Detecting and Explaining NASDAQ Bubbles: Evidence from GSADF Tests and a DSGE–Asset Pricing Model

Abstract This paper provides a timely detection method of an emerging bubble in the NASDAQ Composite Index using the Generalized Supremum Augmented Dickey-Fuller test. The detected bubbles are then analyzed with the help of a Dynamic Stochastic General Equilibrium–Asset Pricing (AP) model, which combines economic fundamentals and financial dynamics. Through a comprehensive analysis of booms and busts of the NASDAQ Composite Index over the past 50 years, particularly the in-depth analysis of the Dotcom Bubble around 2000, the Housing Bubble prior to the 2008–09 Global Financial Crisis, and the current bubble after 2020, we have formulated and confirmed the Connectedness Hypothesis (tight links between economic fundamentals and stock dynamics) and the Overshooting Hypothesis (excessive adjustments in share prices and returns). Three drivers of bubbles are identified: (i) unusual market sentiments (stock return shocks), (ii) unanticipated technological hype (technological progress shocks), and (iii) unexpected credit expansions (monetary policy shocks). The current bubble, which burst in 2022 but re-emerged after the AI boom in 2023, appears to be driven by all three factors, making the index easy to expand but also more vulnerable to downside risks or shocks.