Geographic Diversification in Passive Investing

Konstantin Zhdanovich

Aug 24, 2023

When engaging in passive investing to reduce volatility, investors generally prefer to diversify their investments. While the potential diversification effect between asset classes (stocks and bonds) is intuitively clear (the higher the share of bonds, the lower the portfolio volatility and the lower the return over a long period), the dependencies in geographic diversification are not obvious. To demonstrate the issue, let's compare the returns of a portfolio consisting of 100% S&P 500 with some of its more diversified counterparts (S&P 500 with the addition of various MSCI indices — see the list of portfolios in the chart legends) over the period 1993–2023. As calculations show, over the considered period, the portfolio consisting of 100% S&P 500 (black round marker on chart 1) demonstrated a better return-to-volatility ratio than its more diversified counterparts (blue markers on chart 1). Moreover, the concentrated portfolio consisting of 100% S&P 500 would have shown the smallest maximum drawdown over the period at 50.9%, whereas diversified portfolios had drawdowns of 52–56%. Thus, over the considered 30-year period, adding regional indices to the portfolio did not increase the investor's return, reduce volatility, or decrease the maximum drawdown of the portfolio. Let's consider the portfolio returns in dynamics. Diversified portfolios were significantly better than the concentrated portfolio in only 9 out of 30 cases (annual returns were analyzed), with most of the 'successful' years being grouped in the 2000s. Thus, over the considered period, a passive investor would not have gained significant advantages by adhering to geographic diversification. The performance of their portfolio could have been improved with more active investing, implying changes in the portfolio structure according to changing macroeconomic conditions.

Notes: - Calculations are based on Bloomberg data. - The MSCI Middle East and Africa index has a short history — since 1997. For the purposes of this calculation, the index's history was extrapolated using the MSCI World index for 1993–1996. - Maximum drawdown calculations were made based on monthly data.