A few have asked, so I’ve written up what is actually a shorter piece on this SLR business is all about. First, SLR stands for Supplementary Leverage Ratio (and it’s not SLF, as I managed to leave two of the same typos in the main article referenced below, to the point the mistake made it into the headline). Parts of the SLR were set aside last year as a COVID crisis tool, but are scheduled to come back at the end of March 2021.

That’s the “cliff.”

What’s the SLR and where did this thing come from?

Short version: capital ratios had failed miserably to alert bank regulators or anyone else in the pre-crisis era to a dangerous and sustained systemic increase in leverage. What used to be called commercial banks (or investment banks) figured out how to manipulate them, relying heavily on instruments like credit default swaps (the majority of AIG’s CDS book, for example, had been used for just this purpose), to achieve something called regulatory capital relief.

Basically, how to make risky assets seem safe according to the rules for the purposes of being forced to hold less capital against the whole bunch. Hidden leverage.

To fix, or attempt to fix the problem, regulators got together in the smoldering aftermath of what they called a “subprime mortgage crisis” deciding to supplement capital ratios with another calculation that doesn’t weight assets for different risk “buckets.” By treating everything the same, it was hoped the new number would better and more accurately depict the full leverage used by every institution.

No matter if assets are safe or downright dangerous, this SLR measures the whole balance sheet in more homogenized fashion (including derivative books and off-balance sheet positions; though, it needs to be pointed out, this is neither straightforward nor uncontroversial, but that’s beyond the scope of this post). In other words, where a UST security is assigned a very low risk-weighting when calculating Tier 1 Capital ratios, it is weighted the same as any other asset for the purposes of the SLR.

In the aftermath of GFC2 (the COVID crisis, according to central bankers), in May 2020 the US bank regulator triplets approved the following change in the SLR calculation:

Issued by the Federal Reserve Board, the Federal Deposit Insurance Corporation, and the Office of the Comptroller of the Currency, the interim final rule permits depository institutions to choose to exclude U.S. Treasury securities and deposits at Federal Reserve Banks from the calculation of the supplementary leverage ratio.

This “relief” was deemed necessary since the federal government was in the process of selling an immense amount of debt which the Federal Reserve thought had been a huge problem back in March of last year (when that hadn’t been the problem at all). In addition, the same third of the redundant trio of regulators was at that time creating a flood of digital bank reserves (“deposits at Federal Reserve Banks”) which likewise would’ve bumped down SLR ratios for the banks doing QE business with the Fed (note: all banks are now “depository institutions” for the purposes of regulations, even those which aren’t).

Hoping to avoid a situation where banks balked at either QE or UST issuance because of what it might do to their SLR numbers, the triad agreed to this temporary carveout for Treasuries and bank reserves.

Again, the relief was set to expire on March 31, 2021. Our SLR “cliff.”

What that means is come April Fool’s all the bank reserves and UST’s banks have acquired over the ten and a half months in between will suddenly accrue to SLR calculations (denominator), possibly reducing them to the point of being too low or incurring a surcharge applicable to designated Globally Systemically Important Banks (or G-SIB’s).

You might recall that this same issue had come up in the aftermath of September 2019’s repo rumble, with Jamie Dimon attempting to blame the SLR and possible surcharges arising from it for the debacle (as well as his bank’s role in it). From October 2019:

To put it very simply, regulators would require designated G-SIB’s (there are 8 in the US) to hold an additional liquidity buffer (SLR) based upon their reported leverage (one lesson authorities did learn from Bear, Lehman, and the rest). It would be a liquidity surcharge which would have to be met by a set percentage of holdings in unencumbered cash and highly liquid assets like UST’s over and above other regulatory and good standing requirements.

But, you’ll undoubtedly notice, UST’s and highly liquid assets (such as bank reserves) are the whole problem with the SLR “cliff” to begin with. Meaning, that G-SIB’s who bought all these USTs and transacted with Fed QE to end up with bank reserves raising their SLR denominators will then be required to hold more liquid assets in the form of UST’s and bank reserves they already acquired?

While it sounds circular, and a small bit ridiculous, the key word in my quote above is “unencumbered.” Bank reserves are that, sure, but with UST’s it may not be so simple.

Treasuries in general are the best collateral, especially bills which are the best of the best. But if you can’t have them unencumbered (can’t repledge or rehypothecate), maybe the SLR cliff would drive balance sheet capacity scarcity for repo/collateral factors.

What to do?

If it requires offsetting positions to make sure that a bank’s balance sheet hasn’t grown “too much” during the SLR relief period, this might entail having to sell assets to get in under thresholds. Given a choice, while these banks can use bank reserves to satisfy some surcharges, maybe they’d prefer selling off long end UST’s in favor of holding onto T-bills (or even buying more).

UST notes and bonds are exhibiting price risk as we speak, whereas bills are reliably in huge demand (too much demand?) Might this account for the market-based twist (falling rates in bills, rising in bonds and longer notes) in the UST curve, and not just in the US?



Perhaps.

The bigger consideration, and the one completely missed by what limited discussion has taken place, is that no one nor any banks should care about the SLR or its cliff possibilities. For a very long time (the eurodollar system’s “golden” age) banks would sell their own mothers (presuming banks ever had maternal origins) to surmount otherwise onerous bank regulations in the past – only starting with capital ratios.

As I wrote earlier today (recalling the story of Herstatt, and how its closure in 1974 set off the chain of events leading to a Basel Committee then capital ratios and finally this SLR), once past all the initialism typos:

Balance sheet space used to run cheap and easy, and the whole system worked based on that, it is now hugely prized and is treated way too precious – and that alone accounts for this tendency for markets to behave intermittently like they once had in the early summer of 1974. And that includes September 2019 and March 2020. Money dealers didn’t need the SLR to just disappear.

It had been a convenient excuse for at least one of the CEO’s of those dealers.

That’s the problem; balance sheet capacity and space has become “too expensive” as a consequence of behavioral changes (not regulations which years afterward simply codified already modified behavior). If the shadow money was truly flowing, with legitimate inflationary opportunity around every corner, banks big and small wouldn’t do anything about the SLR cliff other than pay the toll (in whatever form) positively confident in being able to get it paid back by a truly booming system.

If anyone, and I mean anyone, is slightly bothered by this SLR stuff then balance sheet space (including collateral concerns) isn’t being conducted in any sort of inflationary manner.

In other words, just the possibility that the SLR might be causing some upset is yet more proof there hasn’t been a flood, or even a trickle, of effective money creation – which is balance sheet capacity governed by a willingness to take on risk that, still, after going on thirteen years, just can’t be there.

In that sense, this SLR cliff might tell us something else entirely about that state of things.