Sectors Or Geography How Should You Invest?
One of the enduring philosophical questions about portfolio design revolves around whether the underlying implementation of diversification should be based on sectors or geography. Both have merit. Both have adherents. How should one proceed?
Much of the decision revolves around one’s perspectives (let’s be clear here – we’re talking about go old-fashioned biases – and we all have them) and attitudes... Will the portfolio’s main approach be strategic or tactical? Are taxes a significant consideration or not? How much time and energy are you willing to commit? How confident are you in your ability to determine market cycles and the like? If you favour one approach, how willing are you to risk that the other approach might do better for a prolonged period? – call it “investor’s remorse”.
Sector Diversification
Let’s begin with sectors. My sense is that this style of investing is more of an old school/ traditional approach. My experience is that most adherents have a greater appetite for risk and a greater confidence in their ability to exploit fundamental and technical analysis to their advantage. There are perhaps a little over a dozen sectors in the opinion of most commentators. For greater clarity, I’ll point out that these include, but are not limited to: materials, energy, healthcare, consumer staples, utilities, technology, real estate and, of course, financials. Some people include cash as a “sector”, as well. Using sectors as the primary lens for investment decision-making requires that the investor have a distinct outlook on whether any given sector will do better or worse than average over the foreseeable future.* The idea here is one of going to where the grass is greenest and opportunistically moving from one sector to another as circumstances shift. The phrase to describe the strategy is often called “sector rotation” and the primary considerations include whether the economy is expanding or contracting, whether monetary policy is accommodative, tight or neutral. Generally speaking, the approach is referred to as “tactical asset allocation”.
Examples abound. For instance, if the economy seems set to expand, things like infrastructure and consumer discretionary might do particularly well. Conversely, if the economy were to be in a contraction mode (such as we have experienced in Canada for much of 2015), then more defensive sectors such as health care and utilities might perform better - or less poorly, at least.
Sector rotation strategies come in various shapes and sizes. Some managers go “all in” for those sectors that they favour, while others simply overweight and underweight particular sectors (i.e. engage in ‘tilts’) in order to express their opinion while simultaneously hedging their bets. As alluded to earlier, some managers are comfortable in going substantially to cash, while others are not (or cannot- depending on their mandate / investment policy).
Geographical Diversification
People who diversify by geography belong primarily to the more contemporary school of modern portfolio theory adherents and are generally proponents of traditional Markowitz optimization strategies. These folks are generally a bit more concerned about correlation co-efficients and risk management. There are limits to this approach, however. For instance, when it comes to income investments, the most I have seen anyone offer is to have domestic income and foreign income. Further segmentation simply isn’t really done, it seems. When it comes to geography, most Canadian investors would agree that there are four primary asset classes that can be used: Canadian equity, American equity, International equity and Emerging Markets equity. Setting target weights to these different asset classes and re-balancing to those targets as market conditions change is called “strategic asset allocation”.
Some people (myself included) use a hybrid approach. I mostly diversify my clients’ accounts by geography, but I also have an express allocation to tangibles (traditional inflation hedges like commodities) as a separate asset class. With that disclosed, I should add that my preference has long been for “soft commodity” global sectors like agriculture, water and infrastructure - and not the beaten-up “hard commodities” like materials, oil and gold. Obviously, this discussion does not have to be of the all or nothing variety – there’s plenty of room for personal variations.
Let’s take a moment to compare and contrast the two approaches. In so doing, I believe it will quickly become clear that I mostly favour the geographical approach. That said, I was on BNN’s Market Call recently and a caller phoned in to say that any allocation that deviates from the global capitalization weights is a de facto “call” on those parts of the world. Although I think most people would say his argument is a bit of a stretch, one could certainly see the gentleman’s logic in pointing out the (technical) irony.
