What makes a good core holding?

Does it capture a broad market segment?

One commonly held definition of a core holding is that the investment provides broad exposure to a large part of the market and gives you a lot of diversification in a single shot. Broad-market equity index funds like TD Canadian Index and RBC Canadian Index, for example, are commonly accepted as core holdings because they mirror the performance of broad market indexes. Many large-cap actively managed funds also fall under the core heading for a similar reason--when Canadian stocks in aggregate are up, they will be too, give or take a few percentage points.

On the flip side, investments that focus on and capture the performance of just one or two market sectors aren't considered "core" under this framework. Because they're less diversified, they may perform spectacularly well at times when broad-market performance is just ho-hum, but they may dramatically underperform in other market environments.

This definition of core versus noncore is useful in bond-land. Broad-market bond indexes like the DEX Universe Bond Index and most funds in Morningstar's short-, intermediate-, and long-term bond categories emphasize high-quality bonds, which usually hold their ground much better than high-yield bonds in recessionary environments.

Given that many, if not most, fixed-income investors are looking to bonds for downside protection and to diversify their stock holdings, it makes sense for them to seek out core bond funds that are anchored in high-quality bonds.

By contrast, a bond fund that is not considered a core holding is TD High Yield Bond  , which generally eschews the highest-quality bonds and instead puts a big emphasis on mid- and lower-quality corporate bonds. As a result of that aggressive strategy, its performance will often be dramatically different than broad bond-market index funds, and its potential for real losses will be much higher.

A statistic called R-squared can be a helpful gauge of an investment's correlation with a broad market index for its asset class. Many investments we often think of as core pass the sniff test by this measure. For example, Brandes Sionna Canadian Equity  , a Morningstar Fund Analyst Pick, has an R-squared with the S&P/TSX Composite Index of 95 during the past three years, meaning that 95% of its performance pattern can be explained by movements in the index. (Note that a high R-squared doesn't necessarily indicate an investment is a closet indexer, as evidenced by high measures at truly active funds like CI Harbour  .)

I think it's core, therefore it is

However, there's a big problem with strictly defining what is core by its level of market correlation. Many great investors -- in fact, I'd argue, most great investors -- couldn't give a hoot about whether their portfolio or performance tracks that of a broad market benchmark. Instead, they go where the opportunities beckon, regardless of whether they end up heavily skewed toward a single market segment or not.

Such investors' performance may veer significantly from that of broad market indexes at various points in time -- sometimes for better, sometimes for worse. They know that getting ahead of the market over the long haul is what matters, and most would probably argue that concepts like R-squared and are the domain of money managers, not real investors.

Warren Buffett's Berkshire Hathaway BRK.B provides a vivid case in point. During the past five calendar years, the fund has trailed the S&P 500 by three percentage points (2005), beaten it by 11, 25, and six percentage points (2006, 2007, and 2008), and lagged it by a stunning 21 percentage points (2009). But no one could argue with a straight face that Berkshire hasn't been a fabulous core holding for investors. (Incidentally, Berkshire stock is having another great performance run this year.)

The same holds true for scores of great mutual funds, from Mackenzie Cundill Canadian Security Series   (which has an R-squared with the S&P/TSX Composite Index of 55) to TD Dividend Growth   (R-squared of 79). Their managers actively focus on exploiting opportunities that other investors are missing, not simply matching the market over the past 10 years. Due to that mindset, they've been able to deliver performance that's different, and better, than the market. Those funds clearly demonstrate that an investor who employs the "core equals broad-market exposure" mindset will be ignoring some solid offerings. They also prove that low market correlation doesn't have to equal above-market risk.

Bottom line

As I mentioned at the outset, there are no strict definitions about what constitutes a core holding. Both index trackers and out-of-the-box funds -- such as CI Signature High Income  , which hunts far and wide for cheap stocks and, in some cases, bonds -- can make worthy cores. The key is to understand what definition you're using, and in turn use that to inform what expectations you have for your holdings.

For example, it's OK if your definition of "core" encompasses very active investments. But making good use of such holdings also requires a thorough understanding of the strategies they employ and when they're likely to out- and underperform. If you approach them with the proper set of expectations, you'll be much more inclined to sit tight or even add more when they're down.

This is also one of those areas where you don't have to be dogmatic. Investors can find the best of both worlds by holding truly active core funds -- or picking individual stocks -- then anchoring them with low-cost offerings that provide broad market exposure.

--With files from Ashley Redmond and Esko Mickels