Devangshu Datta:Inverse relationship

There is a rough and ready method of comparing price-earnings ratios to interest rates. Invert PE ratios and normalise to a percentage. If it's close to the interest rate, the stock is fairly priced. If it's higher, the stock is under-valued. Conversely, the stock is over-valued if it's lower.

For example, a PE of 15 can be expressed acccording to this formula as (1/15 x100) or the equivalent of 6.67 per cent. A PE of 20 will work out to an equivalent of an interest rate of 5 per cent. It's an easy calculation to make.

The logic is as simplistic as the calculation is simple. An investor receives earnings from a share, which is equivalent to the interest received from a loan. If the interest rate is up, returns from equity must also rise "" hence the PE must fall.

There are several obvious drawbacks to fudging like this. Apart from vagaries of compounding, there is the plain and simple fact that earnings are not received in toto. No company pays out all profits as dividends. In fact, few investors would be happy with a company, which distributed all its profits.

This inversion also ignores the unpredictability of earnings versus the certainty of interest. Earnings could grow quickly or deteriorate equally fast and the calculation implicitly suggests that neither of these will happen.

Despite the flaws, the calculation does work reasonably as a first approximation of valuation. If the inverted PE is much lower than the interest rate, there has to be a very compelling reason (usually very high growth) to make a stock purchase justifiable.

Right now, the Sensex is available at a PE of just over 19 (with a dividend yield of 1.7 per cent and a nominal value of 5885 points). The invertion-calculation suggests that this is equivalent to an interest rate of about 5 per cent to 6 per cent. That's high enough to make one wonder if stocks are worth buying. The debt market is offering returns some 150-200 basis points higher.

The devil lies in the details, as always. This average is being calculated on a trailing PE and doesn't reflect the likely 20 per cent to 25 per cent earnings growth in 2003-04. If we factor that growth in, the Sensex is actually trading at a current average PE of around 15.

Markets always discount future rather than current earnings. Again, factoring a likely 20 per cent earnings growth for 2004-05, the Sensex is trading at a projected PE of about 12. Those numbers translate to interest equivalents of 6.67 per cent for 2003-04 PE and 8.3 per cent for 2004-05 PE, respectively.

Projected returns of 6.7 per cent to 8.5 per cent would just about balance off current rates. A conservative would still reckon the market is a little too highly-priced for comfort. It will probably make this fiscal's estimates but next year's earnings gains will have to come on a high base.

It will require an amazing corporate performance for India's 30 largest companies to log better than 20 per cent EPS growth for two years in succession. It's happened before but only during the Manmohan Singh period when good policy was backed by a sequence of excellent monsoons.