For example, let’s say someone had a passive portfolio invested equally in six (mostly geographical) asset classes: income, tangibles, Canadian equity, U.S. equity, International equity and Emerging Market equity. Using global market capitalization as a guide, this portfolio would still be significantly underweight in income and U.S. equity and significantly overweight in Canadian equity. I could say that this allocation is my strategic target and defend it as such. A critic could say that because I am holding some asset classes in weightings that are different from their global market capitalizations, I am effectively making a “tactical” bet on those regions. That person would not give credence to the fact that the weightings are strategic and unchanging due to circumstances. To me, this is a specious argument. Here are the considerations as I see them:
Taxes
Obviously, tax considerations are moot in registered plans. If the money being managed is held in taxable accounts, however, then portfolio turnover can quickly become a real consideration. The more you trade, the greater the tax implications of your activity. As a general rule, the geographical approach is more tax effective for two reasons. First off, most geographical people have a more or less allocation for each part of the world. As such, their re-balancings are usually modest in comparison to the tactical sector folks who sometimes move in and out of sectors entirely. Secondly, since the geographic people can re-balance on the margin, they can do so with money in motion. In other words, they can buy low by adding to positions that are below the target if money is being added or they can sell high if there are sells that are required (retirement income, for instance). In both instances, the tax impacts are relatively modest compared to what would likely be involved in a sector-based strategy.
De Facto Overlap
Most people should realize that if they invest in a diversified basket of global securities, they would likely be close to the global sector weights as a consequence. The same might not be true if using global sectors. Most mutual funds and ETFs offer exposure to a Canadian sector or an American sector or whatever other part of the world you care to name. Getting a product that offers true exposure to (say) global health care is harder to do. Even if you can find a product that does this, can you find one for all the other sectors, as well? Even if you find other global sector products, can you find them in C$ - or will you have to buy them in U.S. $ and face conversion costs, exchange differentials or both?
Home Bias
Another concern that I have is Home Country Bias. It applies in varying degrees to both strategies. According to Bloomberg, as of June 30, 2015, the Canadian stock market was dominated by three sectors: financials (35.01%), energy (20.43%) and materials (10.78%). Taken together, these three sectors alone make up for nearly 2/3 of the stock market capitalization of our homeland. You should know that I have been working with the people at Blackrock to develop a product (to be released in early in 2016, we hope) that offers exposure to the TSX while totally excluding those three sectors. Canada, in turn, makes up only about 4% of the earth’s total stock market capitalization. If you’re just buying sectors and using Canadian securities and products to do so – you’re only getting exposure to 1/25 of the stocks available. Imagine going to Canadian Tire while there’s a sale on, but only shopping in one aisle. There may be some good deals in that aisle, but there will be plenty of good deals in the other aisles, too. By investing exclusively (or even primarily) in Canada, you are ignoring a world of opportunity. The new product- should it be released – would allow people to invest in Canada without being massively overexposed to the “big three” sectors. In the interest of completeness, you may be interested to know that, according to Dimensional Fund Advisors, world stock market capitalization is about 52% ($28 trillion) based in the U.S., 31% ($17 trillion) in International Developed, 12% ($8 trillion) in Emerging Markets and 4% ($2 trillion) in Canada. About 1% of the World’s stock market capitalization is invested in so-called Frontier Markets.
Historical Returns And Correlations
One thing related to the problem of home bias is the fact that, historically, other parts of the world have experienced better long-term returns than the Canadian stock market. Combine that with the fact that correlations, while increasingly positive, also allow for a moderate amount of risk reduction. According to Morningstar, historical annualized stock returns from 1950 to June 30, 2014 are 10.1% for Canada and 11.4% for the U.S. Other parts of the world have also outpaced Canada over long timeframes.
Conclusion
As with many things in life, there are many ways to approach investing. Each has its advantages and drawbacks. To my mind, a strategic approach using geographically-defined asset classes is superior, but that is not to say that a tactical, sector-based approach is without merit. As always, personal circumstances must be taken into account. Some of the considerations itemized here might not apply to you.
*Note that in the real world, nothing can be reliably foreseen.
John J. De Goey, CFP, CIM, Fellow Of FPSC is a Vice President and Portfolio Manager at BBSL. The views expressed may not be shared by BBSL.
E: john.degoey@bbsl.ca; T: @JohnDeGoey_BBSL.
www.johndegoey.com