fuel could set a floor on further downward movements in headline inflation and trigger second-order effects”. The March 2017, headline CPI, however, wasn’t just under 5%, it was under 4%—in other words, RBI was projecting inflation one month down the line and it was off by 110 basis points! For April 2017, the number is down further to 3%. Given that inflation was high in both May 2016 and June 2016—what RBI calls a favourable base effect—it is possible inflation could also be low in May and June 2017. Keep in mind, RBI’s forecast for June 2017—this was made in April 2017—is 4.2%.
One reason why inflation could remain low is the monsoon is forecast to be good, a good crop is likely and, with large unutilised capacity, pricing power is likely to remain poor—HSBC, in fact, is looking at the output gap closing only by the end of 2018. Indeed, while RBI is wont to talk of inflationary pressures in areas excluding food, it is worth keeping in mind that, in terms of contribution to inflation, non-food items have fallen from 4.1 ppt in January 2014 to 2.6 ppt in January 2016, and it has remained at around that level since, the surest evidence of low pricing power.
One reason why RBI could be getting it so wrong could be the fact that, though the central bank says it doesn’t pay much attention to its household expectations survey, it may just be—otherwise, why not just junk it? The latest survey, interestingly, has a median inflation expectation of 7.5% in FY18! The survey of professional forecasters is better than the household one, but it too has been wrong. Why even private sector economists should be getting it wrong is not clear, but one reason could be in the definition of ‘core’ inflation. While RBI is of the view ‘core’ inflation is sticky, as FE columnist Surjit Bhalla and Bank of America-Merrill Lynch’s India economist Indranil Sen Gupta point out, the measurement of ‘core’ is faulty due to not stripping out fuel prices completely—according to Bhalla, ‘true core’ is at its lowest levels since 2011. Whether RBI cuts rates or not, we need a better explanation for why its forecasts have gone so wrong. If the review is thorough, it is possible rates could also be cut.