Informational efficiency in theTokyoStockExchange, 1931–40

Abstract

This article relies on a unique dataset of daily price indices for stocks and bonds to analyse the functioning of the Tokyo Stock Exchange (TSE) in the period 1931-40. We find that this market deviated from weak-form efficiency, in a context of crossmarket segmentation, short-run spillovers, and turmoil surrounding major events. In this context, zaibatsu insiders were able to make abnormal returns via informed trading, while other uninformed investors could rely on technical rules to make abnormal profits. Such findings call for a micro-level analysis of the interwar TSE corporate financing function.

Measuring the Time-Varying Market Efficiency in the Prewar Japanese Stock Market

“…Then, Fumimaro Konoe's cabinet established the "National Mobilization Law" in April 1938 to severely suppress price volatility in various markets. This study finds that the 1930s Japanese stock market was almost efficient, but it contradicts the finding of Bassino and Lagoarde-Segot (2015). This difference can be attributed to the dataset and the empirical method used in Bassino and Lagoarde-Segot (2015).…”

“…This study finds that the 1930s Japanese stock market was almost efficient, but it contradicts the finding of Bassino and Lagoarde-Segot (2015). This difference can be attributed to the dataset and the empirical method used in Bassino and Lagoarde-Segot (2015). Their dataset contains periods of high-level price volatility, such as the Great Depression and WWII.…”

“…Their dataset contains periods of high-level price volatility, such as the Great Depression and WWII. In fact, the volatility of the stock return was too intensive, as shown in Figure .1 of Bassino and Lagoarde-Segot (2015). They also estimate a time-invariant GARCH-in-mean model to investigate information efficiency using the aforementioned periods' dataset; it is obvious that market efficiency tends to be inefficient when a high-level price volatility dataset is used.…”

“…2 He concludes that market efficiency declined after the start of the Pacific War and that the breakpoints are consistent with the variation of stock prices. Bassino and Lagoarde-Segot (2015) use the daily stock prices from 1931 through 1940 when applying Engle et al's (1987) generalized AR conditional heteroskedasticity (GARCH)-in-mean model to investigate information efficiency in the prewar Japanese stock market. They find that the 1930s Japanese stock market deviated from weak-form efficiency.…”

“…In practice, we estimate the time-varying degree of market efficiency using the 1930s Japanese stock market monthly data similar toBassino and Lagoarde-Segot (2015) and find that, during most parts of the 1930s, Japanese stock market was statistically inefficient. See the online appendix for more details, available at https://at-noda.com/appendix/prewar_stock_ appendix.pdf…”