Fund firms find ways to fudge numbers

He insists fund companies don't actually do much wrong because they follow the rules, and they'd get in trouble if they violated those standards.

While he's right from a legal standpoint, he ignores the simple truth that the rules leave fund companies a lot of ways to fudge the statistics. If these numbers factor into your investment decisions, look at their meaning more closely:

• Past performance, Part 1: The candy of the mutual fund world, past performance is where a fund "tastes great" and there are no consequences of indulging. Fund executives publish fine-print warnings that past results are not a reliable indicator of what to expect going forward, but that's always below the big-type hype using those results as a big reason why you should buy a fund now.

• Past performance, Part 2: Some funds achieve their record the old-fashioned way, through shenanigans and financial engineering. Fund companies routinely merge away their bad track records. If XYZ Growth is a laggard but XYZ Large-Cap - run by the same management team - has reasonable performance, the growth fund will get the ax, and the strong record survives. Never mind that many investors had a lesser experience - or that management has shown an ability to underperform - the snapshot view looks good.

• Past performance, Part 3: The long-term annualized average record looks good but ignores the question "What have you done for me lately?" Some funds live off of great past performance; they haven't been solid performers for years, but big numbers produced in the more-distant past make them look solid.

• Average costs: While there is no guarantee that cheap management is good management, costs matter. That's why many investors set their cost barometer based on the average cost for the type of fund they are buying. For a stock fund, the average expense ratio is roughly 1.3 percent. It's roughly 1 percent on bond funds.

In general, investors think that "below average" is sufficient. What they don't know is that the average is skewed dramatically by the way an average is calculated.

Using a "dollar-weighted average" - so that a fund with $10 billion in assets affects the average more than a fund with just $10 million in it - drops the "average" expense ratio significantly, so that the typical costs for investors in stock funds drops below 1.0 percent (that's good, because it means investors gravitate to low-cost funds).

• Returns aren't adjusted for taxes: The fund company doesn't pay Uncle Sam, but you do. Funds tell you what they earned, when what's most important is what you get to keep.

• Time-weighted performance measurement: This boils down to "your mileage may vary." The typical pattern for a hot mutual fund is that the assets flow in only after a period of great performance; in other words, investors tend to "buy high." If the fund suffers thereafter and the shareholder bails out, they have sold low. Meanwhile, the average performance numbers can continue to look pretty good.