In earlier pieces we've discussed the nature of portfolio rebalancing, and discovered that there is often an excess return to be earned from it. In general, when a small amount of a volatile asset is added to a portfolio, and rebalanced regularly, its actual return is usually considerably higher than the stand alone return. This is because the interaction of asset volatility and policy rebalancing produces a natural "buy low/sell high" bias into the rebalancing transactions. In addition, rebalancing benefits most with risky assets with similar returns.

In previous work we have shown that this "rebalancing bonus" is approximated by

where r is the correlation of the asset with the rest of the portfolio, and Var is its variance. Precious metals equity is a good case in point. The variance of a diversified portfolio metals equity is in the 0.1 range, and its correlation with most diversified portfolios is zero. Thus, an approximately 5% (1 x 0.1 x 1/2) excess return on the asset is earned simply by rebalancing it regularly.

Are there other assets which behave the same way? Indeed there are; the equity assets of emerging market nations. For monthly returns for the period 1/88-3/96 the following nations had a correlation of less than 0.2 with large US stocks (the S&P500): Brazil, Chile, India, Korea, Taiwan, and Turkey. Brazil and Turkey both had variances in the 0.5 range as well, suggesting the possibility of up to a 25% "rebalancing bonus" for the disciplined investor in these volatile markets. To be sure, there are significant transaction costs for both the large and small investor in the securities of these nations, but they are dwarfed by the potential excess returns. To say nothing of the emotional rigor required to buy equity in a country in the wake of a coup, civil war, or natural disaster.

If you are a small investor, and you decide to go this route, and wish a reasonably diversified approach to a given national market, you have only one choice -- single country closed end funds. They trade on the major US exchanges, usually the NYSE. These securities do have drawbacks. Firstly, their expenses are quite high, usually about 2% per year. Next, trading costs will be on the order of 2%-4% round trip even with a deep discount brokerage, because of the wide bid-ask spread of most of these securities. Since they trade in the US, and can sell at a premium or a discount to their NAV, will they not be subject to the sentiment of the US market, thus increasing their correlation with it? If so, then the increased correlation with the US market would greatly lessen their value to the US investor. How well are these funds managed? Is the US investor likely to see their returns lessened by incompetent active management?

In order to answer these questions I decided to examine the performance and correlation of the available emerging market single country funds. The data are summarized below. For each country the senior fund was chosen. Monthly returns for each fund were studied beginning either at the inception of the fund or January 1988, whichever was later. (January 1988 being the beginning of the available emerging markets data from MSCI.) Returns for the study period for the market return, NAV, and appropriate MSCI country index are tabulated, as are the correlations of each of the three returns with the S&P500.

The returns data are pleasantly surprising. Although the average market return is significantly lower than the average MSCI index return, this is largely attributable to the well known proclivity of these funds to develop fairly wide discounts after issuance. Many of the countries studied have two or more funds available, causing further widening of their discount. However, there has been few recent single country IPOs, and it seems that further significant widening of the fund discounts is unlikely. The NAV return is a better index of management skill, and it is somewhat surprising that this is actually a full percent higher than the average MSCI benchmark, an impressive accomplishment in the face of the typical 2% fund expense. It would seem, then, that the purchase of these shares at a historically high discount would more likely than not result in benchmark beating performance.

It is reassuring that the correlations of the NAV returns with the S&P are nearly identical with the benchmark returns -- a quite low 0.20. Unfortunately, as expected, the fact that these funds trade in the US markets adversely affects the fund market return/S&P correlation. The correlation of the market returns with the US market is considerably higher at 0.34. A correlation of 0.34 means that you will lose about one third of the "rebalancing bonus" estimated from the "half variance" approximation. Further, you are going to be losing about 3% per year in return to fund expense and trading expense. Still, for an asset with SD = 0.5 and Var = 0.25, that still amounts to an about 5% per year return increment.

Emerging Markets closed end funds are no walk in the park. However, for the small investor with nerves of steel and the discipline to buy into disaster and sell into euphoria, the potential returns are considerable.

copyright (c) 1997, William J. Bernstein