Misleading with statistics

I'm sure you've heard the old saying about liars, damn liars and statistical liars.
I wouldn't go so far as to call the folks at the Investment Company Institute liars, so let's just call them deceivers -- statistical deceivers.
The offense came in October when the mutual fund trade group issued its latest study of mutual fund costs. As in the past, the ICI implied that its members deserve a big pat on the back, proclaiming that "the shareholder cost of investing in each major type of mutual fund continued to decline between 1998 and 2001. ..."
The institute, which represents fund companies, not investors, said the cost of investing in stock funds dropped 5 percent over that period, bringing the total decline since 1980 to an impressive 43 percent.
Over the two periods, bond-fund costs have fallen by 17 percent and 41 percent, it said.
But it's not that simple.
"Cost" refers to things such as management fees and sales commissions called loads. It covers everything from the cost of employing stock pickers and paying for mailings to customers, to the fund company's profits. The lower the cost the better.
Investors are more familiar with the term "expense ratio." A fund that charges you $1.50 in expenses for every $100 you have invested has a 1.5 percent expense ratio.
Misleading
I have no reason to think the institute's figures are wrong, but clearly they are misleading.
That's because the institute gave more weight to funds that took in more investment dollars. Hence, a low-cost fund that took in lots of money would pull the average cost down, while a high-cost fund that took in a small amount of money would do little to hold the average up.
The institute's system produced a lower cost figure than you would get by simply averaging the costs of those two funds.
Moreover, by looking only at new fund investments, the institute ignores the costs borne by investors holding shares they'd bought in previous years.
By simply averaging each fund's costs in the standard way, fund-tracking company Lipper Inc. says expense ratios last year averaged 1.57 percent for stock funds vs. the institute's 1.28 percent. And Lipper says costs have been going up, not down.
Stock funds charged 0.94 percent in 1980 and 1.5 percent in 1988, while bond funds charged 0.77 percent in 1980, 0.93 percent in 1988 and 0.972 percent in 2001, Lipper says.
That's quite different from the declines of more than 40 percent that the institute reports for 1980 through 2001.
Loads not included
Lipper's cost figures would be even higher if the company accounted for loads, found most often in funds sold by brokers.
In fact, it is investors, not fund companies, who are behind the cost declines the institute claims. Investors have shown a preference for low-cost no-load funds, giving those funds more weight in the institute's calculations, the institute says.
Am I quibbling? No. Fund-industry critics have quite rightly pointed out that costs have not come down, as they should have given the greater efficiencies companies have enjoyed as funds soared in size in the 1990s.
Also, costs that look small can deeply damage fund returns in the long run. If you invested $5,000 a year for 30 years at 7 percent a year, you'd end up with $472,000. But if bigger costs reduced your return to 6 percent, you'd finish with $395,000.
Costs are especially corrosive when fund returns are low. An investor can be excused for ignoring a 2 percent expense ratio when a fund is returning 15 percent or 20 percent a year, as many did in the 1990s. But if you're making only 6 percent a year -- a more likely figure for the next few years -- those costs chew up a third of your gains.
Many investors are getting the message -- explaining their growing preference for low-cost funds. They -- you, I mean -- deserve a hearty congratulations.
And the Investment Company Institute? For it, a big raspberry.
XJeff Brown is a business columnist for The Philadelphia Inquirer. E-mail him at jeff.brown@phillynews.com.