» Business » Why India's ability to forecast is so poor

Why India's ability to forecast is so poor

By T N Ninan
December 27, 2008 14:00 IST
Get Rediff News in your Inbox:

In January 2008, inflation (as measured by the wholesale price index) was about 4 per cent. By August, it had climbed to a decadal high of 13 per cent. Now it is half that.

No year in recent memory has seen such sharp swings in price data. Specific prices underline the volatility even more: Oil rose during the year from $100 to $147 and then crashed to $35.

The stock price index has halved. Global food prices rose 50 per cent in the year to June, then the entire increase was wiped out in six months. Industrial commodity prices, which had been shooting up, are now half the level of a year ago.

It is not just prices. The job market has gone from feverish activity to a deathly stillness. Stock trading volumes have also moved into reverse gear, dropping by 40 per cent after going up 80 per cent in the last financial year.

Office rental values soared in 2007, then halved from the start of 2008 to the end. One could go on, but the point should be clear.

To be sure, 2008 has been an exceptional year, starting with a boom and ending with a bust -- from credit being abundant one day to a zero credit situation three months later, and from buoyant demand in the summer to dead markets and piled up stocks in the winter.

That is one reason why no one was prepared for either the speed of change or the sharpness in directional swing. As is obvious, all years will not be like this one which, mercifully, will come to a close in four days.

But the questions being asked today are no less important: How long will the downturn last? Will it get worse before it gets better? If, by general consensus, 2009 is going to be worse than 2008, what about 2010?

The question before those questions is whether our ability to peer into the future, poor through most of 2008, is any better today. The answer, unfortunately, is: No, it is not.

The government's statistical system is being overhauled with painful slowness. The base years for the indices are too far back in time, the weights for different sub-sets are out of date, the methods of data collection are defective, and the statistical methods used simplistic.

Without reliable numbers and up-to-date methods, everyone is indulging in guesswork about the true level of industrial output, what services sector growth actually is, and whether the crop data can be relied on.

We don't have a formal index of leading indicators, to give early warnings about which way the economy is headed.

Twenty years after economists in the finance ministry talked of using de-seasonalised numbers to get the true meaning of raw data, this technique is still to become a formal part of the official statistical system.

And the methods by which many forecasters work out their GDP predictions seem sometimes like a combination of adopting Feng Shui and reading tarot cards.

Everyone pays a price for this. If you watch the (slow) speed of government response to evolving trends, like interest rate changes or adjustments to money supply, or look at official growth forecasts as they evolve, it is evident that the response times are too slow.

Till as recently as late last month, the prime minister was talking of 8 per cent GDP growth this year. The Reserve Bank did not begin to shift its stance on monetary policy till mid-October. Such mistakes have serious consequences.

What the country needs as a New Year gift is a modern statistical system. But for reasons too complex to go into here, it will be some future New Year in which we will (hopefully) get one.

  • Global Meltdown & Impact: Complete Coverage
    • Get Rediff News in your Inbox:
      T N Ninan
      Source: source

      Moneywiz Live!