In the first article in this series, we looked at the impact of fees and charges on investment returns. This time we're going to see what the cost of investing in a particular fund can tell us about its likely future performance.
It’s a curious irony of the investing industry that, despite constant reminders that past performance is not a reliable indicator of future results, many investors (and alas, a sizeable proportion of financial professionals) continue to behave as if it is.
There’s a very good reason why regulators insist on such warnings being included in marketing material. Past performance, especially a track record shorter than ten years, tells us next to nothing about how a particular fund will perform in the future.
If anything, a very strong record over, say, two, three or five years, could be seen as an indicator that a fund is due for a period of underperformance.
The most reliable predictor
What, then, really does give us a clue as to future performance? Academics and other serious researchers have consistently found that the most reliable predictor of all is cost; more precisely, the more the investor pays, the lower their net returns are likely to be.
Researchers at Morningstar have conducted detailed ongoing research on this issue, for which they divide funds into five quintiles based on how much they cost. They also separate funds into two groups — “successful” ones, i.e. those that survived for the whole of the period in question and outperformed their peers, and unsuccessful funds, i.e. funds that were either liquidated or merged with other funds, or underperformed their peers.
Their most recent analysis showed that, in the five years to the end of 2015, funds in the cheapest quintile were three times more likely to succeed than those in the most expensive quintile. They found, moreover, that using expense ratios to choose funds helped in every asset class and in every quintile from 2010 to 2015.
You may at this point be wondering about the impact of falling fees. Ongoing charges figures have indeed been coming down. UK investors can now access ETFs for just a few basis points, and further fee reductions are likely across the industry.
As fees fall, you might have thought that cost would become less important as a factor in predicting future returns. Analysts at Morningstar have recently addressed this question too, and found, perhaps surprisingly, that the opposite is true.
Funds are performing more alike
Researchers examined the rolling returns and fees of all domestic US equity funds over the period from September 1998 to August 2018, including those that didn’t survive the full 20 years.
They started by measuring the return difference between the best-performing funds and average-performing funds, and also between the best performers and the worst performers. What they found was a significant narrowing of the performance gaps between funds. In the early 2002, for instance, it wasn’t unusual for there to be differences in performance of 10% or more. In recent years, however, returns have been much more closely bunched together.
Fee differences have remained consistent
Next the researchers measured the difference in fees between the cheapest and the costliest funds. For the five-year period to the end of August 2003, the difference in annual expenses between the average fund in the cheapest quintile and the average fund in the third-cheapest quintile was 0.64%. For the five-year period ending on 31st August 2018, the figure was only slightly smaller, 0.60%.
Similarly, there was very little difference in the average fees charged by the cheapest funds and the most expensive funds. For the 60-month period up to the end of August 2003, the difference was 1.63%. For the five-year period to the end of August 2018, the difference was only slightly lower,1.56%. In other words, fees have come down fairly consistently across the board.
The relationship between fees and performance
Now for the crucial part. The difference in returns, we’ve established, is much narrower now than it used to be, and the difference in fees is only very slightly narrower. What, then, can we conclude about the connection between fees and performance?
To work that out, Morningstar calculated the average fees charged over time by funds across all five return quintiles, and then compared the differences in their fees to the differences in their returns.
The conclusion the researchers came to was that cost has not become any less important as a contributory factor in the outperformance delivered by the best-performing funds period over the 20-year period. On the contrary, expense ratios have become far more important.
Over the entire period, the researchers calculated, the average fee advantage between the top-performing and third-highest-performing quintiles was around 1% of the overall return advantage. But for the five years to the end of August 2018, it accounted for around 10% of the return advantage.
The lesson for advisers
So, what can we take away from this latest research?
To quote Morningstar’s global research director Jeffrey Ptak, “fee differences appear to account for an even greater share of performance differences than before, suggesting that investors are well advised to continue to factor cost heavily into their fund assessments.”
Identifying, ex ante, the relatively few funds that are going to outperform the market over the long term is a very difficult task.
Quantifying costs is altogether easier. True, we don’t know what the transaction costs are going to be, but we can come up with an estimate based on what they’ve been in the past. And we do know in advance the annual management charge.
In the absence of any more reliable predictors of future fund performance, then, advisers who genuinely want to act in their clients’ best interests should focus on cost.
Next time: How do you work out how much your client will pay to invest in a particular fund?