I’m often asked whether it is sufficient for a Singapore-based investor to hold the STI instead of the world portfolio. After all, the reasoning goes, “the biggest SGX-listed companies have extensive businesses all over the world”.
My response is two-fold. First, many investors may not know that although there are 30 stocks in the STI, six of these 30 firms account for more than half the index’s market capitalization. These six companies are: Jardine Strategic, Jardine Matheson, Singapore Telecoms, and the three local banks. It is true that these firms do have meaningful businesses outside Singapore. Nonetheless, these businesses do not span the globe. The two Jardines for example are anchored in Greater China. So is the overseas business of DBS, Singapore’s largest bank. Even if you include other big firms like Keppel Corp, Sembawang Corp and SIA, you don’t get a globe’s eye view of businesses as far as the STI is concerned. What this implies is that the STI may not be as plugged into the world economy as investors might think.
Second, for more concrete evidence, study the data. Run a regression of the STI’s returns against the returns of the world portfolio. If the STI is globally diversified, we should isee ts returns move tightly in step with the returns of the world portfolio. In other words, the STI should have a world beta coefficient of around one relative to the world index. To see if this is the case, I downloaded weekly returns for the STI and the MSCI all-country index from Thomson Reuters Datastream for the 5-year period from November 2012 to November 2017. All returns are monthly and are adjusted for rights issues and dividends.
What did I find? The average world beta coefficient across the 30 STI component stocks is 0.49, which is well below one. This confirms that the typical STI stock is relatively insensitive to movements in the world index. Put differently, investing in the STI isn’t the same as investing in a globally diversified portfolio. The benefits of holding such a world portfolio have already been covered in my previous blogs, so I will not repeat the message here (please refer to Blogs 31 to 33).
A single number (0.49) doesn’t tell you how similar or different are the world beta coefficients of each STI component stock. This information is shown in the following chart, where the red bars are the world beta coefficients of the individual STI firms.
Only three stocks (CapitaLand, Keppel Corp and Jardine Cycle & Carriage) have a world beta coefficient of at least 0.7. The median world beta coefficient for the other 27 stocks is only 0.47.
So, if the 30 stocks aren’t so responsive to movements in the world index, what moves their stock prices? The simple, but rather unsatisfactory answer is that each stock is mainly driven by the STI itself! Evidence for this can be seen from the green bars in the above chart.
Each green bar shows the local beta coefficient for the 30 stocks, calculated by regressing a stock’s returns against the STI’s returns. As you can see, the green bars are generally taller than the red bars. The average local beta coefficient is 0.88, well above the average world beta coefficient of 0.49. This is pretty solid evidence that even the biggest firms on SGX are more home-biased than global in outlook.
To sum up, the popular belief that Singapore stocks provide as much diversification as a world equity index is not supported by the data. Singapore might be a highly open economy, but the reality is that many of our leading firms have some way to go before they can be regarded as fully plugged into the world economy. Until then, the SGX will likely remain a self-referential market, with Singapore stocks largely be driven by, well..other Singapore stocks.
This spreadsheet contains the data used in running the regressions.