Some market analysts have signed diversification’s death certificate.They claim 2008’s meltdown showed searching for negatively correlated assets is a fool’s errand. Traditionally divergent asset classes, they argue, are now tightly correlated.

But while market conditions are complex, the ever-present threat of overconcentration can still be staved off with sound judgement and the right analytical framework.

Understanding the problem

Canadian investors’ tendency toward home bias makes them particularly vulnerable to overconcentration, says James Price, director of fixed income for Macquarie Private Wealth.

He cites the past year as an example of failing to look beyond our borders.The S&P 500 gained 13% in 2012, while the TSX posted a sluggish 3%.

Nor is it enough to pull together holdings from several asset classes. The appearance of diversification can conceal high correlations, says Neil McIver, director of wealth management and portfolio manager, McIver Wealth Management, Richardson GMP. Say you’re building a portfolio, and pick one of the big banks. Then you add some preferred shares, such as Bell and Brookfield Properties. For your anchor, a 10-year Government of Canada bond.

This may appear diversified, but McIver says a correlation analysis would show it’s actually highly concentrated. Rising interest rates cut into borrowing, hurting the big bank’s profitability. The preferred shares trade like 30-year bonds, so a 1%interest rate increase reduces their value by 30%. It will also bring down a 10-year bond’s value by 10%.

It gets worse. Correlations also have tightened between national markets, Price says, because they’re all reacting to the same global macroeconomic pressures.

The 2008 meltdown, and its lingering aftermath, is the best example, says Stephen Lingard, director of research and portfolio manager, Franklin Templeton Multi-Asset Strategies. “That heightened fear causes a similar reaction across markets and asset classes—there’s really no place to hide.”

Buying some U.S. and European equities and adding alternatives like commodities or hedge funds used to work, “but since the global financial crisis we’ve seen correlations on those asset classes get closer to 1,” Price explains.

He adds commodity ETFs — which have widened access to assets that were once hard to get—have caused higher correlations with traditional equity asset classes.

The news isn’t all bad, though. “There are still lots of products that can make a portfolio diversified. A perfect example is sovereign bonds of non-troubled nations like the U.S.,Canada,Australia and Germany. They’ve been inversely correlated with equities for the past five years.”

Lingard adds, “Some bond investors got into trouble reaching for yield in 2008, thinking high-yield or low-quality corporate bonds were the same as government bonds. The only things that held or went up in value were long government bonds, since interest rates collapsed.

“So even in a shocking scenario like 2008, there were still some great rewards from diversification,” he says.

Lingard also suggests sky-high correlations are largely cyclical. “I won’t say we’re going back to the way it was 50 years ago, when markets were separate and distinct. But cyclically there will be falling correlations. The longer we go without a major global shock, the more correlations between asset classes and markets will recede,” he explains.

The pieces

The process begins with an assessment of past, current and projected market conditions. “You have to understand where we are in the cycle. The last great bull market, from 1982 to 2000, rewarded risk. From 2000 to today, risk has not been rewarded and we probably have between two and five years to go in this current bear market,” McIver explains.

McIver chooses among 40 asset classes — Canadian equities, U.S. large-cap growth, U.S. small-cap value, international equities, etc. Based on his assessment, seven or eight asset classes make the cut.

With the aid of modelling software, he’ll then determine how to weight each asset class. Currently he’s cool on international equities, and restricts that class to a maximum of 12%. The software helps determine if the percentage weights he’s stipulated for each asset class make the portfolio efficient. If not, he tweaks the percentages until the numbers fit.

The next analysis involves populating each slice of the portfolio with individual securities.

Arriving at a potential buy list for each asset class requires looking at P/E ratios, debt-service coverage ratios and other analytical metrics. He narrows the list through technical analysis, which looks at where a company is in the business cycle.

Final cuts are based on a correlation analysis of individual securities. Say Bell and RBC make it this far.A correlation analysis may reveal “these companies moving together like dolphins,” McIver says. So one has to go.

Big banks and gold companies, on the other hand, can provide negative correlation even in the current environment. “Companies like Yamana Gold and the big banks move in opposite directions in response to inflation and interest rates,” he notes.

But which big bank is on the final buy list? And on the gold side, will it be Yamana, Barrick or another player? The tie-breaker will emerge in the technical and fundamental analysis.

“Volatility is what kills portfolios. So the lack of volatility over time actually increases the portfolio’s rate of return,” McIver says, adding that “diversification as a measure of safety is an old concept. It’s actually for performance.”

5 Model Portfolios

Portfolio 1 fits investors who can be expected to use up their capital during their lifetimes. They have fairly high, ongoing income needs, and need protection from market fluctuations.

“A typical example would be a retired couple that sold a business. They have fairly small RRSP holdings and about $1 million to invest,” McIver explains.

Most of McIver’s clients fit Portfolio 3. These are clients with $3 million-to-$5 million to invest who own businesses or are professionals. They may also have real estate, like additional residences or vacation properties.

“They can be more aggressively invested because they’re backstopped by income or assets not held in the portfolio,” McIver explains.

Portfolio 5 is reserved for investors with a wide range of assets and more than $10 million in their portfolios. “They aren’t going to come close to spending the capital that they’ve accumulated — in their lifetimes or their children’s lifetimes. It’s really for the grandkids,” McIver explains, adding these investors have lived through market corrections.

And some clients may be suited on paper to a higher-risk portfolio, but on an emotional level can only handle Portfolio 2.

He also notes these portfolios are baseline allocations, which he alters on a tactical basis if warranted. The entire portfolio is reviewed every 90 days, but each component is monitored daily.

“Each slice of those pie charts is on our smartphones. If they move more than one or two percentage points on a given day, we get a alarm. Then the investment committee has a conversation about whether to punt on a position if it becomes too risky, or do the opposite if there is an opportunity,” McIver explains.