Devising a well-diversified portfolio
Here’s a ‘control portfolio’ to help investors with their own efforts at broadening their equity exposure
The Canadian market has many high-quality public companies, but in the eyes of global investors the S&P/TSX composite, with performance dominated by commodity and bank stocks, is a mediocre investor benchmark.
Market history and financial research strongly implies that a domestic investor buying an ETF tracking the S&P/TSX composite can expect to outperform active investors choosing individual domestic stocks. But for Canadians who’ve got domestic stocks covered in this way the next obvious questions are okay, what next?
What other indexes, asset classes and sectors do I need for diversification? And what’s the mix? They recognize that the domestic benchmark fails to benefit significantly from secular growth sectors like artificial intelligence, luxury goods, emerging markets, cloud computing, small-cap stocks, global infrastructure development, online retail and health care.
The equity-only portfolio shown here is intended as a benchmarking control portfolio against which investors can compare their own stock-portfolio performance. The intention is to be as diversified and long-term focused as possible – minimizing the need for economic forecasting and the number of sector-oriented investing decisions investors are forced to make – while not unnecessarily adding to the risk of permanent loss of capital.
Saving for retirement is the most common portfolio objective for investors and, since retirement expenses will be primarily in Canadian dollars, it makes sense to emphasize domestic stocks in the portfolio. An ETF tracking the S&P/TSX composite forms just under half of the benchmark portfolio.
The S&P 500 makes up 32 per cent of the portfolio. The U.S. index provides the exposure to technology, industrial and healthcare stocks that the domestic benchmark lacks, and the multinational behemoths that drive the index’s performance generate revenue worldwide, benefiting from global economic growth.
I chose the U.S.-traded S&P 500 ETF because I didn’t want to fully hedge against currency risk in this segment of the portfolio. In performance terms, hedging the currency implies a prediction about the future value of the loonie, and I didn’t want to make that call or lose the potential portfolio benefits if the U.S. dollar keeps rising.
Currency risks are partly mitigated in the portfolio, however, in another way, in another section of the portfolio – with an allocation to a gold bullion ETF. Gold, known as the anti-dollar, historically moves in the opposite direction of the U.S. dollar and provides a hedge against currency risk. Bullion can also act as insurance against political and financial risk (although its minor appreciation during the financial crisis means its function as a hedge against financial risk is somewhat debatable).
Income remains a major investor focus so global dividends account for 5 per cent of the portfolio. Annoyingly, I had to split the allocation in two parts to gain exposure to both U.S. and global dividend stocks. The global focus is intended to benefit from the higher dividends available in European and developing world markets.
The same issue occurred in the small-cap portion of the portfolio – there’s one for U.S. small-caps and another for global smaller companies. The small-cap allocation represents an attempt to boost returns. Financial research consistently shows that smaller company stocks outperform large caps over the longer term, but at the expense of higher volatility and risk. At 5 per cent of assets, a bad patch of small cap returns won’t damage overall performance while providing diversification and (one hopes) growth.
I made one sector call in the portfolio – an exception to my general premise of making as few investment decisions and predictions as possible – with healthcare stocks. This is a measure of my confidence that developed world (and Chinese) demographics will result in consistent, longterm revenue growth for companies in the sector.
The accompanying line and bar charts detail performance for the portfolio relative to the S&P/TSX composite, while the table shows returns for each investment. I don’t want to emphasize past returns too much – everybody’s hindsight investing is great. It is likely more important that diversification appears to limit volatility, as evidenced by the smaller downdrafts in the period from August, 2015, to February, 2016, and the fact that the five-year annualized standard deviation for the portfolio, at 10.8, is below the benchmark’s 11.7 (not shown).
There isn’t space to detail all of the reasons I picked each segment of the portfolio, but I should note the Efficient Frontier methodology that won Harry Markowitz a Nobel Prize in 1952 was not used. There are a bunch of reasons for this, but the main one (and most investors aren’t aware of this) is it requires extensive “adjustments” in terms of analyst assumptions before it can be used. These adjustments almost all require an opinion about the future performance of asset classes, and I was trying to avoid forecasting.
In addition, the work of Nassim Taleb, author of The Black Swan, has identified serious mathematical shortcomings for the Efficient Frontier method as he proved it significantly understates actual market risk.
The selection of investments for this portfolio was informed primarily by research on how and why passive investing outperforms almost all active investors over the long term. An older book about famed Legg Mason portfolio manager Bill Miller was also part of the thinking. In The Man Who Beats the S&P: Investing with Bill Miller, Mr. Miller notes that equity benchmarks such as the S&P 500 and the S&P/TSX composite are actually actively managed – stocks are chosen and removed based on set criteria – which is a realization I’ve been thinking about consistently for over a decade.
But the lack of formal investment selection methodology for this portfolio is the primary reason I don’t know what to call it. It can’t be a model portfolio because there’s no specific model behind it, so “control portfolio” is maybe best. My hope is that investors can find different uses for it, comparative or analytical, as time goes on.
Edited on 4 May 2017