Courtesy of Stefano Lugo
Credit ratings assigned by Credit Rating Agencies (CRAs) like Moody’s and Standard & Poor’s have long been relied on by investors and regulators alike. This gives CRAs significant influence over our financial markets. Ratings too severe can lead to a credit crunch, and ratings too generous can lead to a credit bubble—which became evident during the 2007 subprime crisis. For this reason, the Dodd-Frank Act of 2010 requires all US Federal agencies to remove references to ratings from their own regulations.
Whereas the problems associated with the regulatory use of ratings are clear, little is known about the actual consequences of disallowing their use by regulators. On the one hand, these reforms can foster independent credit risk assessments by investors, thus reducing the excessive influence of credit ratings. On the other hand, by eliminating clear thresholds based on ratings, regulators may provide investors with the incentive and opportunity to take on more risk.
In a recent working paper, I investigate the effects of removing rating-based rules by focusing on one of the first actual implementations of the Dodd-Frank act: the revision of rule 2a-7 under the Investment Company Act 1940.
Rule 2a-7 and the use of credit ratings in the regulation of Money Market Funds
Rule 2a-7 governs Money Market Funds (MMFs). MMFs are mutual funds required to invest predominantly in short-term, safe securities. One of the key aspects of MMFs is that they are open-end funds whose shares are created or redeemed at a constant price of one dollar per share. This feature makes investing in MMF shares similar to putting money in a bank account, as investors can at any point in time sell their shares at the same price paid to purchase them. The obvious key difference between MMF shares and a bank account is that the latter benefits from the protection of FDIC deposit insurance, whereas the former does not. As a result, MMFs are exposed to runs by investors, which became evident in 2008 when the Reserve Primary Fund was unable to redeem its shares at one dollar (known as “breaking the buck”) due to severe losses caused by the default of Lehman Brothers. To limit the risk of funds breaking the buck, pre-reform rule 2a-7 formally set explicit boundaries on the maximum exposure to credit risk that prime MMFs can undertake. Securities were labelled as first tier, second tier, or below second tier based on the ratings assigned by CRAs. Prior to the crisis, prime MMFs were not allowed to purchase securities rated below second tier, and they were not allowed to allocate more than 5% of their portfolios to second tier securities (the threshold for second tier securities was reduced to 3% in 2010).
Following Dodd-Frank’s mandate, in September, 2015, the Securities and Exchange Commission published a revised rule 2a-7 that removed any reference to credit ratings. Securities are no longer categorized as first tier, second tier, and below second tier. Prime funds are now allowed to purchase previously non-eligible securities and to invest any share of their portfolios into securities previously categorized as second tier. While removing any rating-based threshold, the new rule also explicitly requires MMFs to “retain a similar degree of credit quality standards” as before the reform.
The elimination of rating references came on the heels of another revision to rule 2a-7 in 2014 that required prime funds targeted at institutional investors to price their shares using the effective, market-based value of their assets (so called “floating net asset value” (NAV)). Government funds and retail prime funds were allowed to keep their shares fixed at one dollar per share.
Empirical evidence on the effects of the reform
Since the removal of rating-based rules, prime funds have significantly increased the share of second tier (or below) securities—as defined before the reform—in their portfolios. At the end of August 2018, around 14% of active prime funds had more than 3% of their portfolios allocated to non-first tier securities. This fact does not prove that the reform has led to an increase in risk-taking by MMFs; in principle, funds may simply have started performing independent credit risk assessments, as required by the reform, investing more only in second tier securities less risky than their rating-based categorization would imply. However, the data reveals the opposite – investments in second tier securities have increased on average but have decreased for second tier securities characterized by a level of credit risk more in line with a first tier, rather than second tier, categorization. Compared to the pre-reform period, funds appear to be investing more in securities promising higher returns but with more risk.
The post-reform increase in the share allocated to non-first tier securities is particularly high for funds more likely to have been constrained by the previous regulatory cap of 3%, further supporting a causal interpretation of the relation between the reform and the verified change in investment behavior by MMFs. The increase in non-first tier holdings cannot be attributed to the mandate that institutional MMFs use a floating NAV. Retail funds, who maintain a fixed price of one dollar per share, have increased their allocation to non-first tier securities significantly more than institutional funds. If anything, imposing the use of a floating NAV may have partially counterbalanced the effect of removing rating-based rules.
The increase in the average share of non-first tier securities in the portfolios of MMFs is entirely driven by a shift in the allocation within specific asset classes, in particular, commercial paper issued by non-financial corporations. This has interesting consequences for how markets price credit risk; the sharp increase in the relative demand by MMFs for second tier commercial paper is associated, all else equal, with a significant reduction in the spread paid at issuance on those securities. Over time, this could lead to firms characterized by a relatively high level of credit risk relying more on short-term financing, thus increasing their exposure to roll-over risk.
Implications for policy makers
The results presented in my paper provide empirical evidence of the perils of removing credit ratings from financial regulation. Despite their flaws, credit ratings establish simple, clear-cut rules. Without such rules, MMFs have the ability, and incentive, to significantly increase their exposure to credit risk. One possible solution to this dilemma would be for regulators to first define new rules that are at least as clear and objective as the rating-based ones they aim to replace. Even if agencies want to give more responsibility to regulated investors over their credit risk assessments, objective thresholds should be kept in place to set a boundary. A good example is provided by the December 2017 reform of the Basel regulatory framework. The new output floor mandates that capital requirements determined by banks using their internal models cannot fall below a minimum level computed using standardized (i.e., rating-based) rules. Other regulatory agencies may want to imitate this mixed approach where applicable.
 As discussed below, rule 2a-7 now mandates that prime institutional funds price their shares using their effective net asset value.
 See the document by the Securities and Exchange Commission “Removal of certain references to credit ratings and amendment to the issuer diversification requirement in the money market fund rule”, published in its final form on September 25th, 2015.
to form PF”, published in its definitive form on July 23rd, 2014.