Are Brokers Acting in Their Clients' Best Interest?

By Jasmin Sethi and Aron Szapiro

In 2015, the U.S. Department of Labor proposed the “fiduciary rule,” a regulation aimed at mitigating conflicts of interest in investment advice and ensuring that brokers act in the best interests of their clients. After the 5th Circuit Court of Appeals struck down the DOL’s final rule last spring, the SEC proposed a new standard of conduct similarly aimed at addressing conflicts of interest in April.

What Do We Know About Conflicts of Interest? Historically, a key conflict of interest that could distort brokers' recommendations derived from load-sharing arrangements. To assess how bad these conflicts were for investors, the DOL relied on an academic paper by Susan Christoffersen, Richard Evans, and David Musto (referred to as the CEM paper from here on) to justify the fiduciary rule.[1] The CEM paper found that funds that paid higher-than-expected loads to brokers tended to see higher inflows as a result. A significant finding of the CEM paper was that for every 100 basis points in higher-than- expected loads, future returns were reduced by about 0.34%. The academics also found that every $1 in higher-than-expected load sharing increases flows by $14.20. However, the CEM paper relied on data from 1993 until 2009. It's reasonable to wonder if the market has evolved since then, an argument that industry commenters raised in response to the DOL's fiduciary rule.

Our Approach Morningstar's policy research team set out to answer this question in a recent paper, combining public filings from the SEC with Morningstar's own data on flows, performance, and mutual fund attributes.[2] Morningstar's policy research team replicated the methodology used in the CEM paper and combined Morningstar's own data on fund performance, flows, and other characteristics with public filings collected by the SEC to update the analysis. Our analysis suggests that brokers and advisors are more rigorously screening funds on performance when making recommendations to their clients.

The key findings from our work are:

• Consistent with CEM, load sharing drove higher flows to funds with lower returns during the period prior to 2009.
• While that effect was stable before 2014, it decreased significantly with the proposal in 2015 and finalization in 2016 of the fiduciary rule.
• We also confirm that unusually high loads are associated with lower performance, but the effect declines over time.
• In fact, after 2009, we do not find any statistically significant association between unusually high loads and performance, if we control for previous returns.

The Proposed Rule Seems to Have Redirected Fund Flows We interpret this empirical evidence to suggest that regulation played a part in the decisions brokers made about which funds to sell. While loads appeared to play a big role in directing fund flows from 1993–2014, they do not seem to affect fund flows in a statistically significant way from 2015 and on. In short, brokers may have been swayed to steer their clients to funds that shared more loads with them, but that does not seem to be the case anymore.

As shown in EXHIBIT 1, when we examine five-year rolling time periods over the past 25 years, the effect of higher-than-expected loads on inflows is consistently positive over time. Excess loads paid by a fund to unaffiliated brokers led to systematically higher inflows to those funds, even after controlling for fund characteristics such as returns, performance, and fees.[3] For example, from 1993 to 1997, these results lead us to estimate that a 100-basis-point increase in excess loads paid to unaffiliated brokers is associated with about a 0.02% increase in monthly inflows to that fund.

- source: Morningstar Analysts

We find evidence that the DOL rule may have reversed the trend of inflows flowing to funds that paid excess loads to unaffiliated brokers, as shown in EXHIBIT 2. We also deployed the CEM methodology with some additional variables (described in our full paper) to assess the impact of regulation. We used the same specifications that we deployed in EXHIBIT 1, but with a dummy variable for before and after the DOL proposed the fiduciary rule in 2015.[4] Although the rule was technically proposed in April 2015, the department had signaled it would propose such a rule throughout the end of 2014. In any case, the result is generally robust (at 10% level of statistical significance) given different definitions of the dummy indicating the timing of the DOL rule. EXHIBIT 2 illustrates how, as the DOL began to signal it would promulgate regulations to rein in conflicts of interest, the conflicts embodied by excess load sharing stopped driving flows.

- source: Morningstar Analysts

To test the robustness of these results, we also use alternative definitions of past performance, as described in the full paper. We find no significant impact on these findings. In summary, we believe these results show that the DOL’s proposal of the fiduciary rule had a statistically significant impact on flows to funds. In particular, funds that paid excess loads to unaffiliated brokers previously saw higher inflows, but the proposal of the fiduciary rule appears to have reversed that effect.

The Proposal Does Not Have the Same Effects on Returns While our analysis tells a clear story on the association of the DOL's proposed fiduciary rule with flows, the association with returns is less visible. In fact, we see little association with loads and returns after 2009, a significant shift because that was the last year CEM analyzed, and that work in turn provided the basis of the DOL's regulatory impact analysis of the benefits of the fiduciary rule. We were able to largely replicate the CEM results for returns on the 1994–2009 period. Again, we obtained coefficients that were similar in magnitude and the same in statistical significance to those obtained by CEM, meaning during this time period, excess loads drove flows to worse performing funds.

