March 26, 2006

Poor fund classifications
CIFSC missing the mark

The Globe and Mail recently reported that Morningstar Canada plans to leave the Canadian Investment Funds Standards Committee. Morningstar CEO Scott Mackenzie, a founding member of CIFSC, cited the group's ineffectiveness and slow response to the fund universe's evolution. Morningstar, instead, plans to work on a better classification system. I applaud their decision.

Misclassification

I can't quantify how many times I've encountered a fund that was assigned to a category that made no sense to me. Here are a couple of examples.

Standard Life Canadian Small Cap is one of my favourite small cap funds. It's listed in the Canadian Equity (Pure) category merely because the average market capitalization of its stock portfolio is $1.3 billion. This is based on market value and puts it slightly out of reach of the Canadian Small Cap category definition.

Yet, another 'pure' Canadian fund - Mawer New Canada - has an average market cap of $0.9 billion. Mawer New Canada is classified as Canadian Small Cap. But because of the definitions, these two funds are placed in vastly different categories.

The funds in the Canadian Equity (Pure) category sport a median market cap of about $23 billion - about the same as the S&P/TSX Composite index. (I've included seg funds in this ranking.) The average market cap of funds in this class range from a high of $37.9 billion to a low of $0.3 billion, according to PALTrak.

If we look at the Canadian Small Cap category, the median market cap of its constituent funds is $0.9 billion (with a high of $9.5 billion and a low of $0.1 billion). As a result, Standard Life Canadian Small Cap is compared to what basically is a large cap category while the Mawer New Canada resides with fellow small cap funds. This makes no sense.

Another is the High Yield Bond category. Not only is this class home to both domestic and foreign funds (a big difference over the past four years) but it also contains a mix of mandates - from funds with lots of government bonds to those that hold nothing but dicier junk bonds. Again, this is nonsensical.

Poor class definitions

Part of the problem comes down to weak definitions. Following the above examples, inclusion in the Canadian Small Cap class requires at least half of fund assets (and 3/4 of non-cash assets) to be held in Canadian stocks. Also, the weighted average market cap of a fund's stock holdings must be equal to or less than 0.1% of the 'adjusted' market cap of the S&P/TSX Composite Index. This presents two problems.

First, there is no direct tie to investment policy. It's true that looking at what a manager does is as important as (perhaps more so than) looking at what s/he says. However, this measure is bound to fail in an environment like the one that has prevailed recently. Most small cap funds will only buy stocks when they're 'small caps'. The definition of 'small cap' varies, but virtually all such funds buy stocks only when the market caps is under $1 billion (some have lower thresholds). But if the fund is successful, that number - on a market value basis - will rise. This is particularly true when small caps are outperforming large caps - as they have for most of the last six years.

Second, the basis of CIFSC's market cap threshold, the S&P/TSX Composite, only accounts for 69% of the total market cap of the TSX. The S&P/TSX Composite boasts an adjusted market cap of $1.175 trillion, compared to the $1.696 trillion market cap of all domestic issues listed on the TSX (as of the end of 2005). So, the CIFSC's Canadian Small Cap definition ignores nearly 1/3rd of the value of TSX-listed domestic issues. This difference would put funds like Standard Life Canadian Small Cap into the proper category.

In the High Yield Bond category, you'll find some pure (primarily Canadian) high yield bond funds like Dynamic Canadian High Yield Bond. But you'll also find PH&N Canadian High Yield Bond, which has a more flexible mandate. Sure the PH&N fund invests all in corporate bonds but, as of the end of 2005, less than 20% was in high yield bonds.

A better way

Morningstar has stated that it will endeavour to create its own set of fund classifications (which it did prior to CIFSC's birth). A combination of better categories and improved definitions will help to ensure that funds are better classified. More categories are needed.

Back in 1998, when I oversaw a large mutual fund database and had responsibility for proper classifications, we had more than fifty different categories. Eight years later and with a much larger fund universe, the CIFSC categories total less than three dozen. So, clearly a longer list of fund classes is needed. And that will help better define the categories, which will result in overall better classifications.

No system will be perfect since there will always be funds with very unique and unusually flexible mandates. But there is lots of room for improvement on the status quo. Sure, two sets of categories might cause some confusion but I expect that any new set of categories and definitions put out by Morningstar will be a step forward. And I'm all for improvement.