Skip to main content

Arpad Benedek

A few years ago a gentleman called me up asking for a meeting to talk about his portfolio. I don't work directly with retail investors but oftentimes I'm happy to bounce some ideas around over a coffee and help them find an adviser. In this case however, he was adamant that he pay me to only help him benchmark his portfolio. I refused payment, but showed him some simple ways to do just what he was looking for.

In his particular case, he had a well-diversified portfolio with equities spread out globally, and a well-diversified bond portfolio domestically. He had individual stocks for the most part in North America, some mutual funds for developed foreign markets and a mix of active funds and index funds for emerging markets exposure.

He was working with an adviser, but when they discussed performance monitoring, the adviser explained they would use the TSX as their benchmark. Now, this isn't actually an incorrect benchmark, but it depends on your perspective.

Comparing a globally diversified portfolio with a mix of equities and fixed income against the TSX is fine if you are trying to figure out if diversifying out of Canada and out of equities is working for you from an asset-allocation point of view. If his portfolio outperformed the TSX, it's still possible that he was holding bad individual stocks, bonds and mutual funds.

We want to find a simple way to figure out how well the various parts of the portfolio are doing. If the emerging markets index was up 50 per cent, and your emerging markets investments were up 35 per cent, the 15 per cent lag would get lost in the overall portfolio comparison to the TSX. Lagging by 15 per cent there, and "only" earning 35 per cent might be the best performing part of your portfolio. But if you only held Canadian stocks and you lagged the TSX by 15 per cent, the red flag is evident.

One solution is to create a weighted average benchmark. This is really simple if you have an Investment Policy Statement (IPS) that outlines the asset allocation for your portfolio. You simply take the benchmark index for each asset class and multiply that by the percentage to which it is allocated in your portfolio.

For example, let's say you had a portfolio that was 30 per cent Canadian fixed income, 30 per cent Canadian equities, 20 per cent U.S. equities, and 20 per cent global developed equities. The respective returns for the benchmark indices were 3.5 per cent, 7.8 per cent, 5.6 per cent and 4.4 per cent. The equation is as follows:

(30% x 3.5%) + (30% x 7.8%) + (20% x 5.6%) + (20% x 4.4%) = Weighted Average Benchmark

In this case, your custom benchmark return was 5.39 per cent.

If your actual portfolio return was, for argument's sake, exactly 5.39 per cent as well but you used only the TSX as a benchmark, you would see a lag of more than 2 per cent. For this particular year, maybe asset allocation didn't work in your favour (it's not supposed to every single year), but the investment selection seems just fine.

So when it comes to benchmarking your portfolio performance, there is more than one way to skin a cat. A weighted average benchmark is going to be more useful in assessing the investment selection ability because it factors out asset allocation as a variable.

Simply using a weighted average benchmark isn't the be all and end all of performance monitoring as it's possible to outperform in one asset class and underperform in another asset class, with the resulting variances cancelling each other out. Either of these simple monitoring tools are not sufficient to make changes to your portfolio without further, deeper and, hopefully, longer-term analysis.

Preet Banerjee is a senior vice-president with Pro-Financial Asset Management. His website is

Report an error

Editorial code of conduct