| Surjit S Bhalla: Can RBI's forecasts be trusted? |
| Fiscal policy is forced to be expansionary as monetary policy is failing to perform |
| Surjit S Bhalla / New Delhi August 1, 2009, 0:45 IST |
The RBI governor, Dr Subbarao, recently announced that he was seeking discussion and perhaps even criticism from within his organisation. This is definitely newsworthy and Dr Subbarao should be applauded for taking this initiative. The RBI is perhaps the last of the feudal organisations in India (along with all the political parties) and this attempt at an entry into the 20th century is laudable. I wish Dr Subbarao luck; having worked on two RBI committees a decade apart (in 1997 and 2006, under the chairmanship of former Deputy Governor and a true-blue RBI man Mr Tarapore) I can say with some experience that the RBI does not take lightly to anti-feudal forces.
No sooner had Dr Subbarao made his plea for dissension than the empire struck back. In its quarterly review on Tuesday, July 21, the RBI viewed the economy in a dour manner (why so serious?). It kept tight monetary policy tight (highest real interest rates in the world if one uses the GDP deflator) and warned of impending inflationary dangers. Believing full throttle in this gloomy stagflation outlook, the RBI lowered the forecast for GDP growth for 2009/10 from 7.5-8 per cent (made in January 2009) to 6 per cent. Correspondingly, it raised its forecast for WPI inflation in March next year from 3 per cent to 5 per cent. These pronouncements are put into focus by noting three facts. First, internationally, India is the only economy that is lowering its GDP forecast, while most are debating not that GDP will be higher in 2009, but how much higher. Second, while all expect inflation to be higher than zero inflation, there is no central banker of a non-banana republic (that I know) who is forecasting this high inflation.
The third fact is perhaps the most damning. The table shows the past forecasts and the errors on both. Note that it is nobody’s contention that the forecast errors should be zero. That would be like forecasting the past. What is desirable is that the forecasts have a randomness to them such that over time the errors add up to zero. Unfortunately, nothing of the sort occurs with RBI forecasts. Very consistently, the RBI under-estimates GDP growth by about 1 to 1.5 per cent — it gloriously missed the entire growth acceleration between 2004 and 2007. In May 2004, the growth forecast for 2004/5 was 8.1 per cent, but in October it got lowered by 2 percentage points. Which means that the RBI was expecting GDP growth (in October 2004) to average only 4 per cent for the next two quarters! It turned out to be twice that rate.
In 2008, perhaps the RBI noted its erroneous ways and started forecasting higher GDP growth for the great crisis year of 2008/9. At the peak of the crisis (July 2008), it forecast GDP growth of 8 per cent. A month later, year-on-year industrial production was reported to be negative — the very first negative number in the developing world, suggesting that the great Indian slowdown of 2008 was almost entirely a home-grown affair (note that the world collapsed a full three months after the Indian collapse and after Lehman in September).
The inflation forecasts are no better, and in many respects shockingly worse. (That this might have something to do with the deeply flawed quantity theory of money model that the RBI uses has been commented upon ad nauseum in these columns.) The data are from quarterly reports of the RBI. In end January 2009, which is two months before the target of the forecast (March 2009), the RBI’s considered assessment was that year-on-year WPI inflation would be 3 per cent. At that time, the WPI index was 229.6 and the March 2008 WPI figure was 225.5. A 3 per cent year-on-year increase would mean an index level of 232.3 in March 2009. Which means that in just two months the RBI was expecting the index to rise by 1.2 per cent or close to 7 per cent at an annual rate. (If seasonal factors are incorporated into the exercise, which they should, but which the RBI adamantly refuses to incorporate into its thinking, the “performance” would be worse.) This when the world was rightfully talking of the genuine possibility of a second great depression worldwide.
| RBI: GDP AND INFLATION, FORECASTS AND ERRORS, 2004-09 (in %) | ||||||
| RBI forecast in: | GDP Growth | WPI Inflation | ||||
| Forecast | Actual | Error | Forecast | Actual | Error | |
| May-04 | 8.1 | 8.3 | -0.2 | 5.4 | ||
| October 2004 | 6.0-6.5 | 8.3 | -2.1 | 6.5 | 5.4 | 1.1 |
| April 2005 | 7.0 | 9.2 | -2.2 | 3.9 | ||
| October 2005 | 7.0-7.5 | 9.2 | -2.0 | 5.0-5.5 | 3.9 | 1.4 |
| April 2006 | 7.5-8.0 | 9.7 | -2.0 | 5.0-5.5 | 6.6 | -1.4 |
| October 2006 | 8.0 | 9.7 | -1.7 | 5.0-5.5 | 6.6 | -1.4 |
| April 2007 | 8.5 | 9.0 | -0.5 | 5.0 | 7.5 | -2.5 |
| October 2007 | 8.5 | 9.0 | -0.5 | 5.0 | 7.5 | -2.5 |
| January 2008 | 8.5 | 9.0 | -0.5 | 5.0 | 7.5 | -2.5 |
| April 2008 | 8.0-8.5 | 6.7 | 1.6 | 5.5 | 0.7 | 4.8 |
| July 2008 | 8.0 | 6.7 | 1.3 | 5.0 | 0.7 | 4.3 |
| October 2008 | 7.5-8.0 | 6.7 | 1.1 | 5.0 | 0.7 | 4.3 |
| January 2009 | 7.5-8.0 | 6.7 | 1.1 | 3.0 | 0.7 | 2.3 |
| April 2009 | 7.0 | 4.0 | ||||
| July 2009 | 6.0 | 5.0 | ||||