I was recently reading a commentary I wrote two years ago on the merits of holding Canadian banks. I was addressing a question from a reader as to suggestions on how to structure a portfolio that would generate a comfortable nest egg that in five years time would provide for the reader’s retirement.

It was a 50-something couple — call them the Smiths — with an 18-year-old daughter. Their investments were geared to provide low risk returns to retirement.

Their daughter’s education was taken care of and would not impact any portfolio decisions. Nor were they counting on contributions coming from the wife’s income, as her salary supported staple items and a horseback riding hobby.

The Smiths’ principal home was valued at $480,000. They had a combined portfolio — half in RRSPs and half in non-registered plans — valued at approximately $358,000. They owned a rental property — a condominium valued at $200,000 with $70,000 left on the mortgage — with the rent paying for the mortgage. Finally, the husband had a pension from a former employer that would pay $1,500 per month beginning at age 65. Their question was simple: what did they need to do if they wanted to retire in five years, at age 60, with a $40,000 annual income?

In that article, I reasoned their goals could be satisfied with a 7.5% rate of return, without adding money to their portfolio. That conclusion was based on a recognition that the Smiths’ portfolio need only deliver $40,000 per year indexed at, say, 2% per annum, from age 60 to 65. At 65, the cash flow from the portfolio could be reduced to the equivalent of $22,000 adjusted by the 2% per annum inflation computation. The remainder would be topped up by the $18,000 ($1,500 per month) pension income.

A return of 6.6% per year would still meet the required income to age 90. At that point, the Smiths’ portfolio would be depleted. A simple portfolio model like the FPX income index or the FPX balanced index (the FPX indexes were developed for the National Post and are reported daily, with historical values archived at my Web site, at www.croftgroup.com, under the benchmark heading) would be expected to meet those financial objectives; and that was the basic recommendation.

What drew this to my attention was that the FPX solution was a basic answer to a more complex question. When you dug deeper into the Smiths’ portfolio, 56% of the assets were invested in the banking sector. The rest was in medium-risk mutual funds, which offered broader diversification.

The real issue for me in January 2006 — when the article was originally penned — was the concentration within the banking sector. On the one hand, modern portfolio theory would never allow the bulk of one’s assets to be put into a single basket — in this case, bank stocks. On the other hand, it was hard back then to argue against banks as a long-term investment.

In the 2006 article, I referred to a story about the U.S. Federal Bureau of Investigation’s capture of notorious bank robber William Sutton. When asked why he targeted banks, Sutton replied, “Because that’s where the money is.”

Sutton knew what most long-term investors have, until recently, known instinctively. Since the mid-1950s, the total return from banks has outpaced the broader market by a factor of four to one.

It leads one to ask whether the same holds true today. The last 18 months have been one of the most difficult periods for Canadian banks. Is this an example of mean reversion — meaning, a longer-term trend of underperformance? Or is this a short-term period where banks are challenged, but eventually will recover and again take their place as market leaders.

At the time, I was torn between suggesting a reduction in the Smiths’ exposure to banks and maintaining the status quo based on 50 years of performance history. Clearly, a recommendation to reduce exposure would invite performance comparisons down the road, which presumably, would favour the banks.

But if one applies portfolio theory, you cannot escape the fact that banks are a sector within the economy with risks that are unique to it. Sector exposure has a tendency to ramp up portfolio risk; and portfolio management is all about reducing risk.

@page_break@But here’s the rub: The Smiths were obviously comfortable holding an overweighted position in banks. If a good portfolio is one that keeps you invested for long periods, I argued back then, the comfort factor should play a role. So, when it came to the banks, it was left to the Smiths to decide.

Now, two years later, it is apparent again that the conventional approach to portfolio management is always the right approach: a well-diversified, low-risk portfolio should always take precedence over the psychological comfort investors get from holding what they think they understand.

And so, we come full circle. The role of investment professionals is defined by their ability to steer investors away from the comfort factor toward the right portfolio. Two years later, I wonder how comfortable the Smiths are with their banks while the FPX portfolios delivered the required rate of return. IE