What Factors Drive Global Stock Returns?

Outline Data and Summary Statistics What Factors Explain Global Stock Returns? Cross-sectional tests with individual stocks Time-series regression tests with country, industry, and characteristic-sorted portfolios Country-Specific or Global Factors?

Data and Summary Statistics The sample construction begins with all publicly- traded firms provided by Datastream from July 1981 to December 2003. Select those stocks with sufficient information to calculate B/M, C/P, D/P, E/P, L/B, and market value of equity. Then select common stocks that are traded on the country’s major exchange(s). Apply several screening procedures for monthly returns.

Data and Summary Statistics Match the year-end financial statement data for year t-1 with monthly returns from July of year t to June of year t+1. The final sample encompasses 27,488 common stocks from 49 countries and 34 industries. Figure 1 exhibits the distribution of sample stocks across the 49 countries. Figure 2 shows the development of the sample over time. Table 1 presents the summary statistics of monthly returns and other firm characteristics for the final sample.

Data and Summary Statistics For some tests, betas with respect to the value- weighted global and country-portfolios are employed. These betas are estimated annually for each stock at the end of June of each year, using the stock’s previous 36 monthly returns (12-month minimum).

What factors explain global returns? The experiment involves two types of asset- pricing tests. The first test employ the cross-sectional regression approach proposed by Fama and ManBeth (1973) using individual stocks. The second test adopts the time-series regression approach, in which returns on country, industry, and characteristic-sorted test portfolios are regressed on returns of various factor- mimicking portfolios. Cross-sectional tests with individual stocks

What factors explain global returns? Table 2 reports the time-series average of the slope coefficients (with t-statistics) from monthly Fama- MacBeth(FM) regressions of individual stock returns on betas and other firm-level characteristics. 1.Results are reported for “univariate” regression involving only one independent variable per regression model and “multivariate” regressions. For C/P, D/P, E/P, and L/B, dummy variables are used (follow FF(1992)) to separate firms with negative cash flows, no dividend, negative earnings, and no leverage.

What factors explain global returns? 2. Panel A represents the regression results across all stocks from all countries. The univariate FM regressions show that global and country betas do not explain the cross-section of average stock returns. In contrast, most other firm-level characteristics show reliable explanatory power: Size, B/M, Mom, C(+)/P, D(+)/P, and E(+)/P. In the mutivariate regression (beta and leverage ratio are not included), the slope coefficients for size, B/M,

What factors explain global returns? Mom, C(+)/P, D(+)/P, and E(+)/P remains significant. 3. Panel B presents the results for the US only. The 9,840 US stocks constitute more than one-third of the final global sample (figure 1). The univariate regression results run in parallel to those for all countries (panel A). The D(+)/P coefficient is much smaller and is insignificant. In the multivariate regressions, the coefficients on size, Mom, C(+)/P, and B/M remain significant, while those on E(+)/P and D(+)/P becomes insignificant.

What factors explain global returns? 4. Panel C present the results for developed (excluding US), and emerging markets. The results obtained from all countries largely hold for stocks from development markets outside US, with the exception of the size and E/P effects. For emerging markets, only C(+)/P retains a significant slope coefficient. The C/P, D/P, and E/P effect is stronger in the second half of the sample period (B/M effect is opposite). Size effect is concentrated in January, whereas the momentum effect reverse in January (-2.45% vs. 1.29%).

What factors explain global returns? 5.The cross-sectional tests indicate that alternative measures of the value-growth effect are not interchangeable, and the selection and identification of the numerators of these “inverse-price” ratios is important. Constructing the factor-mimicking portfolios In order to further explore the characteristics that best account for the variation in global stock returns, this paper construct proxy factors as returns on zero- investment portfolios that go long in stocks with high values for a certain characteristic (such as B/M) and short in stocks with low values for that characteristic.

What factors explain global returns? An FMP (factor mimicking portfolio) is constructed for each firm-level characteristic and the summary statistics of the FMP is provided in Table 3. For each of the characteristics, global quintiles portfolios are formed at the end of June of each year t and calculate the value-weighted returns of each portfolio from July of year t to June of year t+1. The FMP return then is computed as the highest-quintiles returns minus the lowest-quintile returns (except for size FMP returns).

