Money Find/Feedback/Site Index
March 22, 2000

Indian Tax
Mutual Funds
Personal Finance
Personal Banking
Real Estate

E-Mail this report to a friend
Devangshu Datta

Dot com fever

The elasticity of Internet share valuations appears to be in direct proportion to the surrealism of the rationales for those valuations. The original explanation for the high valuation of dot coms was that the Internet was a fast growth new industry, so it deserved higher valuations of future cash flows.

Quite so, but how much higher should these valuations be? The usual rule of thumb for fast growth industries is that price-earnings ratios should not be more than twice the percentage rate of earnings per share (EPS) growth. That is, a company with an earnings growth at 100 per cent per annum can, at best, be priced around 200 times earnings.

This model naturally pre-supposes a positive earnings record. The reality in the Internet business is that the average Internet company has only losses and the profitable ones trade at price-earning multiples (P/Es) that are often ten times their EPS growth rates.

The original explanation wore thin when, after several years of rocketing share prices and mega-deals in the dot com sector, there was still little sign of profit. The next rationale was that dot coms must actually be evaluated on the basis of sales growth rates rather than earnings growth rates.

After all, they may not show profits but most Internet companies do show triple-figure revenue growth rates. Sooner or later, they will leverage their first-mover advantage into a hammerlock on the market and start making money hand over fist. Or so this argument goes.

There are several problems with this valuation model. One is that price to revenue multiples to revenue growth rates are hugely skewed almost as much as price-earning to EPS multiples are. There is the old, value-investing caveat about a business that grows too fast by borrowing expensive money. In the Internet era, it requires modification because the money burned comes from IPOs and stock swaps. But in the long run, equity is expensive and it's a permanent charge.

Another problem is that the first-mover advantage itself may be over-rated. There are few bars to competition on the Net and a newcomer may well find a winning formula. Take the example of Amazon versus CDnow. CDnow is the newcomer in the retail book and music market and it draws between half and two-thirds as much traffic as Amazon. But recent sales and profitability numbers suggest that CDnow actually sells more, and at better margins than Amazon.

Again, to extend the example, bookstore chain Barnes & Noble came late on the Net but it has grabbed revenues and market share on the basis of its established brand name and its ability to deliver quicker than Amazon. There will be more such click-and-mortar players in every industry. Yahoo! was revolutionary with its search engine driven Customer Relationship Model and its electronic marketplaces. But it now faces competition from some 1,500-odd engines and perhaps ten of those are already serious rivals.

The latest fad in Internet valuations is that they ought to be justified by discounted cash flow (DCF) analysis. DCF does have several advantages over more traditional valuation models in that it ignores tax and accounting-entry conventions and it allows for analysis over longer timeframes than P/E or Revenue-multiple models.

But there is a basic stumbling block to applying DCF analysis to dot coms. How do you discount cash flow without cash? To make DCF "work", one ingenious method is to take a probabilistic approach. Here the analyst assumes several possible scenarios over say, a ten-year timeframe.

In each scenario, the dot com in question is assumed to reach a certain state of sustainability. By comparison with brick-and-mortar businesses in similar industries, a range of cash flows is derived for the dot com. Now you can do DCF across the range, use a standard discount rate, and see whether the share valuations start making sense.

For example, a leading dot com supermarket would be compared to Wal-Mart and assumed to possess similar market share and margins after a decade. If the dot com is assumed to be less successful, a second lower market share-margin set will be derived. And so on with assigned probabilities of occurrence to each such DCF scenario. One key to this process is that coherent projections can be made about the industry growth rate for supermarkets.

Consulting firm McKinsey adopted this approach in a case study of Amazon sometime in the last quarter of 1999. According to its projections, the current market value of Amazon would be justifiable only if it became an absolute market leader. In that extremely positive scenario, it would still be reckoned fully valued on a ten-year basis. In every other circumstance, (four cases were analysed in detail), it was over-valued and by large amounts. In no case, was it currently under-valued according to the study.

Assuming that similar models hold for other dot coms, there is little money to be made at current prices even by a very long-term investor. Perhaps, the recent correction in US Internet stock prices is a belated realisation of this possibility. Internet prices are still high but the bubble may now be close to deflation. Or, it could be that somebody will soon find another rationale for Internet share prices staying sky-high.

What is disturbing is that Internet fever is now catching on here even as it seems to be abating in its homeland. There could soon be an Indian bubble of unprecedented dimensions. There are four pre-conditions that the Indian dot com industry needs for financial take-off.

The first two are the rationalising of the tax-treatment of Indian employee stock option plans (ESOPs) and the relaxation of the SEBI guideline for a three-year profitability record before a company can go public. Since the entire IT and pharma industries need these breaks, these two practices will be cleared soon.

The other two pre-conditions are quintessentially American practices, which have helped ramp up prices and hype on the NASDAQ. One is already common here while the other is likely to arrive seen. The first is the process of bookbuilding for an initial public offer (IPO). This includes an opaque and discriminatory allotment of shares, a process that creates an artificial scarcity to boost demand at the time of listing. Momentum trading then further hikes prices.

However, 89 per cent of American dot coms trade below their IPO price, a year after the IPO. The bookbuilding process has already been adopted in the recent past by several Indian IT companies such as Hughes Software and HCL Technologies.

The second practice is a side effect of compensating employees with ESOPs. It is normal for an American sweat equity holder to buy a collar on his stock options. A collar is a two-way derivative that combines a put (sell) at one price and a call (buy) at another price-perfect protection at a small premium. The clearance for Indian derivatives trading came through with this year's Budget. Assuming that ESOP tax treatment is rationalised, I foresee a great demand for collars.

Devangshu Datta

Tell us what you think of this column