Returns and Return Premia of Size and Investment Portfolios in Japan—A Conspectus

This article conducts an overview of the performance of Japanese firm size- and firms’ investment-sorted stock portfolios from 1990 to 2020, and we derive the following contributions. First, we find that in our second half sub-period, the size effect is much clearer; while overall, the effect of investment is not so clear, suggesting that the portfolio constructions by firms’ investment are not so effective in Japan. Second, as we analyzed the performance of Japanese size- and investment-sorted portfolios using the data, which are in US dollars and for almost 30 years, our findings should be highly meaningful for both industrial practitioners and academic researchers, much deepening our understanding of stock portfolio returns and return premia in Japan.

find that in our second half sub-period, the size effect is much clearer in Japan; while overall, the effect of investment is not so clear, indicating that the portfolio constructions sorting by firms' investment are not so effectual in Japan. Second, as we examined the performance of size-and investment-sorted Japanese stock portfolios using the data, which are in US dollars and for almost 30 years, our key findings presented in this article should be highly meaningful for both industrial practitioners and academic researchers.
The rest of this article is organized as below. First, Section 2 is a brief related literature review; Section 3 explains our data and research design; Section 4 exhibits our key results; and Section 5 concludes the article.

Short Literature Review
This section conducts a short review of related literature with our current study.
First, Fama and French (1993) developed a representative three-factor empirical asset pricing model, adding size and value factors to the capital asset pricing model. Later, Fama and French (1995) also empirically examined the size and value effects in the US stock market. Afterwards, Fama and French (2015) extended their three-factor model to a five-factor asset pricing model, adding operating profitability and firms' investment factors to their three-factor model. (2017)  In contrast, for Japan, using the Japanese data from 1981 to 2005, Tsuji (2012) explored the size-and book-to-market ratio-sorted stock portfolios; and Tsuji (2020) also provided an overview of the size and value effects in Japan for the period from 1990 to 2020. However, we note that the analysis of the firms' investment, which is a new element for the Fama-French five-factor model, is limited particularly for the Japanese stock market.

Recently, Fama and French
As these backgrounds of the past studies clearly show, providing an overview of the performance of Japanese stock portfolios by focusing on the firms' investment effect shall be significantly beneficial for both academic researchers and industry practitioners. Therefore, based on this, in the following sections, this article attempts to conduct an effective overview of the return and return premia of Japanese stock portfolios sorted not only by size but also by firms' investment.    . Furthermore, we also analyze the latest sub-period though not a five-year period (January 2020 through to October 2020).
In Figure 1, the dynamic evolutions of cumulative raw returns as regards the above six Japanese size and investment portfolios and that of the Japanese market portfolio are displayed. In Panels A, B, and C of Figure 1, those for our full period, our first half period, and our second half period are exhibited, respectively (Stock prices are 100 in July 1990 (Panels A and B) and 100 in January 2006 (Panel C)). Table 1 exhibits key statistics for our full, first half, and second half periods, and Table 2 provides those for our seven sub-periods. Our key results from Table 1 are as follows. 1) For our full period, the highest return is seen in the Small-Low

Conclusion
This article has conducted an overview of the performance of firm size-and firms' investment-sorted stock portfolios in Japan. Using the data in US dollars, we derived key statistics and measures for many sample periods; thus our analysis should be significantly beneficial both for international investors and academic researchers.
As a result of our analysis, the small-size effect was generally clear, and in our second half sub-period, the size effect was particularly clearer; while overall, the effect of firms' investment was not so clear. This suggests that the portfolio constructions sorting by firms' investment are not so effective in Japan. Moreover, as we analyzed the performance of these portfolios by using the data for almost 30 years, our overview presented in this article should be highly informative.
Furthermore, we used the data from Fama and French and in Fama and French (2015), the investment-sorted portfolios are constructed by focusing on the change in total assets from the fiscal year ending in the year before last to the fiscal year ending in the last year (divided by the total assets of the fiscal year ending in the year before last). The matter of portfolio selection is continuously a hot topic as such recent studies as Chen et al. (2017Chen et al. ( , 2021 indicate, and we can use alternative definitions for firms' investment; hence further research of equity portfolios by using some alternative modified measures may be one of our future tasks.