Expense
Expenses are 100% certain while past performance IS NOT an indication of future results. Is there any evidence that
expenses are predictive? The truth is, NOTHING is predictive for picking winners. However, expenses do tend to have some
statistical connection, and a not surprising relationship to keeping the odds of getting burned minimized. This statistical
connection and relationship of expenses to results has been documented in numerous academic and industry studies.
The problem is in how most grading systems grade funds. Normally, funds are graded primarily or exclusively on their
returns as compared to what someone deems as a "peer" data point instead of relative to an asset class you might be selecting
the fund to represent in your overall strategic asset allocation.
ALL of our grading routines are free from the bias that may be introduced by "peer group" or "universe" ratings (other
than the routine for expenses which is covered more fully below). It is a well known fact that in most fund ranking systems
the best way to get a good score is to be misclassified into the wrong "peer group." For example, look at ANY other grading/rating
system out there and compare a total market index fund to an S&P 500 index fund. If the data covers three or more years, the total
market index fund will probably be rated higher than an S&P 500 index fund. It isn't a better fund, it is a DIFFERENT fund that
owns small and mid caps that have done well over the last three years. THIS difference would be exposed in the
FUNDGRA+DES®
analysis where many other systems would lump them together and artificially rate the total market fund "better" for large
cap blend even though it doesn't belong in that asset class! The reverse of this artificial "better" rating would happen as well
in these other rating systems when small and or mid cap stocks under-perform large cap stocks.
The only peer group ranking used in our grading methods is for fund expenses. Our expense grades take the lowest 10% of
expense ratios for all funds based on their best fit asset class, and grade them relative to this expense benchmark. LESS THAN
4% of all funds make the expense honor roll! We wanted to avoid all peer group biases in our routine by using the lowest cost
available fund for each asset class as an expense benchmark however, for all but a handful of the asset classes, more than 80%
of all gradable funds have expenses that are MORE THAN THREE TIMES as expensive as the lowest cost alternative.
Thus, we were compelled by the preponderance of expensive funds to use a method like the peer group approach (with benchmarking)
for the ranking of expenses. To do this, we take the lowest 10% in expense of all funds that pass our diversification grades for
the particular asset class and set this as the expense ratio "benchmark". The methodology will award a grade of B- if a fund's
expense is plus or minus 30% of this benchmark expense. (For example, the lowest 10% expense ratio of all small cap growth funds
in a test of fund data for the period between April 2004-March 2007 had an expense ratio of 1.01%. In order to get graded a B- for
this asset class in expenses, a fund's expense ratio would need to be approximately +/- 0.30% of this benchmark or between 0.71%
and 1.31%.)
The grades for any particular fund's expenses are stated relative to the lowest 10% of all funds' expenses for funds within a
given asset class based on the following ratios:
- A+ = Expenses less than 30% of the lowest 10th percentile (i.e. less than 0.30% if the lowest ten percent of all funds in a
class is 1.00%).
- A = Expenses less than 40% but more than or equal to 30% of the lowest 10th percentile
- A- = Expenses less than half, but more than or equal to 40% of the lowest 10th percentile
- B+ = Expenses less than 60%, but more than or equal to half of the lowest 10th percentile
- B = Expenses less than 70%, but more than or equal to 60% of the lowest 10th percentile
Thus, to receive an honor roll expense grade for expenses, a fund's expense must be at least 30%
below the benchmark established by the lowest 10% in expenses of all funds for a particular asset class.
- B- = Expenses within +/- 30% of the lowest 10th percentile
- C+ = Expenses more than or equal to 130% but less than 150% of the lowest 10th percentile (i.e. if the lowest 10th percentile expense for a particular asset class were 0.50%, a fund would be graded a C+ at an expense ratio of 0.65%-0.75%).
- C = Expenses more than or equal to 150% but less than 175% of the lowest 10th percentile
- C- = Expenses more than or equal to 175% but less than double the cost of the lowest 10th percentile funds
Note: Using fund data for the period April 2004-March 2007, less than 43% of all funds that passed the
diversification grading step had expenses less than DOUBLE the lowest 10% of all funds in the asset class.
Stated another way, more than HALF the funds graded for expenses had expenses that were more than double the
expenses charged by the funds with expenses in the lowest 10%.
- D+ = Expenses more than or equal to 200% but less than 250% of the lowest 10th percentile expense
- D = Expenses more than or equal to 250% but less than 300% of the lowest 10th percentile expense
- D- = Expenses that are more than or equal to 300% but less than 4 TIMES the cost of the lowest 10th percentile expenses (for example, if the lowest 10% of a particular asset class's expense ratio was 0.70%, as it was for funds in the High Yield Bond asset class, a grade of D- would be assigned to funds charging 2.1%-2.8% in expenses.) More than 80% of all graded funds are generally graded D- or better.
- F = Expenses FOUR TIMES OR MORE than the lowest 10th percentile expense. Almost one in five funds is typically graded as F for expenses.
(Note: If a fund has a redemption fee or contingent deferred sales charge it is lowered by one grade.)