Financial misconduct or fraud has consequences for both the firm and its customers. In other words, the manager of a firm alleged to have committed fraud needs to react to the incident being mindful of the consequences for shareholders, the product market, and the manager’s own career concerns. Such a multifaceted problem is generally not acknowledged, but rather separately examined in terms of stock price losses, regulatory penalties, product market reputational damages, and loss of corporate control and managerial reputation. Similarly, in the U.S. mutual fund industry, academic researchers have addressed the related question of penalties for mutual fund financial misconduct and fraud, finding performance declines and significant investor redemptions. However, a corporate manager dealing with fraud serves multiple masters and is virtually unacknowledged. In this article, we examine the performance and flow consequences of alleged financial misconduct, which we refer to as the external reaction, alongside the internal response of mutual funds to the fraud (the internal reaction) through asset fire sales and downgrading fund fees.
The mutual fund industry is particularly well suited for analyzing how an agent with a problem of two masters reacts to allegations of fraud. First, the financial press reports accounts of financial misconduct and fraud by mutual funds and their portfolio managers. In 2012, the U.S. Securities and Exchange Commission (SEC) reported 147 such allegations against investment advisers, one short of the previous year’s record.1 Professor Eric Zitzewitz from Darmouth College, in his 2006 journal article, documents that at the height of the late trading scandals, market timing by favored investors in some mutual funds cost long-term investors around $400 million per year from 1998 to 2003. Thus, one gets the impression that both investors and fund companies should pay a great deal of attention to allegations of fraud by the individuals and firms that manage their money.
Second, in the mutual fund setting, both external and internal reactions are measurable. To estimate external penalties, we examine the performance and money flows of U.S. mutual funds following fraud allegations. We also ask the following questions to examine the internal actions of fund companies in response to fraud: How do fund managers manage their portfolio holdings in response to predictable money outflows? Do fund managers engage in or resist asset fire sales? Do fund companies change their fee structures to protect their firms against redemptions in the aftermath of the fraud allegations?
To answer the above questions, we identify a sample of 432 equity mutual funds that are involved in a scandal throughout the nine-year period between 2002 and 2010 using various sources of mutual fund databases, SEC filings documents, and the Wall Street Journal. We then provide a breakdown of scandal types and the regulators that initiated allegations against scandal funds. The timing scandal denotes only scandals in which the mutual fund was alleged to have allowed market timing in its fund to occur. The disclosure scandal indicates scandals in which the fund failed to adequately disclose material information to its investors.
In many cases, this involves directed brokerage agreements between the fund’s adviser and broker–dealer agents who sold mutual fund shares in return for a commission paid out of the fund’s assets. The other scandal includes all remaining cases of regulatory violations.
Typically, such cases involve the incorrect charging of management compensation (e.g., Numeric Investors, LLC), the unfair allocation of “hot” initial public offering issues to selected individuals (e.g., Monetta Financial Services Inc.), or the acceptance of entertainment and travel gifts from affiliated brokerages (e.g., Fidelity Management & Research Company).
In Table 1, the Securities and Exchange Commission (SEC) appears to be the dominant regulator, having carried out investigations against 342 of the 432 mutual funds in our sample. The New York Attorney General (NYAG) was the second most active regulator within our sample, having pursued 47 charges, which were exclusively against timing-related scandals. The Self-Regulatory Organizations (SROs) are the least active one among the regulatory bodies. See table 1
What are the characteristics of scandal versus non-scandal funds? In Table 2, we perform a difference-in-means test to determine if there are significant differences between funds involved in scandals and their non-scandal counterparts. We compare the means of both groups across all mutual fund characteristics and performance measures. Scandal funds are significantly bigger and belong to larger fund families than non-scandal funds. The average size of a scandal fund is $1.938 billion and it belongs to a family with TNA of $187.928 billion.
