Overvalued on earnings discount model

MARKET INSIGHT

If EPS growth lags in second half, be ready for massive downsides.

One of the most conservative equity valuation models is dividend discount. Here, the dividend yield is compared to a risk-free discount rate. The problem with dividend discount is that a lot of high-tech companies never pay dividend as a matter of principle.

The internal rate of return (IRR) is high enough to justify zero payouts. Take a look at Intel or Microsoft in its heyday and, if you're curious, imagine what Infosys' turnover would be if it had never paid dividend and earned its IRR consistently on that extra cash.

An earnings discount model is a compromise. It assumes that the investors' payout is equivalent to the earnings per share (EPS) and compares that yield with discount rate. A high-growth rate is expected to translate into capital gains.

In effect, this turns the Price -Earnings (PE) ratio upside down to calculate an earnings yield (E-Yield). The importance of rate movements becomes crystal-clear.

If the debt yield is low, the E-yield can be low. In other words, the PE can be high. If the debt yield is high or rising, the E-yield must be high or, the PE must be low.

Thus, a rising discount rate should be reflected in a lower PE. Of course, a lower PE doesn't necessarily mean a lower price if EPS is growing very fast.

And, any investor worth his salt is punting on future projections so, if EPS growth expectations are very high, the current PE could even rise in tandem with a rising discount rate. This last "" the rising prices/rising rates combo "" is exactly the situation in the Indian market at present and it has been the case over the past several years.

The logic is that investors have bet heavily on fast EPS growth since 2004-05 and ignored the rise in rates. Those expectations of high growth have so far, been justified.

If there is a stumble in EPS growth in the second half of 2006-07 or 2007-08, the market could be exposed to massive downsides. One reasonable proxy for a discount rate is the 364-day Treasury Bill.

The RBI holds fortnightly auctions of this instrument; it's risk-free; the cut-off yield changes to reflect rate expectations and liquidity. It moves in the same direction as the 91-Day T-Bill or money market instruments.

In June 2003, the 364-day T-Bill auctioned at implied yields of about 4.95 per cent. The Nifty was at 1045 points and the Nifty's average PE was at 12, with an implied E-Yield of about 8.4 per cent.

That was a buy on very conservative metrics even if we assumed that the Nifty's EPS would remain static. By June 2004, the T-Bill was at 4.6 per cent, the Nifty was at 1550 with a PE of 12.5 and the E-Yield at 8.02 per cent. Still a very clear buy. The equity price-trend went according to plan during this phase.

By June 2005, the yield had hit 5.6 per cent. The Nifty was at 2112 with a PE of 14 and an E-yield of 7.1 per cent. Still quite reasonable since the E-Yield is comfortably higher than the T-Bill yield

By June 2006, the T-Bill yield was over 7 per cent, the Nifty was at 2923 and the PE was 16.6 with an E-Yield of 6.02 per cent. No longer comfortable.

In early January 2007, the yield was around 7.2 per cent, the Nifty is at 3900 with a PE of 21 and an E-Yield of 4.75 per cent. The market is very overvalued on a earnings discount model.

Between 2003 (rates bottomed in late 2003 ) and now, rates have risen by almost 300 basis points from a bottom yield at 4.3 per cent.

The saving grace is that EPS rose by around 114 per cent (absolute) between March 2003 and 2006. That's a compounded rate of 28 per cent EPS growth and unless the current quarter is terrible, there won't be a big slowdown in 2006-7.

But what happens if EPS growth expectations are belied over 2007-08? The PE will then correct to "sustainable" levels of 14-15. A 35 per cent drop in valuation may translate into somewhat lower capital loss given that there will be some EPS growth. But a loss is a big possibility given the slightest slowdown.