ICI Viewpoints

In April 2024, the Department of Labor (DOL) issued a controversial rule intended to broaden the definition of fiduciary advice to cover broker recommendations on individual retirement accounts (IRAs). The rule was stayed indefinitely by the courts in July 2024. That same month, the Government Accountability Office (GAO) nevertheless issued a report supporting the controversial rule.

The GAO report argues that investors risk losing “tens of thousands of dollars” by purchasing mutual funds through brokers. In particular, the report claims that actively managed mutual funds that compensate brokers (“bundled funds”) have lower gross returns (returns before fees) than actively managed funds that do not compensate brokers (“unbundled funds”).[1] According to the report, this means brokers put their own interests before those of clients. ICI voiced skepticism about the GAO’s results, surmising that any differences in gross performance across funds had little to do with whether they were bundled or unbundled.

Our investigation confirmed this hunch. Using “factor models”—the widely accepted standard in finance for risk-adjusting funds’ portfolios—we found that gross risk-adjusted returns are statistically the same for bundled and unbundled funds. In short, there is no evidence that investors risk losing “tens of thousands of dollars” by purchasing mutual funds through brokers.

GAO’s Analysis Failed to Adjust Adequately for Risk Exposure 

The GAO’s claim that bundled funds underperform holds only for a subset of actively managed mutual funds: those that invest most of their assets in US listed stocks (“domestic equity funds”).[2] The report indicates that bundled domestic equity funds underperformed their unbundled counterparts by 89 basis points (0.89 percent) per year from 2018 to 2021, a difference which, if correct, is economically significant.[3]

We were able to replicate this result using the GAO’s statistical approach. However, looking closer, we discovered that the result is driven by an even smaller subset of funds: those actively managed domestic equity funds that Morningstar labels as “large cap value funds” or “large cap blend funds.” For example, when we dropped large cap value funds from the analysis, the GAO’s results are no longer statistically significant. When we dropped both large cap value and large cap blend funds, the GAO’s results become both statistically and economically insignificant. In other words, there is no evidence of underperformance among bundled funds in the seven remaining Morningstar domestic equity categories, which account for nearly half of the $5 trillion in assets in domestic equity funds.[4]

It is not bundling or unbundling that drives the GAO’s results, but differences in funds’ portfolio holdings that create varying risk-return profiles even for funds within the same investment category.

To be sure, the GAO’s analysis tried to adjust for this by including in its statistical approach Morningstar category level factors (“dummy variables” in the jargon of statistics). This implicitly (and incorrectly) assumes all funds in the same Morningstar category have identical portfolios and investment strategies.

Portfolios and investment strategies vary across funds, even across funds in the same Morningstar category. Consider, for example, funds that Morningstar classifies as “large cap value.” In 2021, the last year in the GAO’s dataset, bundled large cap value funds held 88 percent of their assets in US stocks and 7 percent in international stocks, compared to 91 percent and 6 percent, respectively, for unbundled funds (Figure 1). While these differences might seem too small to matter, the only way to know for sure is to use a standard approach (i.e., factor models) to adjust for the risk in each fund’s underlying portfolio, even among those funds in the same Morningstar category.

Figure 1
Bundled and Unbundled Funds Hold Different Portfolios
Percentage of total net assets of domestic equity large-cap value funds, year-end 2021

Source: ICI calculations of Morningstar data

Note: Chart components may not add up to 100 percent due to rounding.

Risk Adjustment Matters When Comparing Fund Performance

Much academic literature shows that the best way to evaluate a fund’s risk-adjusted performance relative to other funds is through factor models. A factor model estimates a fund’s risk-adjusted return by benchmarking its performance to returns on broad market indexes (e.g., S&P 500; FTSE 100). The performance of each fund is measured relative to a common set of market indexes, otherwise known as “factors,” which puts all funds in the sample on the same basis, irrespective of how Morningstar classifies each fund.

We use three different factor models to estimate the risk-adjusted returns for each fund. These models capture a fund’s exposure to:

 (1) domestic equity markets; plus

 (2) international equity markets; and additionally

 (3) bond markets.

Given these factor models, a fund’s gross return can be broken down into the portion explained by market factors and a remainder, which is the fund’s risk-adjusted return (i.e., the standard risk-adjusted measure of fund performance).

In all three models, we find the same result: there is no evidence that bundled funds underperform (Figure 2). The leftmost bar in Figure 2 shows the performance of bundled funds reported by the GAO (-89 basis points per year)—the negative sign indicates that bundled funds “underperform.” The next three bars show risk-adjusted relative performance based on our factor models. As seen below, estimated “underperformance” improves sharply to between -1 and -3.5 basis points. Although still negative, these estimates are economically small and not statistically different from zero.

Figure 2
“Underperformance” of Bundled Funds Vanishes with Proper Risk Adjustment
Differences in gross annual returns on bundled versus unbundled funds, 2018–2021

¹ Statistically different from zero.

² Not statistically different from zero.

Note: Domestic equity market factors are the four Fama-French-Carhart equity market factors; international equity market factors include a developed market factor and an emerging market factor; and bond market factors include US bond market, US bond default, and US bond prepayment factors. See Qureshi (2024) for the definitions of these factors.

Note: One basis point equals one-hundredth of one percentage point, i.e., 0.01%.

Sources: GAO report and ICI calculations of data from Morningstar, Bloomberg, MSCI, and Ken French’s data library.

By putting all funds on the same risk-adjusted basis, the gross returns on bundled and unbundled funds are statistically the same.

GAO’s Analysis Is Flawed

Bundled funds do not have lower risk-adjusted gross returns than unbundled funds, which means there is no evidence to the GAO’s claim that investors risk losing “tens of thousands of dollars” by purchasing mutual funds through brokers. Policymakers shouldn’t rely on the GAO’s findings as a rationale for resuscitating the DOL fiduciary rule or any other new rule proposals related to it.

Notes

[1] The GAO defines “bundled funds” as mutual fund share classes (rather than funds per se) that are classified by Morningstar as paying compensation to investment advisers (share classes that offer front-end, back-end, or level-loads and/or revenue sharing with investment advisers). “Unbundled funds” are fund share classes that pay no compensation, no revenue sharing, and no reimbursements of expenses to investment advisers. There are also “semi-bundled funds,” which the GAO defines as fund share classes that Morningstar classifies as lying between bundled and unbundled share classes with respect to compensation for investment advice and assistance.

[2] The GAO also examined the gross returns of index funds, international and global equity funds, and target date funds. It found no consistent evidence that bundled index funds underperformed, no evidence that bundled international and global equity funds underperformed, and evidence for target date funds that bundled funds outperformed.

[3] For example, it is more than the average expense ratio (0.62 percent) that investors incurred in equity mutual funds in 2022. And it is well-known that investors focus intensely on fund expenses.

[4] The seven other Morningstar categories are: large cap growth; small cap value, growth, and blend; mid cap value, growth, and blend.