Technology stock valuations -- Driven by Old Economy variables

B. Venkatesh

Anup Menon

Despite nine recessions since World War-II, the stock market is up 63-fold since earnings are up 54-fold. Earnings drive the market. -- A Fidelity Investments ad quoted in Robert Shiller's book Irrational Exuberance.

IN THE recent boom in technology stocks, a perception that has gained ground is that these valuations are driven more by expectations than by current earnings. But, as it happens, the tech stock valuation is largely driven by the same, or similar, variables that govern stocks in the traditional sectors, or Old Economy, if you prefer.

This is based on two findings from a Business Line analysis of a sample of tech stocks. The first is that the market seems to be giving more importance to quarter-on-quarter (September 2000 over September 1999) financial performance growth rather than prior-quarter growth (September 2000 over June 2000) -- just the way the market values cyclical stocks. Second, the market seems to be using accounting information substantially to value tech stocks.

The price-earnings multiple (PEM) of the Business Line-250 Software Index is nearly 6 times the PEM of the BL-250 composite index. Does this mean the market is pricing technology stocks differently? Or, do the earnings growth of technology companies justify such high PEMs?

We sought answers in the accounting information easily available to retail investors, who constitute a significant segment in the equity market.

We started on the premise that the market is valuing tech stocks correctly. Again, the reason has to do with the retail investors. The alternative is to use the discounted cash-flow basis (DCF). This discounts the expected cash-flows of the company to the present, and compares that value with the current market price. If the discounted value is more or less equal to the current market price, it can be concluded that the high PEM for tech stocks is, indeed, justified.

The DCF approach was not adopted for two reasons: First, it requires estimating the future cash flows, which a retail investor cannot undertake on his own due to lack of information. And, second, it requires adopting discount rate, which is highly subjective.

Now, the assumption that the market is correctly valuing tech stocks means that the PEM is an important measure for an investment decision. And PEM is but a function of the market price and the earnings per share (EPS). We, therefore, set out to find whether the market valued tech stocks based on EPS and other related factors. Indeed, it did!

Simple relationship

We first tried to find out whether EPS is the only factor that determines the valuation of tech stocks. We found that about 57 per cent of the valuation of tech stocks can be explained by EPS. We took the average stock price of each company and ran a simple linear regression with EPS as an independent variable.

We faced a problem in arriving at the average price of each stock. This is because tech stocks were very volatile last year. We, therefore, decided to find the average price in the last 50, 100 and 200 days, and run a regression for each price. Our analysis indicates that the explanatory power of the model decreased gradually as we increased the number of days from 50 to 100 and to 200 days to find the average stock price.

We interpret the decreasing nature of the explanatory power of EPS thus: During early 2000, tech stocks traded at high PEMs due to heavy buying interest from mutual funds, financial institutional investors (FIIs) and retail investors. These stocks, however, since returned most of the gains.

The higher explanatory power of the 50-day average stock price may be due to the tech stocks trading now at more realistic levels in relation to their EPS. This is clearly reflected by the sharp fall in the PEMs of tech stocks since then. Infosys, for instance, now trades at a PEM of 169 against 268 last June.

But our study did not stop with this. We decided to delve deeper to discover if the market was using other accounting data to value tech stocks.

More accounting variables

We decided to test four variables available to retail investors, along with the EPS -- sales growth, operating profit growth, net profit growth and market capitalisation. We found that these factors, indeed, help the market value tech stocks.

For instance, taking the 50-day average stock price for the quarter-on-quarter model (September 2000 over September 1999), the EPS and accounting variables explained 77 per cent of the valuation of tech stocks. Against this, the prior-quarter model (September-over-June) using EPS and accounting variables explained only 66 per cent of the valuation of tech stocks.

The analysis indicates that the market places more emphasis on the quarter-on-quarter growth rates rather than the prior-quarter growth rates. This, indeed, came as a surprise; for it suggests that the market is pricing tech stocks the same way as Old Economy stocks.

For instance, a hotel company. Such a company typically performs well from October to March, which is the tourist season. The market, therefore, compares, say, the October-December quarter financials with the corresponding previous period to find out how well the company has performed in the current year. Specifically, it does not compare the October-December quarter performance with July-September quarter.

The technology industry, however, is not subject to such cyclical trends in a year. Why, then, was the market pricing companies that way? One reason could be that retail investors are strongly influenced by media reportage. And media, especially, the financial newspapers, report quarter-on-quarter growth for technology companies and not the prior-quarter growth.

But are retail investors influenced by media reports? We draw reference to the behavioural social scientist, Prof Robert Shiller's book Irrational Exuberance to stress that media can have a strong influence on retail investors and, therefore, the stock market.

Conclusion

While suggesting that accounting variables are important, our analysis shows that these variables fail to fully explain the valuation of tech stocks. This may be because the market is factoring other information to value such stocks.

A key variable, for instance, is the expected growth rates in earnings. In developed markets, such as the US, investors have been known to factor in the consensus estimates of futures growth rates to value stocks. This apart, information about projects on hand and expenditure incurred in research and development could also prove useful to them. It would be interesting to see whether these factors help fully explain the valuation of tech stocks.

Further, while we have stated the relative importance of accounting variables from an overall perspective, we have not attempted to prioritise each variable according to order of importance. In other words, our analysis does not indicate which variable has a higher level of predictive power on a standalone basis, compared to the others. Hence, further work can be done on reasearching the relative important of each variable.