There's an old expression in investing that "there's no free lunch" - you can't get something for nothing and any investment that offers the prospect of higher returns inevitably brings greater risk with it.
Until recently, it was uniformly believed the one exception was the promise of diversification across asset classes. Based on Nobel-winning research into modern portfolio theory, the underlying notion was that mixing the right combination of assets could both increase return and reduce risk. The key was taking advantage of the tendency of different assets to react in contrasting ways to economic events.
To many, modern portfolio theory was among the casualties of the global financial meltdown of 2008, as some assets that were supposed to move in different directions to one another went down in tandem. This caused many critics to question whether the principle of diversification is still relevant.
Recently, I sat down with Michael Nairne and his team at Tacita Capital in Toronto to explore this. We first examined the performance of six different investments from January, 1972, to July of 2010, looking at both their return and their standard deviation, or volatility. (Standard deviation is the most common way the investment industry measures risk - the higher the number, the bigger the swings and the higher the risk.)
We looked at stocks in Canada, the United States and the rest of the world (using the MSCI EAFE index), as well as real estate investment trusts, commodities and gold (in Canadian dollars). The return and volatility of these investments varied considerably over the nearly 39-year period. REITs, for instance, returned an average of 11.93 per cent a year, while gold returned a mere 8.97 per cent. Volatility varied from a standard deviation of 22.47 per cent (for gold) to a low of 16.18 per cent for U.S. stocks.
If these were your only choices, investors searching for the best return would select REITs; those seeking the lowest volatility would invest in U.S stocks.
We then took this one step further and looked at what happened if we combined asset classes into portfolios. Starting with just Canadian stocks, we sequentially added additional investments - first stocks from the U.S. and then from the rest of the world, then REITs, commodities and finally gold, reducing the allocation to the existing asset classes to make room for the newcomers. In the process, the stock component went from 100 per cent in the first three portfolios to 60 per cent in the last one, as we assigned 20 per cent to REITs and 10 per cent each to commodities and gold.
And at the end of each year, we rebalanced the allocation back to the target weighting - so that we increased the allocation to underperformers in the past year and trimmed the amount held in assets that had outperformed over the last 12 months.
As each new investment category was added, something remarkable happened. The overall return of the portfolio went up and the volatility came down. The more portfolios were diversified, the better the investor experience. Combining all six assets produced a portfolio with average annual returns nearly equal to the best performing asset class, but with far less volatility.
Diversification doesn't always work in the short term, but it certainly has over time. A key reason is because of the annual rebalancing every twelve months - buying more of what's done poorly, selling some of what has done well.
Mr. Nairne summarized it this way: "Diversification's free lunch is still being served. To enjoy it though, you have to order a robust assortment of assets and remain seated until the entire meal is on the table."
To which I'd add that you also have to eat your vegetables. As our results show, smart investors must have the discipline to consume not just what's tasty but the stuff that will build muscle to withstand nasty spells of bad weather and unpleasant surprises.
Dan Richards is president of Clientinsights. He is a faculty member in the MBA program at the Rotman School at the University of Toronto.
dan@clientinsights.ca
Dan Richards