In my never-ending quest to find data with predictive
value, I like to put new data points to the test. The
latest is encouraging. Morningstar’s data team has
been gathering brokerage cost information from
mutual funds’ Statements of Additional Information,
also known as
485BPOS
. The figures are reported
in dollar terms, and we then divide by average net
assets to understand what percentage of assets are
going to brokerage commissions.
Funds incur brokerage commissions when they trade
stocks. Because most fund companies trade huge
sums of money, they are able to negotiate low com-
missions, but they still add up and they are not
included expense ratios. (We’re not talking about
sales loads, which are fees you pay a broker to
buy a mutual fund.)
Brokerage commissions themselves are fairly modest
costs that are typically well below a fund’s ex-
pense ratio. For example,
Fidelity Value
FDVLX
ran
up about
0
.
11%
in brokerage commissions while
charging an expense ratio of
0
.
68%
. If trading costs
were limited to those
11
basis points, then there
wouldn’t be much information there. However, as
trade cost consultants love to say, they are just
the tip of the iceberg.
Consultants and academics have argued that trading
costs substantially exceed reported brokerage
expenses. Funds drive up the price of stocks as they
are buying and drive them down as they are selling.
In addition, sometimes funds fail to buy as many
shares as they want to. Another problem occurs when
a fund manager trades stocks simply to accommo-
date flows rather than because the manager has an
information advantage.
So, if brokerage commissions are an indicator of over-
all trading costs, then they would be quite useful
as a predictor of future performance. You may recall I
put turnover ratios to the test last year for similar
reasons. The results were that turnover wasn’t much
help. They showed very little difference in returns
and success rates from the lowest-turnover funds to
the second-highest quintile. Only the highest quintile
of turnover showed any sign of impairment.
The results on brokerage costs are much better,
although we have data only going back to end of
2008
,
so I had a smaller time period to test. We’ve posted
a
PDF
file at
mfi.morningstar.com
that shows the bro-
kerage costs for all the U.S. equity funds in the
Morningstar
500
along with their percentile ranks so
that you can see how they stack up.
How We Ran the Data
We grouped funds into quintiles based on brokerage
costs within their categories and then examined
four-year returns on a pre-expense ratio basis. We
backed out expenses in order to isolate brokerage
commissions. Then we calculated success rates for
each quintile. The success rate tells you what per-
centage survived and outperformed peers.
What We Found
Our brokerage-cost figures indicate this is a promising
area. Funds in the lowest brokerage-cost quintile
FundInvestor
Continued on Page 2
Research and recommendatio s for the s rious fund investor
Brokerage Costs Signal
Funds’ True Cost Hurdle
Fund Reports
4
T. Rowe Price New Asia
PIMCO All Asset All Authority
Artisan Global Opportunities
T. Rowe Price Balanced
Morningstar Research
8
SRI Funds Don’t Have an Edge
The Contrarian
10
A Bad Year for Active
and Passive?
Red Flags
11
Avoid the Loved
Market Overview
12
Leaders & Laggards
13
Manager Changes and News
14
FundInvestor Focused 10
16
Tracking Morningstar
18
Analyst Ratings
Income Strategist
20
FundInvestor 500
22
FundInvestor 500 Spotlight
23
Follow Russ on Twitter
@RussKinnel
RusselKinnel,
Director of FundResearch and Editor
SM
February
2013
Volume
21
Number
6
