Today’s article in Barron’s profiles the Matthews Japan Fund (MJFOX; investor share class). This $452 million fund has a sensible 1.20% expense ratio, no 12b-1 fees and a 23% turnover (although this figure, reported on the fund’s website, is calculated differently than the 49% turnover published by Morningstar®).
The article indicates that the fund has recently performed quite well, mostly as a result of Japan government’s stimulatory policy:
Over the past 12 months, Matthews Japan rose 22%, compared with 15.3% for the Japan-only funds category, according to Morningstar… For investors, much depends on whether Abenomics, the economic policies of Prime Minister Shinzo Abe, will be able to jolt Japan’s economy out of its seemingly endless slump… [Fund managers] see welcome signs in the positive turn in core inflation last year, as companies began to raise prices for the first time in years.
The fund invests mostly in large-cap stocks, although compared to its primary benchmark, it is tilted toward mid- and small-caps:
The 63-stock portfolio is a mix of big names—including auto makers Toyota Motor and Honda Motor and financial firms ORIX and Mitsubishi UFJ Financial Group —and mid- and small-company stocks that often fly under the radar. At year end, more than 40% of assets were invested in companies with a market value of $5 billion or less, versus just 12% for the MSCI Japan index.
The article compares the annualized one-, three- and five-year returns of Matthews Japan to those of the MSCI EAFE index, which is not appropriate for this single-country fund. A better measure of performance would be the Sharpe ratio, which is the simplest form of adjustment for risk. According to that measure, the fund beat a practical implementation of its primary benchmark, the iShares MSCI Japan ETF (EWJ) in all of three-, five-, ten- and fifteen-year periods through February 2014. The fund also generated better annualized returns than the ETF in those periods.
With that, let’s take a deeper look at the fund’s performance using Alpholio™’s methodology. For this analysis, instead of a diverse pool of reference ETFs, we chose only Japan-oriented ETFs, as well as one domestic ETF to model the fixed-income holdings of the fund. Since some of these reference ETFs only have a limited history, the overall analysis period has been reduced by a couple of years compared to our standard approach. Nevertheless, the results are still informative.
Here is the cumulative RealAlpha™ chart for the fund:
From early 2008 through 2012, the fund generated hardly any RealAlpha™. The cumulative RealAlpha™ started to increase dramatically only in the second quarter of 2013, but subsequently flattened out in the fourth quarter.
The lag cumulative RealAlpha™ curve was generally below the regular one. This indicates that in most analysis sub-periods, investors would be better off by sticking to the reference ETF portfolio rather than adjusting the positions to match the fund’s returns. In other words, not all new investment ideas of the fund panned out. (To learn more about the regular and lag RealAlpha™, please see our FAQ.)
Finally, at over 18%, the fund’s annualized volatility in the entire analysis period was higher by more than 2% than that of the reference portfolio.
The following chart shows the ETF weights in the reference portfolio for Matthews Japan in the same analysis period:
The fund had top equivalent equity positions in the iShares MSCI Japan ETF (EWJ; average weight of 38.6%), iShares Japan Large-Cap ETF (ITF; 14.2%), iShares MSCI Japan Small-Cap ETF (SCJ; 9.7%), SPDR® Russell/Nomura Small Cap™ Japan ETF (JSC; 8.6%), and WisdomTree Japan SmallCap Dividend Fund (DFJ; 8.5%).
The cash and short-term investment position of the fund was modeled by the iShares TIPS Bond ETF (TIP; average weight of 15.9%). The Other component of the chart includes two additional Japan stock ETFs with smaller average weights.
In summary, although the Matthews Japan Fund beat its primary benchmark on a risk-adjusted basis, it generated a mostly flat cumulative RealAlpha™ in the last six years. While the fund significantly outperformed its reference ETF portfolio in the two middle quarters of 2013, there is no guarantee that this will continue in the future. As a matter of fact, the cumulative RealAlpha™ of the fund has been slightly declining since last September. Therefore, investors may want to consider using a dynamic portfolio of ETFs instead.
To learn more about Matthews Japan and other mutual funds, please register on our website.