Style Diversification: An Often Overlooked Source of "Free Lunch"


I recently attended an investment conference and a phrase used by one of the presenters caught my ear. This occurred partly because of the significance of what was said and partly because I thought I had heard it before. Here is the phrase I heard:

“Diversification is the explicit recognition that we don’t know what the future holds.”

Before delving into this phrase’s significance, let me talk about having heard it before. Guru investor, Peter Bernstein1 (1919 – 2009), espoused this diversification message in a 2004 interview for MONEY Magazine2. My recollection of having heard the message before came from this Bernstein interview:

Understanding that we do not know the future is such a simple statement, but it's so important. That's what diversification is for. It's an explicit recognition of ignorance. And I view diversification not only as a survival strategy but as an aggressive strategy, because the next windfall might come from a surprising place. I want to make sure I'm exposed to it.

In short – diversification is the explicit recognition of our “ignorance” about what the future holds.

Let’s look at this “explicit recognition”. Diversification is an important lever for investors. It is surely the single most hands-on, readily-implementable and beneficial risk management tool at an investor’s disposal. There are many forms of diversification, from:

  • Holding securities across a number of asset classes
  • Having exposures to different geographies, currencies and industries
  • Investing in small, medium and large companies

The above list is not exhaustive but the rationale for diversification is the same despite the method(s) used. In fact, the idea of diversification is so old and so woven into our daily existence that we use the phrase “don’t put all your eggs in one basket” even when not referring to an investing context. The fact that few reading this paper can claim to have ever carried eggs in a basket attests to how long-lived this axiom really is. Diversification’s rationale comes from having some things that ‘zig’ while other things ‘zag’. In mathematical terms, diversification happens when lowly correlated items are combined.

This paper focuses on a type of diversification that often gets overlooked – Style Diversification.

What do I mean by investment style? Investment Style is defined as the philosophy a portfolio manager takes in building and managing a portfolio. The philosophy is characterized by the qualities of the stocks sought for the portfolio. Common styles include Value and Growth. Although, there are many others.

Why is style diversification prone to being overlooked? Often, the style a portfolio manager uses is a reflection of some facet of her own personality. Typically, a value manager looks for undervalued stocks (by some definition) and, not coincidentally, also looks for bargains in real life too. A growth manager looks for stocks with exciting future prospects and carries that optimism about the future into daily life. There is often a strong link between how a portfolio manager invests and how she lives her life. This strong link is why style diversification might just be the least employed method of diversification. After all, why would a growth manager ever, let alone consistently, own a value stock? Similarly, it would be blasphemy to suggest to a value manager to own a growth stock. Their investing personality is “dyed in their wool” so to speak. However, overlooking this free diversification opportunity is at an investor’s peril. The benefits missed can be significant.

The Benefits of Style Diversification

For style diversification to provide any benefit, there must be evidence of low correlation between two styles. Let’s first look at the potential to find evidence of low correlation between two different styles by looking at two indexes. The two indexes I chose are both for Canadian Equities. They are the Dow Jones Select Canada Value Index3 and the Dow Jones Select Canada Growth Index4. When combined, these two indexes contain roughly 200 stocks. The constituents of these two Dow Jones indexes are medium and large market capitalization stocks found primarily in the S&P TSX index.

Graph 1, below, shows the rolling one-year returns for the two indexes relative to the return of the S&P TSX index. In other words, for each month graphed, I measured the amount by which each index had out- or under-performed the TSX over the preceding 12 months. I am not showing the return that each index generated over the preceding year. I am showing how much that return exceeded or was exceeded by the return of the S&P TSX.

Graph 1

It is readily evident that these two Dow Jones indexes act in opposing manners. They have in many instances, resembled mirror images with one style producing outperformance above the S&P TSX while its style counterpart is underperforming. There are many points where the amount of outperformance produced by the one style is roughly equivalent to the amount of underperformance produced by the other style. Please note that the winning style is not always the same one. This reality gives credence to Bernstein’s assertion that diversification is necessary because you don’t know which style will be the one to outperform in the future. Each style takes its turn in ‘winning’ and in the long run both of these styles have similar, attractive return prospects. Some might think “couldn’t we do even better if we allocated strategically, moving our money back and forth into the winning strategy?”. Of course, we would! However, this notion flies in the face of Bernstein’s assertion that we are ignorant of the future. Market timing is a losing game. Period.

