Oranges and Apple

IFA's Performance Monitoring Report vs. the fi360 Fiduciary Score

Oranges and Apple

For the purpose of the quarterly monitoring of the funds that IFA advises for use by both our individual and institutional clients (including 401(k) plan sponsors), IFA has chosen to use its own proprietary scoring system rather than the widely used fi360 Fiduciary Score. Below is a summary of the differences between them with respect to equity funds.

The primary difference is the weight given to past performance. IFA’s Performance Monitoring Report (PMR) only assigns 10% of the score to past performance (based on Sharpe ratios). The fi360 methodology assigns up to 42.9% of the score based on past performance relative to the peer group. IFA’s decision to assign a lower weight to past performance is based on the requirement of a large sample size (i.e., number of years) to determine the presence of skill with statistical significance. The calculator below determines the number of years needed for statistical significance at a 95% confidence level given the average alpha (excess return over the benchmark) and the standard deviation of alpha.

As the chart below demonstrates, statistically significant alpha over a long period of time (20 years) is a very rare occurrence.

For the chart above, restricting it to only the funds that had positive alpha, the average alpha was 1.0% with a standard deviation of 5.8%. Plugging those numbers into the above calculator reveals that 135 years of similar data is required before we can conclude that skill explains the returns rather than luck. If more than 20 years is needed to conclude that equities have a higher expected return than Treasury Bills, then making inferences about manager performance based on ten or fewer years of data is unlikely to prove beneficial.

One of the primary determinants of performance is a fund’s expenses, both explicit and implicit. IFA’s PMR assigns 15% of the score to the net expense ratio and another 15% to the turnover ratio, which is a measure of the funds internal trading expenses. The fi360 methodology assigns 11.4% of the score to the net expense ratio, and it does not assign a score for turnover.

IFA’s PMR and fi360 take different approaches to measuring style drift. IFA uses the quantitative measures of average market capitalization and price-to-book ratios, assigning 15% of the score to each measure. The fi360 methodology assigns 11.4% of the score to an analysis of the internal holdings of the fund and an additional 11.4% to Morningstar’s assigned style box compared to its peer group.

IFA’s PMR places a high degree of emphasis on diversification, assigning 15% of the score to the number of holdings and an additional 15% to the percentage of total assets contained in the top ten holdings. The fi360 scoring system does not assign a value to diversification.

There are two categories where fi360 assigns a score and IFA does not—manager turnover and size of the fund. The fi360 scoring system assigns up to 11.4% to each of these. Based on our research and review of academic papers, IFA does not consider manager turnover to be an important indicator of future performance. Returns come from risk exposure—not manager selection. While a fund needs to attain a minimum size to have future viability, IFA would only consider using funds from solid fund families.

Here is a table summarizing the scoring differences between IFA and fi360 for equity funds.

For bond funds, fi360 uses the exact same scoring methodology as for equity funds. IFA, on the other hand alters its scoring system to reflect the basic differences between the two asset classes. IFA assigns a 40% weight to the fund’s duration, which is a reflection of the degree of term (or interest rate) risk. Since some bond index funds can legitimately have a high turnover rate as a byproduct of targeting a particular section of the yield curve, IFA only considers the net expense ratio, to which it assigns a 20% weight. The remaining 40% is assigned to the annualized standard deviation of the fund, which is expected to provide a measure of the degrees of both term and default risk taken relative to peers. Of this 40%, 25% is assigned to the ten-year number, 10% is assigned to the five-year number, and 5% is assigned to the three-year number.

Here is a table summarizing the scoring differences between IFA and fi360 for bond funds.