Are new fund mergers biased?
Recent mergers drive by consolidation

There are too many mutual funds available for sale. In my opinion, there are too many funds for some asset classes (i.e. Canadian equity) and also too many bad funds. As the industry players offering these products continue to merge, many redundant products are merged into the history books. Traditionally, only poor performers were 'rationalized' but the more recent trend isn't so clear. Will the past bias in performance persist going forward as fund merger motivations shift?

Survivor bias

Many funds are launched, only to be subsequently wound up or merged into another. A glance at the old January 1995 13-Year Mutual Fund Tables from the Financial Times of Canada illustrates just how much the industry has grown, and how many products no longer exist.

Scanning the tables brings back such names as AIC Emerging Markets, Gyro Equity, Hyperion Aurora Trust, Ivy Capital Protection, Polymetric Performance (Admax), and Regent Dragon 888. I know what happened to some of the old funds. For instance, Admax and Regent funds eventually ended up part of AIM, which is now AIM-Trimark Investments. However, others don't even ring a bell. Surely some of them were poor performers.

The tasty sounding Gyro Equity (a Canadian stock fund) beat what was then the TSE 300 in four of nine calendar years but did so in very volatile fashion. Also, neither Admax nor Regent funds were known as strong performers.

Hence, these dogs no longer show up in most mutual fund averages, which look better today than they would if the 'dead' underperformers had been retained in the 'average' record. This is the bias. Including 'dead' funds would likely result in lower average or median fund performance.

While John Bogle estimates this bias to be on the order of 1% per year, others like Morningstar U.S. and Micropal have found no significant bias. It's fair to assume that some bias exists. While I've never studied the topic, I'd peg the Canadian bias at higher than 0.5% per year but less than 1% - but that's just an intuitive guess.

Recent mergers

It seems clear that yesterday's mergers often ridded the industry of 'hall of shame' members. But today, mergers are happening - at least in part - for different reasons. As noted, a flurry of fund company mergers has caused firms to 'rationalize' their bulging product lines.

For instance, after last year's acquisition of Synergy Funds, CI Funds recently announced the completion of seven fund mergers. Three of the funds - CI Canadian Stock, CI World Equity, and Synergy Canadian Growth Class - have nicely outperformed their respective benchmarks over the past five years. However, some of their longer-term numbers aren't as pretty. Plus, the other four funds have spotty performance records.

The existence of bias, if any, may depend on the period of measurement. For instance, CI American Growth - one of the merged funds - has a poor record over the past five years but is strong over ten, fifteen, and twenty-five years. While it trailed the S&P 500 C$ benchmark, it handily beat the median U.S. stock fund over most periods.

Many survivorship studies cover ten-year periods but a true study would have to cover longer periods to ensure a complete impact of losing track records is considered.

My sense is that some bias will remain as funds are merged but it should be less significant going forward since some good track records are disappearing. However, being aware that such a bias exists is key when looking at average and median fund figures. On the other hand, I confirmed recently that Morningstar fund averages and medians are free from survivorship bias - hence, a useful research tool.