Treasury troubles revisited

By now the US Treasury market’s unnerving fragility back in 2020 has been pretty comprehensively dissected. But the Office of Financial Research last week published an interesting new take on one of the freakiest financial mishaps of the past decade.

For people blissfully unaware of what happened, in March 2020 Treasuries went from being an investor bomb-shelter — ultra-safe and ultra-liquid assets that rally when markets fall — to being both bizarrely illiquid and selling off in tandem with the wider rout. Here is a good mainFT recap of a period where some investors felt they were “looking at doomsday”.

The US Treasury’s Office of Financial Research — set up in the wake of 2008 expressly to scour the financial world for systemic risks — has published a new paper titled Fragility in Safe Asset Markets revisiting the episode. Here is the abstract:

In March 2020, safe asset markets experienced surprising and unprecedented price crashes. We explain how strategic investor behavior can create such market fragility in a model with investors valuing safety, investors valuing liquidity, and constrained dealers. While safety investors and liquidity investors can interact symbolically with offsetting trades in times of stress, liquidity investors’ strategic interaction harbors the potential for self-fulfilling fragility. When the market is fragile, standard fight-to-safety can have a destabilizing effect and trigger a “dash for cash” by liquidity investors. Well-designed policy interventions can reduce market fragility ex ante and restore orderly functioning ex post.

Basically, the authors (the NY Fed’s Thomas Eisenbach and the OFR’s Gregory Phelan) developed a simplified theoretical model of a market with three distinct types of players: investors that value safety above all, investors that value liquidity above all, and intermediating dealers that face balance sheet constraints.

They found that most of the time — including at times of stress — these players “interact symbiotically with offsetting trades”, but in really acute times of turmoil, liquidity-oriented investors can start to anticipate worse prices and seek to sell preemptively, “leading to self-fulfilling market runs that are like traditional bank runs in some respects”.

From an accompanying blog post on the OFR’s website:

The authors show that flight-to-safety episodes can exacerbate the dash for cash when markets are fragile. Demand by flight-to-safety investors early on in a stress episode increases prices both contemporaneously and, by relaxing dealer balance sheets, in the future. If the strategic concerns of liquidity investors are sufficiently strong, then additional demand from safety-first investors today can induce liquidity investors to sell today, precisely because the market today has a relatively higher capacity to absorb sales. Then, a flight-to-safety triggers a dash for cash, amplifying existing market fragility.

A lot of this might seem like a long-winded, academic and theoretical way of saying something that any PTSD-suffering Treasury trader could have told you in seconds years ago. There have been other more granular and practical examinations of the episode by the likes of the BIS.

But 1) the OFR paper hints at further support for one intriguing potential solution that Alphaville has written about before: shifting the Treasuries market towards all-to-all trading. The NY Fed last year published a report (co-authored with Treasury and SEC staffers) that also floating this as a possibility.

And 2) it makes clear that the Fed’s aggressive repo facilities failed to stem the Treasury ructions, because banks that intermediate Treasury trading were still hindered by regulations that penalise swelling balance sheets. As the blog post notes, balance sheet-constrained banks can become “bottlenecks during times of stress, causing price volatility to increase in safe asset markets”.

It was the Fed’s subsequent shift towards outright Treasury purchases and the announcement that Treasuries and Fed reserves would be exempt from the “supplemental leverage ratio” that helped calm things. As the OFR paper notes:

If the SLR had been relaxed earlier, including for Treasury repos, fewer purchases may have been necessary to ensure that the market remained stable.

The temporary Treasury exclusions from SLR calculations ended in Mach 2021. And there are good reasons why the SLR was expressly designed to be blunt, comprehensive and not distinguish between the riskiness of different assets. As the recent US banking shenanigans have shown, supposedly “safe” assets can also cause carnage in the right circumstances.

But the OFR paper argues that temporary relaxation of the SLR should maybe become a more regular feature to heal any re-emerging Treasury market stresses.

Of course, the SLR and other post-GFC regulatory constraints were introduced for good reason. However, the fact that these constraints interfered with intermediation in safe assets during times of stress seems like an unintended consequence. For future stress episodes, a temporary relaxation of the SLR would therefore improve market stability in our model. Indeed, relaxing the SLR in times of stress would be a way to mitigate the fragility emphasized in our paper while maintaining the stabilizing macroprudential consequences for which the regulation was designed.