臺大管理論叢 NTU Management Review VOL.29 NO.1

3 NTU Management Review Vol. 29 No. 1 Apr. 2019 Pacific markets and test whether this premium is different than it is in the rest of the world. We confirm the finding in Amihud et al. (2015) that the illiquidity premium is significantly positive across 45 markets for the extended sample period. The monthly illiquidity premium for each market is adjusted for exposure to three global and three regional factors comprising of market, size, and value factors. The monthly risk-adjusted global illiquidity premium – the intercept from this regression, denoted alpha – is economically significant at 0.85% ( t -statistic = 7.53), using market indexes which constitute return-weighted average of the returns of the individual stocks in each market. The monthly premium is also larger (smaller) in emerging (developed) markets at 1.11% (0.66%) with a t -value of 7.97 (4.16). These numbers are comparable to those reported by Amihud et al. (2015). 2 For the Asia-Pacific region, which includes 16 markets in our sample, the risk- adjusted return-weighted premium for illiquidity is highly significant at 1.05% per month ( t = 6.05). The statistical and economic significance of the illiquidity premium is robust to alternative weighting methods such as volume weighting or market capitalization value weighting stocks within each market. After controlling for exposure to the six global and regional risk factors, the monthly illiquidity premium of the Asia-Pacific markets using volume-weighted (value-weighted) indexes is 0.92% (0.52%) with t = 4.19 ( t = 3.59, respectively). The median estimates of the risk-adjusted illiquidity premium and non- parametric tests for significance support the robustness of the estimates. For example, the return-weighted risk-adjusted illiquidity premium is positive in 94% of the Asia-Pacific markets, and the proportion of markets with positive risk-adjusted premium is significantly higher than the chance result of 50%. Yet we find that the risk-adjusted illiquidity premium in the Asia-pacific market is not significantly different from the premium in the rest of the world. We also employ an alternative approach to measuring the country specific premium for illiquidity using a cross-sectional regression framework. Specifically, for each market we employ the Fama and MacBeth (1973) method, regressing monthly stock returns on stock illiquidity after controlling for several firm characteristics that affect stock returns including size, book-to-market, volatility, and past returns. The risk premium is the mean 2 Amihud et al. (2015) find that the monthly illiquidity premium, adjusted for exposure to six global and regional factors, during the sample period from 1990 to 2011, is 0.82% ( t = 7.07), 1.16% ( t = 7.55), and 0.57% ( t = 3.79) in the global, emerging, and developed markets, respectively.