Using the ‘Pottery Barn rule’ in the Treasury market

The US often makes policy by the “Pottery Barn rule” — if you break it, you buy it.

This usually just means that elected officials and their political parties are responsible for what they mess up. (It should probably be called the Pier 1 rule, after the liquidated discount home-goods retailer, but the current name is plenty evocative.)

The rule often applies to market-structure regulation, too. A great example is the regulatory focus on certain hedge funds after the Treasury market went haywire in March 2020.

As our colleagues Kate Duguid and Phil Stafford have reported, firms that trade more than $25bn in Treasuries monthly may in the future need to register as dealers with the Securities and Exchange Commission under proposed rules.

Our colleagues have also reported on the blowback the proposal has received from market makers, investment funds and other firms — almost everyone but the primary dealers, who are already heavily regulated, and some outside organisations like Better Markets.

This brings us to the main challenge of the Pottery Barn rule — showing that a certain group broke something.

Enter the Treasury’s Office of Financial Research, or OFR. This week it published new research that sure sounds like an argument for hedge-fund regulation, more than a year after it investigated the impact of hedge-fund cash/futures arbitrage on the March 2020 ructions.

The OFR looked at hedge-fund data that can’t be obtained by the public: positioning reports filed privately (via Form PF) with the SEC. And it argues that hedge funds’ moves have significant impact on Treasury market movements.

The paper estimates that a $41bn monthly change (one standard deviation) in hedge-fund positioning is responsible for around 6 bp movement in the 5-year yield. At the 5-year maturity, at least, the paper estimates that the effects of hedge-fund positioning changes are comparable to changes in inflation:

Based on monthly analysis dating from 2013 to the fourth quarter of 2020, we find economically significant and robust evidence that changes in hedge-fund exposures are related to Treasury yield changes. A one standard deviation increase in the net growth of Treasury exposures, which translates to a $41 billion monthly increase in hedge fund net exposures, is associated with a 6.2 basis point decline in five-year bond yields. The size of this estimate is not sensitive to controlling for well-known macroeconomic drivers of the yield curve, such as economic growth and inflation, and exists at various maturities. It is also robust to controlling for the valuation effects of yields on exposures and changes in other financial entities’ Treasury exposures.

The study’s authors argue that result holds even after adjusting for changes in monetary policy and other investors’ activity. They also found that the effect can’t be chalked up to hedge fund positions’ values changing with market moves.

This is completely intuitive to us. Hedge funds make up a larger share of the market’s investor base than they used to, and it seems entirely reasonable to think that changes in demand will move a market’s price! So the OFR’s dedication of resources to this argument makes the most sense through the lens of the current regulatory debate.

Before we dig into the paper’s other (rather more interesting) conclusions, we should review the years-long saga over rates-market regulation:

Treasuries have been a topic of interest since October 2014’s “flash crash” in yields, when traders at bank primary dealers stopped answering their phones (and at least one admitted to turning off his computer). Prop traders took some blame at the time, but industry executives noted that the moves lacked the classic signs of a meltdown in arbitrage strategies; price relationships that are normally policed by those trades held fast, notably the relationship between futures and cash markets.

Then those heavily arbitraged price relationships blew out in the early days of the Covid-19 pandemic. The strange moves and inefficient pricing led regulators — and our colleagues — to direct their attention toward arbitrageurs, including relative value funds, whose strategies came under pressure. Of course, that occurred in large part because their bank funding got very expensive, very fast. (They were also said to contribute to the repo market mess in September 2019.)

In other words: it could be argued that relative-value strategies broke the market in March 2020, so regulators are making them pay.

One response to this is that the US regulatory reaction — making a large range of traders register as government securities dealers — is overly broad and captures other types of funds that don’t use much leverage.

But the OFR finds that when March 2020 is excluded, the holdings of multi-strategy and managed-futures funds had the strongest relationships with price moves. They also include a chart breaking down the different types of hedge funds in the market by size:

First, this raises some questions about how multi-strategy and relative-value funds are classified. Is a pod shop with relative-value strategies (à la Millennium) multi-strategy, or relative value? The OFR says they use self-identification from Form PF filings, and it would be interesting to get more insight into how these funds classify themselves.

Second, even if you assume all the futures-arbitrage strategies are included in the relative-value category, the conclusion makes plenty of sense. Leveraged Treasury-market arbitrage trades were a problem in March 2020 because a spike in funding costs blew their businesses up, not because their presence makes Treasury trading perpetually volatile.

The paper indirectly touches on a controversial topic that has reportedly gotten the OFR into hot water before — the sometimes-blurred distinctions between systemically important intermediaries, and end-investors’ impact on markets.

After the GFC, officials made compelling arguments that banks’ central roles as intermediaries make them especially important systemically. Investors have to bear the consequences of their own market impact, according to this view, so they didn’t need to face the same level of regulations.

The problem is that the distinction can blur in less-regulated markets like Treasuries. We have heard from thoughtful folks that “making markets with a view” can be a highly lucrative investment strategy. And the OFR’s paper hints that having a large cohort of highly price-sensitive traders in a market can make it more volatile, particularly in times of high issuance and economic uncertainty:

Overall, these findings indicate that the trading activities of hedge funds can be linked to market price movements. At the same time, it is important to recognize that these findings do not show that hedge funds are the sole or decisive driver of price fluctuations in the Treasury market. Neither are they necessarily the source or originator of fundamental shocks that cascade through the financial system. Clearly, there are other forces that drive price movements in those markets. Extrapolating from this last point, it might be difficult to demonstrate that hedge fund trading during the March 2020 episode was the principal force behind the large fluctuations in Treasury yields and the decrease in liquidity. However, they might have served the role of an amplification mechanism.

There isn’t much the SEC can do about high levels of Treasury issuance and economic uncertainty (that’s the job of Treasury, Congress and the Fed). But it can expand its surveillance into Treasury markets to get a better idea of who these price-sensitive traders are, and put de facto limits on leverage.

If the goal were solely to create a robust Treasury market structure independent of political pressures, which is surely a fantasy in the US, regulators could be focusing on systemic solutions to deal with funding costs in times of crisis. They could look into clearing solutions, or make more policy about countercyclical changes in dealer regulations.

But if regulators are instead using the Pottery Barn rule, the hedge-fund focus makes perfect sense.