Diversifying across regions, asset classes, and factors versus market timing

A common mistake that can derail a successful diversification plan is trying to time markets by moving money in and out of asset classes, regions, or style groups in anticipation of, or in reaction to, performance. Among investment behaviours that we addressed in our 1993 article, market timing was the first one: “Market timing rarely works”, we wrote. Markets are complex systems that even professionals have great difficulty forecasting. Not only are there many economic, political, technological, demographic, and other factors that may drive market performance, people’s expectations of those factors, and even their expectations about other investors’ expectations, may take markets in unexpected directions.

To demonstrate the challenge, we collected eighteen years of annual return data for different asset classes and plotted them together, each with a different color, in a single table (Exhibit 1) . The table is organized from top to bottom, from best performer to worst performer, and left to right by year. To time the markets is essentially to try and forecast the colour pattern of this table in each new year.

Exhibit 1: Annual performance of some Canadian and International indices

In the eighteen-year period covered for the stock and bond indices presented in Exhibit 1, the best result that investors could have achieved by picking a single asset class in a single year would have been to invest 100% of their portfolios in the S&P TSX Small Cap Index of Canadian stocks in 2009 (+62.4%). The worst result that investors could have achieved under the same conditions (i.e.: the same time frame, the same opportunity set, and the same, single-choice scenario) would have been to invest 100% of their portfolios in the S&P TSX Small Cap Index of Canadian stocks in 2008 (-45.5%). These results at the boundaries can elicit fear and greed and may encourage what we described in 1993 as “fund-hopping”.

Because the performance of individual asset classes can have such large deviations from year-to-year, investors who fund-hop often switch into a fund just when its performance turns down. This practice (“chasing performance”) can damage returns and can result in sub-optimal long-term performance. When diversified investors do not hop around, diverging annual returns of different asset classes are one of a diversified portfolio’s most important attributes. Consider the following: the average annual return of the S&P TSX Small Cap Index of Canadian stocks during the whole eighteen-year period examined in Exhibit 1 was 6.3%. The average annual return of the S&P 500 during the period was 6.3%. The average annual return of a balanced portfolio rebalanced annually to be 70% in the six stock indices and 30% in the two bond indices was also 6.3%. The variability of the balanced fund, however, was much lower than the variability of the individual stock indices (see white boxes, Exhibit 2).

Exhibit 2: Annual Performance of a Global Balanced Strategy

The lower volatility occurred because different asset classes peaked and troughed at different times. This is a characteristic of a diversified portfolio. In 2001, international large-cap stocks were the worst-performing asset class in the group while US small-cap stocks were one of the best-performing asset classes (benefit of geographic and size diversification). In 2009, the Canada Universe Bond Index was the worst-performing asset class in the group, while Canadian small-cap stocks were the best-performing asset class in the group (benefit of asset class diversification). In 2014 large-cap US stocks were the best-performing asset class in the group while small-cap Canadian stocks were the worst performing (benefit of size and factor diversification). The major benefit of diversification is that these variable performances offset each other and lower the overall variability of a diversified portfolio.