|
Screening
for Mutual Funds
Part
2, Advanced Tools
Alas,
like so many great Web investing sites, Wall Street City is gone.
In my last
column I described a strategy for finding mutual fund candidates using
Morningstar’s free mutual fund screener (www.morningstar.com).
That screen relied mostly on Morningstar’s risk and star-ratings,
along with five-year returns, to pick funds worth researching.
In this
column, I’ll describe advanced fund screening features available, to
my knowledge, only on Wall Street City’s investing site (www.wallstreetcity.com).
Many of WSC’s features require a subscription, but its Mutual Fund
Search is free, although you must register to use it.
Wall Street
City’s fund search offers two advantages over other screeners. First,
it provides search parameters that are, in my view, more sophisticated
than generally available. Even more significant, the program shows you
the returns you would have achieved had you run the same screen up to
one-year ago, and bought the funds turned up by the screen back then.
That feature,
termed backtesting, gives you the opportunity to test different
screening strategies without risking real money.
The program
provides more than 50 search parameters. I don’t have room to discuss
them all, so I’ll describe the ones I found most interesting. I took
advantage of Wall Street City’s backtesting feature to see how I would
have done it I had run each screen a year ago and bought the top 25
funds turned up by each screen.
If you want
to follow along with me, get there from Wall Street City’s homepage
by selecting ProSearch
(top menu), then ProSearch
Engines (left-menu), and finally, Mutual
Fund Search.
The search is
easy to use. Simply click on a category such as “General
Characteristics” in the search window, and then click on a search
parameter such as “Load Charge Maximum” and fill in the desired
value in the window on the right. In most cases the choices are, “high
as possible,” “low as possible,” or you can specify minimum and
maximum values.
For all
tests, I selected zero maximum load charge and $5,000 maximum “initial
minimum purchase requirement” to limit the field to no-load funds
available to individual investors.
Comparison
Search
To establish a standard of comparison, I started with a set of widely
available screening parameters. I selected “high as possible” for
average annual five-year return, “low as possible” for standard
deviation (risk measure) and a 25 maximum portfolio price/earnings ratio
to further limit the risk.
The search
program lists the top 25 funds meeting the screening requirements. It
shows you would have done had you run the screen a year ago and bought
all the funds in a chart labeled “how has this search been working?”
near the bottom. The 25 funds yielded a 20 percent return, on average,
up to the day I ran the test. For comparison, the market as measured by
the S&P 500 index returned around 17 percent.
Market
Probability Improves Results
I improved on that screen’s results slightly by dumping the five-year
return, standard deviation, and maximum P/E parameters in favor of a
parameter called “market probability,” which measures performance
consistency rather than raw performance. For example, a probability
score of 50 means that the fund outperformed the market only half the
time, while a score of 100 means that the fund outperformed the market
in every quarter measured.
I selected
“five-years” from the range of three- to 15-year timeframes
available for market probability. The 25 funds that would have been
picked by the screen 12-months earlier returned around 25 percent,
according to Wall Street City’s backtest chart.
Alpha Is
Even Better
I scored a bigger improvement by replacing market probability with
“alpha,” a statistical measure of a fund’s risk-adjusted return.
The definition is complex. Basically, funds with high alphas have
produced better returns than would be predicted by their historical
volatility (beta). I picked alpha measured over five years for the test
from the one- to 15-year available ranges.
Replacing
market probability with alpha upped my 12-month backtested return to 38
percent.
Best
Results
I achieved my best 12-month backtested returns, around 50 percent, using
the “up market premium” parameter, which gauges how much a fund
outperformed the index, only counting quarters when the S&P 500
delivered positive returns.
I tried a few
parameter combinations such as high “up market premium” and low beta
(risk measure), but I wasn’t able to improve upon the results obtained
using “up market premium” alone. However I came nowhere near to
testing all of the possible combinations.
Wall Street
City’s backtesting feature adds a new dimension to screening, but
it’s not the Holy Grail. Market conditions change; so last year’s
winning strategy may not produce the same results under today’s
circumstances.
Nevertheless,
WSC’s advanced features can help you pick better mutual funds. As with
all screens, consider the results to be research candidates, not a buy
list.
published 10/5/03 &
10/19/03 |