This is in contrast to an average fund size of $991 million for non-scandal funds, which belong to a family with TNA averaging $85.165 billion. Scandal funds also have higher expense ratios. The expenses are made up of higher management fees, 12b-1 fees, and non-12b-1 fees. Scandal funds have higher turnover ratios of 97 percent, in comparison with 84 percent for non-scandal funds. Interestingly, on average, scandal funds have cash holdings that are around 1.2 percent lower than their non-scandal counterparts, possibly to meet redemptions. Finally, consistent with our expectations, we find scandal funds experience a lower level of fund flows and performance. See table 2
We find scandal funds underperform by a larger magnitude of 198 basis points, while other funds that are affiliated with a family linked to scandal underperform by 80 basis points in annualized terms. On further analysis, we find that mutual funds’ post-scandal underperformance is aligned with the severity of the fraud.
Underperformance is greater for market timing scandals, financial misconduct that attracts higher monetary fines, and the involvement of two or more regulatory bodies.
Interestingly, we find underperformance to be associated with actions initiated by non-SEC regulatory bodies, which suggests that federal laws and regulations have not been entirely effective in deterring mutual fund scandals. This is supported by a study by Dyck, Morse, and Zingales (2010) who find that the SEC was responsible for detecting only 7 percent of corporate fraud cases, while other regulatory agencies were responsible for detecting 14 percent of the fraud cases.
In relation to the effect of scandals on money flows, we observe investors punish scandal funds by withdrawing their capital in the aftermath of a scandal in the subsequent three-, six-, nine-, and 12-month periods. As for affiliated funds in the same fund family, the reduced money flows are only observed for the nine- and 12-month periods. In particular, scandal funds had fund flows that were 15.31 percent lower than those of non-scandal funds. If we consider an average scandal fund that has $1.993 billion in total net assets under management at the time of the scandal, this effect would correspond to around $305 million worth of fund outflows. Importantly, from these results, we observe that a reduction in fund flows is taking place alongside the underperformance, which suggests the possibility that fund performance is partly affected by a fire sale of assets to meet investor withdrawals.
What are the characteristics of an internal reaction to mutual fund scandals? We find that the net selling of both scandal funds and other funds that belong to the family of the scandal fund is not immediate and only takes place in the second quarter onward, following the announcement of the scandal. We also find evidence that scandal funds and other funds in the family implicated in the scandal significantly reduce expenditure on non-marketing costs, which suggests fund families attempt to retain investors. Altogether, our findings vindicate our suspicion that fund managers react to scandals in line with the dilemma of their position as servants to two masters: investors and fund companies.
Our overall findings also hold even after controlling for fund- and family-level characteristics, and are not driven by the financial crisis in years 2007 and 2008. Furthermore, when we match scandal funds (treated group) with non-scandal funds (control group) using propensity score matching algorithm, we find similar findings to our main analysis.2 As presented in graphs A to E, we find the performance and fund flows of treated group to decline in the aftermath of the scandal and evidences of treated group to engage in net selling and downgrade of fund fees.
In conclusion, fund managers face the dilemma of serving two masters, raising the question of how they behave in the aftermath of the disclosure of fraud. Our mutual fund setting enables us to explore business unit-level operational effects in ways that are not possible in a corporate context. The findings vindicate our conjecture that it is necessary to look at both external and internal responses to conduct a comprehensive analysis of the effects of fraud on mutual funds. We find evidences of performance declines and investors’ redemptions in the aftermath of the scandals. Furthermore, we also find evidences that mutual funds are resorting to asset fire sales in order to meet investors’ redemptions and to downgrade fund fees in their efforts to retain investors. The results of this study are therefore of relevance to regulators and investors since they add internal dimensions of the effects of corporate fraud that might not be as salient as the external effects these parties are usually exposed to.
- The report is available online at https://www.sec.gov/News/PressRelease/Detail/PressRelease/1365171485830#.UvGgb7TDUg1.
- The variable Treated denotes a fund that was involved in a scandal, while Control denotes a fund that was matched by the following fund characteristics: log(fund age), log(fund size), log(family size), expense ratio, turnover ratio, cash ratio, fund flow, and fund alpha in the prior month of the scandal event.