What factors explain global returns? The momentum FMP is calculated following Jegadeesh and Titman’s (1993) six-month/ six-month strategy. Average premium: Of the various FMPs, the market portfolio achieves an average excess returns f 0.49% per month, and it is only marginally different from zero. The E/P, C/P, D/P, and momentum FMPs achieve high average monthly returns, each with t- statistic greater than two. Volatility and cross correlation

What factors explain global returns? Time-series regression tests with country, industry, and characteristic-sorted test portfolios In time-series regression approach, excess returns on test portfolios are regressed on returns of various FMPs. 1. The test portfolios include country portfolios (20 countries), global industry portfolios (34), and decile portfolios based on each of the firm-level characteristics (size, B/M, momentum, C/P, D/P, and E/P).

What factors explain global returns? 2. Table 4 reports the time-series regression results. The global CAPM is used as a starting point for each test portfolio, and then add various FMPs to the global CAPM. 3. Panel A of Table 4 The first two columns give the raw excess return differences between the highest and lowest-return country, between the highest and lowest-return industry, and between the extreme decile portfolios for each characteristic (“H-L Ret”) and the average absolute returns of the test portfolios.

What factors explain global returns? The global CAPM results are shown in column 3~6. According to the GRS test, in each experiment-except for the country portfolio-the global CAPM model is easily rejected. The remaining columns evaluate the global version of the Fama-French model. The addition of the size and B/M FMPs leads to one fewer model rejection and lower F- statistics for some experiments. 4. Panel B of Table 4 Panel B presents the results of the two-factor model (eq. (3)).

What factors explain global returns? A comparison of the various two-factor model’s performances for the entire range of test portfolios reveals that the model with the market portfolio plus C/P FMP produces the fewest number of rejections and the lowest pricing error (absolute intercepts). Models with the market and the D/P or E/P FMPs are rejected by all test portfolios except B/M portfolios. The model with the market and B/M FMP is rejected by all except for B/M and country test portfolios. Therefore, different value-related FMPs are not easily substitutable.

What factors explain global returns? 5. Panel C of Table 4 From the test results of panel B, the C/P FMP shows promise in terms of pricing test portfolios based on other value-based characteristics, whereas other value-related FMPs do not. The momentum FMP is the only FMP that can explain momentum test portfolio. Consequently, the authors introduce a new three-factor model (“HKK”) that include C/P FMP and momentum FMP in addition to the market portfolio (eq.4).

What factors explain global returns? Compared to the global CAPM and FF three-factor model reported in panel A, this model offers a significant improvement for industry, momentum, C/P, D/P, and E/P test portfolios. Furthermore, it is no worse for country and B/M test portfolios. In none of these seven experiments is the model rejected by the GRS F-statistics. This new three-factor model also show a significant improvement in performance relative the various two-factor models investigated in panel B.

What factors explain global returns? Two final tests are performed in panel C. First, the replacement of C/P FMP with the other value-related FMPs lead to worse performance. Second, a composite five-factor model that nests both the HKK model and the Fama-French model is no better than the more parsimonious HKK model.

Country-specific or global factors? The question of whether markets are locally segmented or globally integrated has been one of the most enduring issue in international asset pricing. The authors compare the relative performance of global, local, and international (including local and foreign components) versions of different multifactor models that combine various FMPs in each country using industry and characteristic-sorted quintile test portfolios (7 sets).

Country-specific or global factors? For each country and each set of test portfolios, the regression models are shown in eq. (5a)~(5c) (CAPM ), eq. (6a)~(6c) (FF three-factor model), and eq. (7a)~(7c) (HKK model). Table 5 reports the results. In summary, the experiments indicate that that local and international version of the models typically outperform the purely global versions in most countries. The international version of the HKK model achieves the lowest rejection rate and pricing error, and rank near the top in terms of model explanatory power of all the different models this paper examine.