U.S. GSIBs recently issued a torrent of debt, with record breaking issuance sizes from JPM and then BAC. Yet, at the same time we know that the banking system has too much liquidity, and that banks are pushing out poor quality deposits to money market funds, who ultimately pour the excess liquidity down the ON RRP drain. The two behaviors can be reconciled by understanding the very strong regulatory incentives put on GSIBs to issue longer term liabilities. In response to the Financial Crisis, regulators devised a set of complementary regulations (Basel III) aimed at preventing future bank runs by setting limits on the size and composition of their balance sheets. Unsecured long term debt is favorably treated under all those constraints, and is especially compelling at near record low yields. We have already discussed how the Liquidity Coverage Ratio encourages longer term debt issuance. In this post we review the what and why of a few more Basel III constraints: the Net Stable Funding Ratio, GSIB Short-Term Wholesale Funding Indicator, and Total Loss Absorbing Capacity.
Net Stable Funding Ratio
The NSFR requires GSIBs to hold a level of “stable funding” in proportion to the liquidity of their assets. The liquidity of an asset is mechanically determined by the category it falls into, where assets created by the public sector like Treasuries are the most liquid and longer term loans to hedge funds/banks are the most illiquid. “Stable funding” is determined in the same way, with longer term debt (>1 year tenor) the most stable form of funding and short-term borrowings from institutional investors the least stable. Stable funding is essentially funding that is unlikely to be withdrawn over a 1 year horizon – either because the funding is contractually bound (longer term debt) or statistically unlikely to be withdrawn (retail deposits).
The NSFR aims to strengthen a bank’s ability to meet outflows over a 1 year time period. For example, Treasuries are highly liquid and do not require any stable funding as they can readily be sold to meet any withdraws. But long term loans to a hedge fund or bank are illiquid (especially in times of market stress). So every $100 of such loans must be backed by $100 in stable funding. In the event of a panic, that $100 in stable funding will stay put so the bank would not be forced to fire sell the $100 loan to meet withdraws. The NSFR was finalized in October 2020 and comes into effect this July, with almost all banks estimated to already be in compliance.
GSIB Short-Term Wholesale Funding
The infamous GSIB Surcharge is actually determined by 5 sub-components, one of which is the Short-Term Wholesale Funding Indicator. This indicator aims to measure the extent and quality of a bank’s short-term funding (i.e. maturity <1 year). A GSIB is asked to report its short term funding activity according to maturity (e.g. 31 to 90 days, 91 to 180 days), counterparty type (e.g. financial sector, retail), and security interest (e.g. secured by Treasuries, Agency MBS). Each funding bucket is then run through a weighting formula to arrive at a weighted sum. The weightings are set by the Fed where greater weights are placed on shorter maturities, lower quality collateral, and financial sector entities. For example, 6 month Treasury repo has a weight of 0 while wholesale unsecured funding from another bank receives the highest 0.75 weight. The weighted sum is then normalized by the GSIB’s risk-weighted assets.
The Short-Term Wholesale Funding Indictor is aimed at discouraging GSIBs from using short-term funding, as a higher indicator score may lead to a higher GSIB surcharge. GSIBs are thus encouraged to use longer term funding (or retail deposits).
Total Loss Absorbing Capacity
TLAC forces a GSIB to hold loss absorbing liabilities in proportion to the size and composition of its balance sheet. Liabilities that qualify as TLAC include equity and longer dated unsecured debt (>1 year tenors). Unlike the NSFR and the GSIB Short-Term Wholesale Indicators (and the LCR), TLAC’s concern is not on liquidity but on post-bankruptcy resolution.
When a company is insolvent, its liabilities exceed its assets, so its equity becomes worthless. The company’s assets are distributed through a bankruptcy proceeding to creditors, who have varying levels of seniority and different security interests. The process involves extensive negotiation and litigation as each class of creditors fights over the company’s assets. At the very bottom of the creditor hierarchy are the subordinated unsecured creditors (aka Tier 2 Capital), then the unsecured creditors. In that sense, unsecured creditors are like equity in that they absorb the bulk of the losses in the event of bankruptcy. A thick layer of unsecured creditors acts as a cushion for the other creditors.
TLAC was no doubt drafted based on lessons from the Lehman bankruptcy. Lehman’s bankruptcy proceeding was a mess that lasted over 10 years because it spanned multiple legal entities/jurisdictions and a web of claims where even senior secured creditors did not know what they would recover. TLAC helps fix this by 1) increasing the loss absorbing capacity of a GSIB so its senior creditors would not panic and 2) consolidating that capacity at the holding company level to backstop all the operating subsidiaries (this way creditors of subsidiaries would not panic). Together this provides more clarity and hopefully stability in the event of a GSIB bankruptcy.
Note that some level of longer dated unsecured debt is required under TLAC, with the required level set as a ratio to both risk weighted assets as well as an unweighted assets measure.
Money Market Funding is Bad
The message to GSIBs is very clear – issue longer dated debt and do not use money market funding. Even though short term rates are lower, the regulatory costs are very high. (This is also a reason why the Fed funds market has died: overnight, unsecured inter-bank loans are the worst possible loans under all Basel III regulations). In a world where rates are low and funding is superabundant, GSIBs have the freedom to move towards their optimal funding structure. That structure is lots of retail deposits, with some long term unsecured debt at low rates.
It is what pleases the gods of Basel.
4 comments On Why Are Banks Issuing So Much Debt?
Thanks for this – v interesting and relevant given these huge deals last week.
To the extent that JPM runs a “matched repo” book, ie same repo counterparty, same maturity repo and even same issuer collateral, would the repo side be included in the above or would that get netted off (in the chart described as “This is JPM’s funding profile as of year-end 2019”)
There are two concepts in your question – 1) matched book repo and 2) netting. Matched book repo is more general than what you described – it is simply borrowing in repo to lend in reverse repo. When the transactions are the same counterparty/collateral/tenor (common when cleared in FICC) then it is possible to net the exposure so that all that is left on the balance sheet is either the net liability or net asset (as always, there are more rules to this but that is the general point). I think they would’ve displayed the netted result in the profile, as that is more accurate (and flattering).
Thank you always Fed guy, your post has been both super informational and educational. Great to see your stuffs. One question tho, as I understand, even tho very short term funding like o/n unsecured repo is not recommended under basel regulations, o/n repo backed by treasuries as collateral is still considered pretty stable funding method right? is there any possible effects of this regulations on o/n secured repo market?
Yes O/N Treasury Repo is fine for liquidity (slight to no regulatory cost, depending on the reg). This is because the lender holds risk-free Treasuries as collateral, so it is assumed that their loans to a bank are stable. However, repo borrowing expands a bank’s balance sheet. This impacts their leverage ratio (both Tier 1 leverage ratio and Basel III SLR). If they are a GSIB it will also impact their GSIB Surcharge, which takes into account the size of a bank’s balance sheet.