Financial regulation can have far-reaching consequences, some of which may not be intended by the regulator. In a new paper, we trace out the effects of recent money market regulatory reform, which had the objective of increasing the resilience of the US money market industry. We show that the reforms effectively decreased the liquidity of money market fund liabilities. Our results suggest that, as a consequence of the reform, the safest funds exited the industry, and the remaining funds took more risk. Ultimately, these reforms have reduced the supply of safe and liquid assets for investors and decreased the supply of funding for the safest borrowers.
Money market funds (MMFs) provide short-term funding to corporate, financial, and governmental issuers. As of February 2018, the total assets of US MMFs amounted to $2.8 trillion. MMFs are generally regarded by investors as profitable substitutes for deposits and other “money-like” securities, such as Treasury bills. While MMFs do not benefit from explicit deposit guarantees, investors typically expect to redeem their investment at par value and to obtain a safe stream of dividends. This expectation is reinforced by the fact that MMFs practically guarantee their investors a constant net asset value of one dollar for a one-dollar investment. Despite this, a MMF may “break the buck,” a rare situation where the marked-to-market value of a fund’s net assets falls to 99.5 cents or less per dollar investment. In such a case, a MMF may experience a run. The most recent example is the Reserve Primary Fund, which, due to its large holdings of Lehman’s commercial paper, experienced a large drop in the market value of the short-term securities it held and suffered a run in September 2008.
While in 2008, the U.S. government ultimately guaranteed the value of investments in MMFs to stave off additional runs, the money market industry has subsequently been at the center of sweeping reform efforts aimed at improving financial stability. In particular, new regulations, which were announced in July 2014, and became effective in October 2016, aim to decrease the possibility of runs on MMFs by decreasing the liquidity of their liabilities. Under the new regulatory regime, prime and tax-exempt MMFs, which primarily invest in short-term corporate and municipal debt, respectively, can no longer guarantee the value of investor claims, but have to trade at their actual net asset value if they are marketed to institutional investors. In addition, all prime and tax-exempt MMFs, including those targeted at retail investors, can impose liquidity fees and redemption gates (i.e., suspend redemptions temporarily) in times of market stress.
These reforms have effectively changed the nature of MMFs’ liabilities. Due to marking-to-market, the liabilities are potentially exposed to more variation in their redemption value, making them less “money-like.” This provided strong incentives for sophisticated investors to increase their efforts to monitor the value of MMFs’ liabilities. Other investors may consider the cost of information acquisition prohibitive, and consequently may not invest in prime and tax-exempt MMFs. More importantly, assiduous monitoring after the implementation of the new regulations will increase the sensitivity of fund flows to performance. In turn, this may push MMF portfolio managers to reach for yield. MMFs’ incentives to take on risk and invest in higher-yielding, less liquid, securities may also have been reinforced by the funds’ possibility of imposing gates and redemption fees. In our paper, we document changes post-reforms that are consistent with this narrative.
Our paper has five main findings. First, we show that the “money-likeness” of investments in MMFs decreased after the announcement and implementation of the 2014 regulation. Following the reform, MMFs’ assets under management became significantly less correlated with proxies for the aggregate demand for money-like securities. Second, we find that many prime MMFs exited the industry or converted into government MMFs, whose claims were unaffected by the change in regulation (after the reform, prime MMFs’ assets more than halved, while assets of government MMFs more than doubled). Importantly, prime MMFs with safer portfolios were more likely to exit. Third, following the announcement, and after the implementation of reforms, flows into prime MMFs have become more sensitive to performance. Fourth, the increased sensitivity of flows to performance have given MMFs stronger incentives to reach for yield. Indeed, we find that portfolios of MMFs have become riskier following the implementation of the regulations.
Finally, we note that the changes in MMF behavior have important consequences for the availability of short-term financing accessible to different types of borrowers. On the one hand, issuers with lower default risk have become less likely to have outstanding liabilities with US MMFs. At the same time, the amount of outstanding liabilities towards US MMFs has increased for issuers with a high risk of default. An ancillary finding is that offshore funds have at least partially substituted for US funds in the provision of funding to the safest US corporate borrowers.
Our paper provides empirical evidence that financial intermediaries’ assets and liabilities are tightly linked. Consequently, any regulation that changes the nature of intermediaries’ liabilities will inevitably have consequences for those intermediaries’ assets, such as their riskiness. As we document, as soon as the claims of MMFs became more sensitive to the value of their assets, investors’ incentives to monitor MMFs’ portfolios and performance strengthened. This has effectively made MMFs more similar to bond mutual funds and decreased the supply of safe assets for investors.
 Retrieved from the Investment Company Institute at https://www.ici.org/research/stats/mmf/mm_02_15_18.
 For example, Chevalier and Ellison (1997) and Sirri and Tufano (1998) document that the sensitivity of flows to performance affect the portfolio decisions of mutual fund managers.
 Our results are consistent with theories arguing that financial intermediaries’ assets and liabilities are jointly determined (Hanson, Shleifer, Stein and Vishny, 2015)