A Multi-Century Perspective on Return Predictability and Price Bubbles

The detection of stock price bubbles has remained contentious in the economics literature. The focus has been primarily on 20th century US data (Shiller Database), which may have led to sample selection bias in econometric... [ view full abstract ]

The detection of stock price bubbles has remained contentious in the economics literature. The focus has been primarily on 20th century US data (Shiller Database), which may have led to sample selection bias in econometric inference. This paper uses a longer sample period for a wider set of stock markets, examining the historical stock data from the year 1815 to 2015 for the United Kingdom, France and the US to identify bubbles over complete period and various sub-periods. We build on the stylized model of West (1987) for return predictability in the presence of possible price bubbles. We use this model to make inferences on the presence of a bubble component in stock prices over the full sample and various sub-samples.