Don’t worry about money market funds

Nick Dunbar is the editor of Risky Finance, where a version of this post first ran.

Cash is cool again, and US money market funds are bigger than ever, hoovering up money from a banking system that has been slow to raise deposit rates.

After flatlining in size around the $3tn mark since the financial crisis, US MMFs now control over $6tn — and growing fast, according to the latest snapshot from the US Treasury’s Office for Financial Research.

Their allocations are also undergoing a big shift. After taking advantage of the Fed’s reverse repurchase facility to position for higher rates, money market funds are now ditching the RRP in favour of funding US government-backed home lending.

Should we worry?

Let’s first go on a short tour of the MMF landscape — and especially the seven biggest funds. These have grown much faster than the overall MMF universe, and are in a sense bellwethers for the concept of cash in an inflationary rising rate era.

The seven large funds are controlled by just four parent companies — Fidelity Investments, JPMorgan, BlackRock and Vanguard. 18 months ago, JPMorgan MMFs had the biggest market share, with $340bn, but now Fidelity has the lions’ share, with $500bn of the total. 

In March 2022, these seven MMFs held investments that earned a paltry average yield of 0.4 per cent, so how could they take advantage of the Fed tightening that was expected to happen? After all, their existing T-bill portfolios had an average maturity of 140 days, which meant they would be forced to wait.

The answer was to pump 40 per cent of their money into the Fed’s RRP, with a one-day maturity that could take full advantage of Fed monetary tightening in real time. This agility allowed the MMFs to deliver an average return close to the Fed funds rate, using a US government-backed facility that was essentially risk-free.

Meanwhile, US bank deposit rates only captured a fraction of monetary tightening, an example of so-called deposit beta. Even today, with short-term rates currently above 5 per cent, the average US savings account rate is just 0.45 per cent according to the Federal Deposit Insurance Corporation. Even one-year deposits offer only 1.36 per cent on average, a third of the yield available in MMF portfolios.

This simple comparison helps explain the $1tn surge in MMF assets (half of which occurred in the seven largest funds), mirroring the $1tn decline in US bank deposits over the past 18 months. Cash is a legit non-awful asset class again. Just look at this swing in yields.

Today, the challenge for the MMFs is different. The Fed funds rate is 5.3 per cent, and while there might be one more hike coming most analysts reckon we are at or near the peak in interest rates. Meanwhile, Treasury bills maturing in one year’s time have a yield of 5.5 per cent.

That’s triggered a big shift out of the RRP facility and back into bills. In the September 2023 data for the seven MMFs, the RRP allocation has fallen to 21 per cent of total assets, or $297bn. The size of the RRP facility itself has shrunk from $2.3tn to $1.2tn, as the Fed unwinds its QE holdings. 

Meanwhile, holdings of treasury debt — mostly T-bills — increased by $142bn, pushing the percentage allocation up to 28 per cent. But percentage-wise, the biggest increase has actually been into two other categories.

About $146 billion went into Treasury repo, where the MMFs are making short-term collateralised loans to non-US banks such as HSBC or Société Générale. 

Then there’s what we call US agency debt, which saw $134bn of inflows for the seven MMFs in this sample, accounting for 15 per cent of investments as of September 2023. The vast majority of this agency debt consists of discount notes issued by the Federal Home Loan Bank system, a US-government backed group of institutions that purchase mortgages issued by local banks. 

In a sense, this means that MMFs are underwriting the US housing market, as they did in 2007. Back then, their investments were in asset-backed commercial paper, a market which froze and resulted in the famous ‘breaking the buck’ when Lehman keeled over, prompting the Fed to backstop MMFs for the first time. 

Unlike ABCP, FHLB debt is effectively government-backed, and comes with strong collateral protections, meaning they are hardly likely to pose a threat to MMFs. In fact, the FHLB system has been criticised for exploiting such protections in order to lend to failed institutions like Silicon Valley Bank and First Republic Bank. 

The seven funds are also providing further indirect housing funding through $137bn allocated to agency repo, where they lend money to banks in return for housing agency bond collateral. 

As for commercial paper, this accounts for only 2.2 per cent of investments in the seven-biggest-funds data set, concentrated in a single fund, JPMorgan’s Prime MMF. Of this commercial paper, ABCP is only a tiny proportion. It’s also worth noting here that most of the growth in MMFs has occurred in government debt-focused funds that specifically exclude commercial paper as an asset class. 

Some people have worried about the risks potentially posed by the latest surge in money market funds, but when you look closely at the data many of these fears look a little overstated.

Concerns about investor runs make liquidity the biggest risk to MMFs. But the weighted average life of investments in the seven-fund $1.4tn data set has declined over the past 18 months, from 48 days to just 32 days at present. And with so much of their assets in short-term repo and T-bills, it’s hard to see how anything short of a cataclysm would cause a liquidity crisis.

And in any case, regulators are determined to punish investors that race each other to the MMF fire exits. This summer the SEC passed a rule requiring MMFs to impose a ‘liquidity fee’ if daily redemptions exceed 5 per cent of a fund’s assets. As the SEC’s chair Gary Gensler put it:

Money market funds — nearly $6 trillion in size today — provide millions of Americans with a deposit alternative to traditional bank accounts. Money market funds, though, have a potential structural liquidity mismatch. As a result, when markets enter times of stress, some investors — fearing dilution or illiquidity — may try to escape the bear. This can lead to large amounts of rapid redemptions. Left unchecked, such stress can undermine these critical funds. I support this adoption because it will enhance these funds’ resiliency and ability to protect against dilution. Taken together, the rules will make money market funds more resilient, liquid, and transparent, including in times of stress.

There’s also rumbling about the potential impact on banks, as competition from MMFs sucks liquidity out of them at a tricky time (those HTM losses are getting worse again).

A recent modelling exercise by the Dallas Fed warned of a non-linear convexity effect in bank deposits — as short-term rates increase, depositors ought to start fleeing en masse unless banks raise deposit rates to a higher and higher proportion of Fed funds rates. 

Given that MMFs are currently offering a deposit beta of effectively one, the question is why it hasn’t happened at an even greater scale. As the Dallas Fed concedes, there’s still $17tn in low-earning US bank deposits — of which $8tn are uninsured — dwarfing the $6tn of MMF assets. That implies pretty clearly that the danger isn’t quite as big as one might fear.

There’s no shortage of people looking for risks — and MMFs have a record of proving an unexpected one — but in this case the risks seem happy to remain out of sight.