Hi Guys,
The negative correlation between mutual fund end performance and costs is well documented. This single graph tells the whole story nicely:
https://im.mstar.com/im/newhomepage/chartofweek_mutualfundcompanies_040215.pngCosts just don’t matter. They matter greatly. Control the mutual fund costs and enhance your payday. The size of each data point is a measure of the money being managed. Dodge and Cox is a winner by the graph measurement scales.
Best Wishes
Comments
OJ
PS: How did you manage to find a subject that didn't involve Monte Carlo simulations?
Thanks. It was a challenge and I was somewhat off my game. I’ll be careful not to let it happen too often.
Very Best Wishes
Please keep your discovery a secret. I don’t want my enemies to find this out. Wait a minute! I don’t want my friends to find this out either!!!!!
Note my familiar hello greeting. What’s happening to me? It can’t be good, or can it? Far to many questions!
Best Wishes
Of course it has to be true that all else being equal cheaper is better. But all else is virtually never equal, and there's a lot being glossed over in this one chart.
* It's comparing apples and oranges - asset weighted performance vs. unweighted share classes:
A fund could have five expensive share classes with few assets and one cheap share class; the performance could look good on an asset weighted basis because of the cheap share class but the fund family would look expensive because of all those "empty" expensive share classes.
Note also that while a load share class might be considered expensive on an absolute basis it could still be rated average or below average in cost. That's because M* groups share classes by type: load, institutional, no load, before ranking their costs as relatively high, average, or low.
For example, LCEVX, a LCV fund with an ER of 1.56% is said to be "below average" in expenses. Its sibling classes include LCEAX, ER 0.81%, called "low", and LCEIX, ER 0.76% called merely "below average" because NL shares are expected to have lower costs than their loaded brethren. (I'm not faulting M* here for how it evaluates costs; just pointing out what's going on beneath this chart.)
* Load families tend to make costs look less relevant. As noted above, A shares can be counted as "cheap" even as their costs drag down their performance.
* At least according to papers I've read, the correlation between costs and performance is stronger for bond funds than for equity funds. That could skew the per-family data, since some families specialize in bonds, while others have more assets in equities.
* How meaningful is data about share classes for families with very few (say, five) share classes total? As opposed to one with hundreds of share classes.
There's the usual caveat of relying on stale data. Here's the source of the chart, from four years ago.
https://www.morningstar.com/videos/691300/the-clear-link-between-fees-and-performance.html
Quoting from that page: If you to play with more extensive data, there's the Morningstar Fund Family 150 (Jan 1, 2019): "a semiannual publication that gives investors access to the same analytical rigor our own analysts use to keep tabs on the 150 largest fund families in the United States."
Highlights: https://www.morningstar.com/blog/2019/02/22/top-fund-families.html
Full paper: https://morningstardirect.morningstar.com/clientcomm/DueDiligenceReports/FundFamily150.pdf
Spreadsheet data: https://morningstardirect.morningstar.com/clientcomm/DueDiligenceReports/FundFamily150_2018_Q4.xlsm