Signalling the Dotcom bubble: A multiple changes in persistence approach

This study investigates multiple changes in persistence in the dividend–price and price–earnings ratio of the NASDAQ Composite Index. Recent time series methods that are capable of signalling and dating asset price bubbles are employed, in particular the method developed by Leybourne, Kim, and Taylor (2007). The method allows for breaks between periods in which the data are integrated of order zero I(0) and integrated of order one I(1). The results confirm the existence of the so-called Dotcom bubble with its start and end dates. Furthermore, an unexpected negative bubble was also identified, extending from the beginning of the 1970s to the beginning of the 1990s, suggesting that the NASDAQ stock prices were below their fundamental values as indicated by their dividend yields, finding not previously reported in the literature. As the tools used by regulators take considerable time to take effect, methods capable of picking up warnings signals of the start of a bubble could be very useful. We conjecture that the methodology can also be applied to study recent phenomena in real estate, commodity and foreign exchange markets.