Graph 1 isn’t about picking a winning style. It is about seeing if styles can be combined to an investor’s advantage. The prospects look good. To see the benefit of style diversification, let’s review Graph 2, below. This graph displays the rolling 3-year volatility of the monthly absolute returns of each of the two indexes relative to the volatility of the S&P TSX. In other words, any time the lines are above 1 (on the left axis) then that index is exhibiting more volatility than the S&P TSX. Any time the lines are below 1 (on the left axis) then that index is exhibiting less volatility than the S&P TSX. As one might expect, the Growth index can produce a level of volatility (risk) that is considerably greater than that of the S&P TSX. This was particularly true of the period 2012 -2014. However, one can see that the volatility of the Value index isn’t always less than the S&P TSX or even less than the volatility of the Growth index. In fact, in 2016-2018, Value was the most volatile of the two styles. The powerful outcome of style diversification though is shown by the blue line plotted in the graph. This line displays the amount of volatility found in a portfolio that consists of a 50% allocation to each of the Growth and Value indexes. For most of the period, the 50/50 portfolio had less risk than either of its two component parts (Growth and Value). And, for a considerable period, the 50/50 portfolio exhibits less risk than the S&P TSX. Having less risk than its component parts is possible because these two styles are lowly correlated.

Graph 2

There are many ways to think of diversification and many ways to think of risk. Volatility is not our favoured means of depicting risk even if it is the most commonly used risk metric in the investment industry. We like to think of risk as the probability of losing money and the expected amount of loss when you do lose. We are long term investors and expect over the long run that the market will increase in value. Therefore, we look at the probability of loss relative to the S&P TSX by measuring the monthly amount of over or under performance relative to the benchmark. Let’s look at risk from this perspective. Again, the power of style diversification shines through.

Using monthly data, since the inception of these two indexes in 2007, the following risks have been encountered:

Let me use the Growth index as an example to ensure that the reader understands the numbers being portrayed in this chart. We calculated the Growth index return minus the return of the S&P TSX in every month since the inception of the Growth index. We then looked at three facets of the Growth index’s relative return:

  • Probability of a losing month – what percentage of months did the Growth index return get beat by S&P TSX return? For the Growth index, 52% means that the index underperformed the S&P TSX in a slight majority of months.
  • Average loss in a losing month – In the months when the Growth index return underperformed the S&P TSX, it did so by an average of 1.4%. In other words, during losing months, the Growth index return was, on average, 1.4% behind the S&P TSX return.
  • Worst loss in a month – what was the largest amount by which the Growth index underperformed the S&P TSX? For the Growth index, the worst month ever was over 6% behind the return of the S&P TSX.

Please turn your attention to the final column in the chart, which exhibits the experience, over the same period, of having half your money in Growth and half in Value. The probability of losing is below 50%, indicating that the portfolio returns exceed S&P TSX returns during the majority of months. Check. The average loss in a losing month is a fraction of the average losses experienced by either of the two indexes. Check. And, finally, the worst month ever represents a significant improvement over the experience of either of the two indexes. Check.

Using either volatility statistics or our probability and extent of loss view of risk, a portfolio of the two indexes combined exhibits considerably less risk. Less risk is precisely why investors diversify. I used stock market indexes to demonstrate this point. A reader might conclude that I am advocating for a passive approach to portfolio management. That is not my intent. I merely used the indexes as proxies for how a style has performed and how a portfolio could perform when styles are combined. We back-tested the effects of combining active Growth and Value strategies and saw the same benefits. More importantly, since December 31, 2013, we have managed our Canadian Equity clients’ portfolios with 50% allocations to each of our Growth and Value strategies. Bernstein’s notion was that he didn’t know from where the next windfall would come. True to that notion, during 15 of the 21 quarters for which we have been using this style diversification approach, one of our style allocations outperformed the S&P TSX while
the other allocation underperformed. The odds were close to even as to which style was the one outperforming (8 times for our Value allocation & 7 times for our Growth allocation). Very beneficially, the risk of the entire portfolio has been less than our benchmark and less than the risk exhibited by most of our peers. In fact, over the five-year period ending December 31, 2018, Pavilion Advisory Group ranked our fund as the 9th least volatile Canadian Equity fund.5

Concluding Thoughts

Let’s return to Bernstein’s idea. In summary, he suggests that 1) we are all ignorant about what the future holds and 2) given this ignorance we should commit to diversification. Allocations to opposing styles are an often-overlooked diversification avenue. In our opinion, allocations to opposing styles are a great defence against the uncertain future.



  5. Source: Pavilion Advisory Group peer performance survey 12/31/18.