Citi Research recently published a research piece titled, "Short-End Notes Impact from MMF reform and AUM expectations," which tells us, "Do not expect a repeat of 2016 this October." Author Jason Williams writes, "We've gotten multiple questions on the front-end impact due to the full implementation of the new money market fund reform, specifically the SEC's liquidity fee which is set to turn on in October. As a reminder, last year the SEC passed the next wave of Money Market Fund (MMF) reform. The big story was that the SEC dropped swing pricing, which would have required the fund's NAV to be adjusted to incorporate transaction costs and liquidity challenges around redemptions across the entire portfolio, not just the subset being redeemed. Instead, the SEC implemented a liquidity fee to be applied when daily withdrawals hit 5% if liquidity costs are 'not small.' Other changes were an increase in requirements for daily and weekly liquid assets -- 25% and 50% respectively -- and the removal of gates and fees tied to weekly liquidity."
He explains, "Likely due to the regulatory cost of running a prime fund, there is growing expectation for more prime fund AUM to shift into government funds, which could impact repo and T-bill pricing and widen CP/OIS and XCCY spreads (XCCY would be indirectly impacted). This year thus far, two large prime funds with AUM totaling around $220bn announced their conversion into government-only. In 2016, a material amount of AUM shifted from prime to government funds [over $1 trillion], as investors wished to avoid ramifications of the first MMF reform.... This sharp move pushed front-end repo and T-bill yields back towards the RRP rate.... We don't expect a repeat of that size this time around. Indeed, thus far, retail prime MMFs -- which are exempt from the liquidity fee -- haven't seen a shift in inflows, which should cushion any material shift in institutional flows."
A section titled, "Fund level analysis does not show impending flight," states, "We see multiple reasons why this cycle will be materially different to 2016. The two key risks are investors in prime funds moving into government only, and prime funds converting into government only due to growing regulatory costs."
It continues, "Institutional prime fund AUM is smaller in aggregate, and retail prime AUM has been growing. Institutional prime MMF AUM is smaller now than it was at the end of 2015, moreso when we consider the two large funds that are set to convert to government-only. More importantly, retail prime AUM has increased materially over the past year, which should more than offset further allocation departing from institutional funds.... Indeed, we suspect MMF AUM may continue to grow for the rest of the year."
Citi tells us, "Institutional prime MMF allocation into CD/CP likely to remain minimal. Even if more institutional prime MMF AUM shifts into government funds later this year, we don't expect there to be a lot of CP to unload, if any. Institutional prime funds are already very underweight CP – especially for maturities beyond 1m.... Indeed, historically prime funds hold much less CP than pre-COVID."
They comment, "This does not necessarily signify they are shifting into government-only, as there are good reasons to be underweight CP. Firstly, spreads are tight. Secondly, the anticipation for some light spread widening as we approach October may limit fund's appetite to buy longer-dated paper, as well as riskier paper such as financials. Thirdly, the current Fed backdrop encourages sitting in shorter-date product such as repo. Hence, even if more funds do shift away, there should be little impact as holdings are mostly short-dated."
Citi adds, "In conclusion, we could not find a strong discernable pattern signifying that multiple funds are lining up to convert in coming months. Having said that, we raise a 'worst case' estimate based on the funds sitting in the lower quadrants of Figure 6. The four funds that are lowest along the Y-axis have only ~$30bn of CP (excluding government converts). Even if we take in the outlier fund in the top-left quadrant, holdings barely surpass $30bn. Thus, we don't see material impact on the CP market resulting from another round of MMF reform. If the repo market had to take in this ~$30bn of new cash, it would not be impacted either."
The Citi update also asks, "Where does MMF AUM go from here?" They answer, "After seasonal decline around April tax receipts, MMF AUM has roughly recovered, with ~$60bn more to go to reach all-time highs. Generally, the increase in MMF AUM over the past few years has likely been driven by flows from the interest rate differential to non-yielding bank deposits as well as the large increase in bank reserves, which increases commercial deposits, due to QE4.... [W]e think investors forget how important bank reserves, and therefore commercial deposits, are to MMF AUM (since more reserves = more deposits, all else equal)."
They also ask, "Where do we go from here? Based on historical patterns, we don't expect a structural decline in MMF AUM over the next 12 months. If anything, some marginal increase should continue from here.... [W]e expect QT to continue into H2 2025, barring a recession.... More QT could drain reserves, assuming RRP is somewhat static in the short-term. Still, we don't expect MMF AUM to be pulled down. For instance, the recent TGA increase around April tax receipts appears to have created only a temporary impact."
Finally, Citi writes, "Will another few hundred billion impact front-end markets? If MMF AUM grows materially, say on the order of 500bn over the next 12 months (in reality, it may just be another $100bn-$200bn), we don't think there will be much as much impact to front-end markets as one might surmise. What we mean by that is such a large increase would probably be the result of banks losing some deposits to MMFs. Presumably, these same banks would need to issue either CP or timed deposits, which would be bought by retail prime funds. Banks may also tap advances, which are funded via discount notes bought by government MMFs."
