The iPod Effect

Apple Computer introduced the iPod in October of 2001 and has not looked back since. Today, the iPod is nearly ubiquitous and arguably the most popular digital music player in the market today, if not the most preferred. And in its wake, Apple's prospects have never looked better. Indeed, if we assume that financial markets are reasonably able to predict a company's value – reflected in its stock price - then it appears that the outlook on Apple's future is rosy indeed. When the company went public in 1984, its stock traded at $3.00 a share; today, it trades at around $60.00 a share.

This raises an interesting question: has Apple Computer's stock price been affected by the introduction of the iPod, and if so to what extent? In the mainstream business media, it is commonplace to associate the fluctuations in a company's stock price to such things as changes in sales data, management decisions or the introduction of new products. Easy explanations, but not necessarily correct or empirically verifiable. Thus it is compelling to ask: is there an iPod effect?

I thought up the above as a topic for my applied econometrics course. It seemed like an interesting question to pose as it entails econometric analysis from the standpoint of a finance framework, particularly that of index models. Assuming that stocks are sensitive to fluctuations in the overall market, and to the extent that a stock market index (like the NASDAQ or S&P500) is a reasonable proxy for the market's performance, then it should be possible to estimate whether company-specific factors, such as the introduction of a revolutionary new product, do in fact affect that company's stock price.

Using the average closing price of Apple Computer stock (AAPL) from September 7, 1984 to April 29, 2006, as well as the average closing value of the NASDAQ composite index (NASDAQ) over the same period, and introducing IPOD as a dummy variable, I set up the following regression model to explain the variation in the company's stock price:

AAPL = &beta₀ + &beta₁(NASDAQ) +&beta₂(IPOD)

Each of the beta coefficients merely reflect the magnitude of the impact that changes in each independent variable have on the company's stock price, while the rationale behind index models tells us that the intercept term specifically represents the implied "expected value" of Apple shares (i.e. how much the stock would be worth holding all other factors equal). In this lies the key to the analysis. Since IPOD is a dummy variable that takes on a value of one (1) in those months where Apple sold iPods and zero (0) if otherwise, if its beta coefficient is statistically significant it will be added (or deducted) from the expected value of Apple Computer stock in "iPod months". As a corollary, in those months where Apple Computer did not sell iPods, the parameter will reduce to zero. Hence, we will be able quantify the effect of the iPod on Apple Computer stock prices if such exists.

The results: corrected for autocorrelation, the overall model is statistically significant at any reasonable level and explains approximately 35% of the variation in Apple Computer's stock price. Further, each of the parameter estimates of the beta coefficients are similarly statistically significant at any reasonable level, with the following values:

AAPL = 3.8866 + 0.005870(NASDAQ) + 11.4109(IPOD)

This tells us that prior to the iPod's introduction, the market priced Apple Computer stock at around $3.89 controlling for macro-level factors (fluctuations in the overall market, as captured by the index), whereas since the iPod was introduced the expected value of the company's average closing share price appears to have increased to approximately $15.30 (i.e. 3.89 + 11.41). Indeed, it would appear that there is in fact an iPod effect, in the amount of approximately $11.41.

Is this good finance, or good economics? I leave it for you to decide. But I will say this much: if I had the cash to burn, I'd buy Apple Computer shares. Or a new iPod.