Monthly Discussion

The NASDAQ as an Ecosystem

There have been many requests for me to try the ecosystem approach on the NASDAQ. My hesitation so far has been due to the fact that the NASDAQ is not as well-defined a "species" as the Dow. The reasons are that the NASDAQ is a relatively young market and still in the process of frequent mutation (too many companies coming and going too often). Moreover, it has less of a culture, less loyalty among its investors, and less consistent behavior than the Dow. But with all these reasons weakening as time goes on, I will attempt here to treat the NASDAQ as an ecosystem with data up to the end of November.

To begin we must once again turn our attention away from prices and toward competitive variables, such as the share volume and the dollar value exchanged over NASDAQ stocks. Exhibits 3(a) and 3(b) show the evolution of these quantities respectively over the last ten years. The purple lines are natural-growth fits (S-curve) on the data.

Exhibit 3(a). The share volume daily exchanged over NASDAQ stocks. The data are reported monthly. The purple line is an S-curve fit.

Exhibit 3(b). The dollar value daily exchanged over NASDAQ stocks. The data are reported monthly. The purple line is an S-curve fit.

Despite spikes around the spring of 1999, the overall trend in both variables is rising and amenable to a description by S-curves (at least the early part of S-curves). However, one may want to see more of a step pattern here, reaching a ceiling in early 1999, particularly in Exhibit 3(b). But a low-ceiling S-curve cannot fit well the whole historical range. If we are talking about a natural growth process since 1990, then the purple line is the best fit. Of course, it could be that the NASDAQ underwent a major mutation and became a totally different species in early 1999. In that case this analysis would not be valid and my hesitation justified. But I don't see any compelling reason to expect such an important mutation of the NASDAQ in early 1999, so I will pursue this approach.

Diving dollar value by share volume gives the average NASDAQ price, both for the data and the S-curves. Thus we obtain a forecast for the price, see Exhibit 4.

Exhibit 4. The average NASDAQ price and a forecast obtained by dividing the curves of Exhibit 3(b) by those of Exhibit 3(a). The short blue line is the Composite index during 2000, (read on the right axes.)

Because of the way it is calculated here the average price is weighted by the share volume. This price is correlated to the simple arithmetic-average price defining the composite index. But the correlation is not as strong as in the case of the DJIA. The correlation coefficient here is 0.8 implying that only 64% of what we see in the Composite index can be explained by what we see on the weighted-average price. We can visually appreciate this correlation in Exhibit 5, in which we also see the long-term forecast for the NASDAQ Composite index.

Exhibit 5. The weighted-average price (purple line) is correlated to the Composite index (green line). The long-term forecast (yellow line) is the same as in Exhibit 4 and forecasts the price. The data points are daily.

Despite NASDAQ's declining trend during the last three months, the long-term forecast points the other way. According to this forecast the NASDAQ composite index was 36% below the level it should be at the end of November. Unlike the DJIA's stagnating long-term trend (see Exhibit 1), NASDAQ's long-term future promises growth. By the end of 2001 it should be around 5,000. The only worry about this forecast is whether NASDAQ will continue behaving like a species reliably; as the Dow does.