Skip to main content
portfolio strategy

The lameness of traditional portfolio diversification is one of the great investing letdowns of the past few years.

Diversification is summed up by the folk wisdom of not putting all your eggs in one basket. This is essential advice, and it still applies. But the last few years of stock market upheaval has shown us that the standard means of diversifying - the mixing Canadian, U.S. and global stocks plus bonds - is too simplistic.

One way to take diversification to a higher level is to look at correlation, or the extent to which different types of investments go up and down in unison. Remember, the essence of effective diversification is not owning lots of different things. It's owning securities that don't go up or down at the same time.

A recent report by the investment dealer CIBC World Markets debunks one myth of diversification, which is that you can help cut risk by investing not only in the Canadian market, but also in the United States and globally.

CIBC found that where one of these regions goes, so goes the other in many instances. Worse, it seems that the rougher the global stock markets look, the more global stock markets start to perform similarly.

"Just at the point when things aren't going well and you want the most diversification, that's the point where diversification seems in short supply," said Meny Grauman, co-author of the CIBC report.

His research shows that Canadian stocks were 90 per cent correlated to global stocks for the 12 months to April 30, including the United States. A correlation of 100 per cent means two entities move identically, while a correlation of 0 means there's no connection and a correlation of -100 per cent means they move oppositely.

Mr. Grauman's numbers show that Canada was 87 per cent correlated to global markets 10 years ago; 15 years ago, it was 62 per cent; 20 years ago, it was only 55 per cent. "The increased globalization and interconnectedness of financial markets has meant that Canadian equities are much more sensitive to what's going on outside our Canadian borders," Mr. Grauman said.

Behind these numbers is a pattern where the correlation between Canadian and global stocks is almost twice as intense in poor markets as it is in good markets.

If global stock markets do a weak job of providing risk-reducing diversification, what are your options? Owning a basket of foreign currencies was the best answer that CIBC came up with, but this isn't really practical for the typical retail investor. Commodity futures, which you can buy into through exchange-traded funds, are another option.

The best all-around choice is bonds, which is obvious to anyone who has held them since the markets began to sink back in 2008.

"In normal down markets, central banks are usually cutting rates and so government bonds become a very good place to wait out the volatility," Mr. Grauman said. Now, however, interest rates are on the rise and bonds and bond funds, therefore, are vulnerable to price declines.

Not that you're seeing much sign of this right now as fear of falling stock markets drives money into bonds. But the outlook for bonds when rates are rising is negative, Mr. Grauman emphasized. "In this environment, positions in government bonds are much riskier."

A diversified portfolio of bonds is what you need, then. Government bonds because they're the preferred refuge when stocks are plunging, plus some corporate bonds. They offer higher yields, as well as a little more resistance to rising rates.

Some analysis by independent mutual fund analyst Dave Paterson of Paterson & Associates shows how well bonds work to provide non-correlated diversification. He took the 46 funds on the recommended list he created for his investment adviser clients and then examined how their returns tracked each other over the five years to April 30.

His view is that a correlation of 70 per cent - it might also be expressed in some cases as 0.7 - is the point at which investments are too similar to go well together. "If I can get it less than that, I'd be happy."

Let's zero in on TD Canadian Bond, the largest bond fund by far in this country and one of Mr. Paterson's chosen funds. It shows low correlation of no more than 26 per cent with each of the five Canadian equity and Canadian equity-focused mutual funds on the list. A particularly good match: CI Harbour, which has an 18-per-cent correlation with TD Canadian Bond.

Prefer a low-fee bond fund? PH&N Bond-D pairs up well with these same funds, although it's just a little more highly correlated to them.

Now let's add some global exposure to a portfolio with the Harbour Fund providing its core Canadian content. Mr. Paterson has five broad global equity funds on his recommended list, but only two partner well from a correlation point of view. One is Mackenzie Ivy Foreign Equity, with a brilliantly low correlation of 17 per cent to CI Harbour, and the Trimark Fund (SC version), with a higher but still OK correlation of 63 per cent.

Mackenzie Cundill Value, Mutual Discovery and AIM International Growth Class are all on Mr. Paterson's recommended list, but correlation analysis tells us they're too similar to the Harbour Fund to work well together in a portfolio.

A few of Mr. Paterson's recommended funds match up well with almost all names on his recommended list: PH&N High Yield Bond-D, TD Real Return Bond, Renaissance Global Health Care and Mackenzie Ivy Foreign Equity.

Of all the non-bond funds on Mr. Paterson's list, Ivy Foreign Equity is the outstanding name in low correlation. "It comes down to this fund's very conservative, value-driven style," Mr. Paterson. "And it still provides a decent level of returns."

Generally, he finds that global bonds have a low correlation to other fund categories, as does global health care. Technology can work well with Canadian-focused portfolios. Two categories with a high level of correlation are Canadian equity funds and emerging market funds. Unfortunately, correlation data is not widely available to investors. So ask your adviser about it, or try your own informal comparison of mutual funds, exchange-traded funds and stocks by looking at their year-by-year returns for at least the past five years.

It still makes sense to diversify your investments the old-fashioned way, but you can do more to protect yourself. Think correlation - it's the reality check on diversification.



Diversification meets correlation

After you've used traditional diversification on your portfolio, check out the correlation of the investments you end up with. Correlation measures the extent to which securities rise and fall together. The ideal portfolio is built with components that are non-correlated or, in other words, that don't all move in unison. Here are some portfolios of mutual funds that work well together from a correlation point of view. The funds were all drawn from the recommended list of independent fund analyst David Paterson.

We start with three popular Canadian equity or Canadian equity focused mutual funds that are on Mr. Paterson's List:



Correlation

Correlation

1. CI Harbour











Good partners include:

PH&N Bond -D

20%

Bad partners include:

Dynamic American Value

85%

Fidelity American Disciplined Eq

65%

Mackenzie Cundill Value

72%

Trimark Fund SC

63%

AGF Emerging Markets

82%

Renaissance Global Health Care

30%

CI Can-Am Small Cap

81%

2. Dynamic Value Fund of Canada





Good partners include:

Dynamic Cdn Bond

11%

Bad partners include:

Dynamic American Value

86%

TD Mortgage

5%

AIM International Growth

72%

CI American Value

54%

CI Signature Cdn Resource

89%

Trimark Fund SC

54%

BMO Guardian Asian G and I

73%

3. Fidelity Canadian Large Cap A

Good partners include:

TD Canadian Bond

24%

Bad partners include:

RBC O'Shaughnessy U.S. Value

73%

TD Real Return Bond

42%

Mackenzie Cundill Recovery

80%

Trimark U.S. Small Companies

44%

Harbour Growth & Income

92%

Mackenzie Ivy Foreign Equity

11%

AGF Canadian Small Cap

85%

Correlation Key

100%

two securities move identically

0

non-correlated, which means there's no connection

-100%

securities move oppositely

Source: David Paterson