We also conducted a subperiod analysis of returns like that conducted for flows. We ran the CEM specification for five-year periods beginning in 1994 and moving through the whole period until 2016. We obtained coefficients on excess loads for all of these regressions. These coefficients are shown in EXHIBIT 3. We can see from these results that excess loads have a negative relationship with excess returns throughout this period; however, the results are not significant in the period after 2010—at least as long as we account for the important variable of lagged returns.

Unlike with flows, when we look at the period before and after the fiduciary rule separately, we do not see a significant effect of excess loads in either period once we control for lagged returns.[5] Similarly, when we ran a pooled specification interacting excess loads with the fiduciary rule indicator dummy, we also did not see a statistically significant effect.

- source: Morningstar Analysts

Pre-Existing Trends Affected Performance We believe that the lack of statistical significance in the performance regressions likely derives from the fact that Dodd-Frank Section 913, which empowered the SEC to promulgate a new standard of conduct for broker-dealers, and the proposal of the DOL fiduciary rule had already influenced the culture around performance accountability. Brokers likely had already been given incentives to direct clients toward higher-quality—or at least higher-performing—funds because of the increased scrutiny of their choices.

Of course, this finding comes in an environment in which loads and, with them, load-sharing payments had been declining for years and continued to decline after 2009. Indeed, the DOL estimated that total loads would continue to fall at approximately 3.2% per year after 2014 when it completed its analysis.[6] Our examination of public filings reveals that this trend of falling loads continued as fewer and fewer flows were subject to loads, as shown in EXHIBIT 4, which illustrates that the total percentage of inflows subject to a load decreased from about 5% in 2010 to less than 2.5% in 2017. However, in an environment of declining loads, it is possible that excess load sharing would have driven more flows than it otherwise would have because fewer funds paid higher-than-expected loads. We do not see such a trend in the data, likely because of the DOL fiduciary rule.

What's Next Our econometric analysis reveals that harms from a key financial conflict of interest—load sharing between mutual funds and intermediaries— appear to have declined since 2010, the last year on which much of the DOL's regulatory impact analysis for its fiduciary rule was based. This benefit for investors could be attributable to regulatory pressure, pre-existing trends away from load sharing, or a mix of the two. We believe that the preponderance of evidence points to regulatory action accelerating a move toward business models in which financial advisors put their clients' interests first. In particular, flows into mutual funds paying unusually high excess loads declined after the DOL proposed its fiduciary rule in 2015, and this shift was statistically significant. This reduction in the distortionary effect of conflicted payments suggests that firms put in place effective policies and procedures to mitigate conflicts of interest in response to the DOL rule and, further, that the SEC's proposal could maintain this important momentum.

The SEC has indicated that a finalized version of Regulation Best Interest is forthcoming, and we expect that the increased scrutiny on conflicted advice brought about by the fiduciary rule will continue. That said, we think if the SEC makes some adjustments to its proposal, as we have previously suggested, it will have a better chance of maintaining this momentum.[7] We believe this implies there is a public benefit in 1) maintaining strong regulations around financial advice to continue to protect investors in this area and 2) nudging advisors to act in the best interest of those investors.

[1] Christoffersen, S.E.K., Evans, R., & Musto, D.K. 2013. “What Do Consumers’ Fund Flows Maximize? Evidence from Their Brokers’ Incentives.” The Journal of Finance, Vol. 68, No. 1, PP. 201–235.

[2] Sethi, J, Spiegel, J., & Szapiro, A. 2018. “Conflicts of Interest in Mutual Fund Sales: What Does the Data Tell Us?” Morningstar White Paper.

[3] Captive brokers are affiliated with a particular fund sponsor and only sell that fund sponsor’s funds. Unaffiliated brokers, on the other hand, are not saddled with such a restriction.

[4] U.S. Department of Labor. 2015. “Definition of the Term ‘Fiduciary’; Conflict of Interest Rule-Retirement Investment Advice.” Federal Register, Vol. 80, No. 75, P. 21928. The DOL also proposed a rule in 2010 but quickly signaled that the proposal needed substantial revisions.

[5] We found that the estimator for excess loads was negative and statistically significant at a 5% level pre-DOL rule proposal if lagged returns were not included in the model.

[6] U.S. Department of Labor. 2016. “Definition of the Term ‘Fiduciary’; Conflicts of Interest-Retirement Investment Advice: Regulatory Impact Analysis for Final Rule and Exemptions.” P. 345.

[7] Szapiro, A. 2018. “Morningstar Regulation Best Interest Comment Letter to the SEC.”

This article originally appeared in the Spring 2019 issue of Morningstar magazine. To learn more about Morningstar magazine, please visit our corporate website.