The Repo Logjam in FICC: A Primer
SEE LAST PAGE OF THIS REPORT FOR IMPORTANT DISCLOSURES
May 27, 2016
The Repo Logjam in FICC: A Primer
“The marginal cost of repo financing is driving the derivatives market and swap spreads in particular” Ken Griffin, CEO Citadel, March 2016
- Regulatory reform has broken an important portion of the swaps market and only intra-day settlement for repo transactions can provide a comprehensive solution; this is not supported by current settlement technology but could be blockchain-enabled. The specifics are that the cost of funding a treasury position in the repo market has increased from ~15bps in November to ~40bps today (generalized collateral finance or GCF rate at DTCC) with a commensurate decline in 10-year swap spreads now at negative 15bps; this has increased the cost, which incorporates the swap spread as an offset, to corporate borrowers of swapping fixed- to floating-rate obligations and hence contributed to higher borrowing costs and adverse volume trends in the fixed-income markets.
- The root problem is that securities-dealers, in calculating the supplemental leverage ratio or SLR, cannot net repo transactions with different counterparties and, in particular, cannot net a reverse-repo transaction with a cash-borrower (also called a securities lender because of the securities pledged as collateral), such as a hedge-fund, against the offsetting “matched book” repo transaction with a cash-lender such as a money-market fund or corporate treasury. Dealers have passed these balance-sheet costs on to clients which is reflected in rising repo rates (i.e. the interest rates paid on loans with securities pledged as collateral) and reduced the supply of pay-fixed swaps (because these are typically provided by hedge funds against treasury positions funded in the repo market). The result is higher swap spreads as swap rates have risen relative to treasury rates.
- A natural solution is to move to comprehensive clearing in the repo market, as in the futures and OTC derivatives markets, where a central-clearing counterparty or CCP is accessed by end-clients via member agency-clearers that, in practice, are securities-dealers. The CCP becomes a common counterparty for cleared repo transactions which therefore can be netted in a calculation of the SLR; the challenge is that clearing transactions for non-bank borrowers increases CCP exposure and hence the default-fund which regulators require members to establish so that CCPs can finance the purchase of a defaulted member’s portfolio – with, for example, the DTCC announcing last year a default fund, referred to as the capped committed liquidity facility or CCLF, of $50bn that is expected to be introduced this summer subject to SEC approval. These CCP default-fund obligations, even if committed and not funded, contribute to exposure in dealer SLRs.
- It follows that dealer SLR is increased by repo transactions for non-bank borrowers whether because of the inability to net non-cleared transactions or the impact of default-fund obligations in cleared transactions. Buy-side clearing, offered only to cash-lending counterparties, such as money market funds, who do not increase clearing exposure (which is always to cash-borrowers) provides some netting benefit without increasing default-fund requirements. However, this will not restore supply-demand balance in the swap market given non-bank borrowers disproportionately pay-fixed. The upshot is an example where regulatory reform has shifted risk (from securities-dealers to clearers), but not reduced it; reducing risk will require new technology to reduce settlement times and errors, and this has motivated DTCC to work with Digital Asset Holdings on a blockchain solution for intra-day settlement and hence netting.
- The specific opportunity is that overnight repo transactions are typically rolled but, because settlement is at best T+1, it is not possible to net the expiring transaction against the replacement transaction. As Murray Pozmanter, representing the systemically important financial utility business at DTCC, remarks “if someone is doing the same transaction two days in a row, when the first trade comes off you should be able to offset it against the new trade that you’ve done; you can only do that if you have intra-day netting which the current technology does not allow”.
Regulatory reform is changing FICC market structure and hence the balance of supply and demand in the interest rate markets particularly at the short-end of the curve. One of the most obvious examples is in the repo market for treasuries and agency MBS where cash lenders, such as money-market funds and corporate treasuries, generate yield by engaging in reverse-repo transactions (buying a security with an agreement to sell it back in the future) which amount to collateralized loans and securities lenders, such as hedge funds, access liquidity and finance positions by engaging in repo transactions (selling a security with an agreement to buy it back in the future) which amount to collateralized borrowing. Beyond meeting their own short-term funding needs as securities lenders, securities-dealers intermediate the repo market through “matched” books where they finance a borrowing client in a reverse repo transaction and re-hypothecate the securities-collateral to a lending client, or other securities-dealer if no client is available, in a repo transaction.
While this activity provides important liquidity to the market, particularly for hedge funds needing to access liquidity, it has become expensive for most dealers because the supplementary leverage ratio (SLR), with a regulatory minimum of 5% by Jan 1, 2018 for covered US bank holding companies according to a rule finalized by the Fed in April 2014, allows the netting of offsetting trades only with the same counterparty and not the netting of a reverse repo trade with one client and its matched trade with a different client. Given it offers relatively low return for use of the balance sheet, banks have scaled back their matched books to improve SLR so that the aggregate balance of inter-dealer reverse repos (and fed funds) at US commercial banks, which had ranged between ~$100-120bn from mid-2011 through mid-2014, albeit closer to $400bn before the financial crisis, fell to ~$50bn over the next year where it remains today. And, the reduced capacity for repo has pushed up rates so that mid-pricing for certain overnight treasury repo rates is now ~40bps from ~15bps as recently as November 2014 (Chart 1).
Chart 1: DTCC Repo Index for Treasuries (General Collateral Finance or GCF rates)
While deep in the plumbing of Wall Street, this change in the treasury repo market has important consequences for the FICC markets generally as Ken Griffin, CEO of Citadel, highlighted in March: “the marginal cost of repo financing is driving the derivatives market and swap spreads in particular.” One specific knock-on effect is that the increased cost of financing treasury inventory has pushed long-term swap spreads, the difference between the swap rate and the treasury rate, into negative territory since mid-2015 (Chart 2); and this has, in turn, reduced liability-hedging by corporate clients. We will clarify the chain of consequence before outlining our bullish thesis on FICC around how the repo logjam can be broken.
Chart 2: 10-Year Swap Spread (Swap Rate Less Constant Maturity Treasury)
The Repo Logjam
Corporations often issue fixed-rate debt to take advantage of pockets of investor demand in a particular term and currency, and swap it into floating-rate to meet internal risk tolerances. Given that fixed-rates are benchmarked to treasury yields while floating rates are benchmarked to swap rates, the swap spread as the difference between swap and treasury rates of the same term, impacts this trade. For example, if the swap spread is 10bps, fixed-rate debt issued at 100bps over treasuries can be swapped into a floating rate debt at 90bps over Libor. Today, however, the swap spread is negative 15bps so, on the same fixed-rate transaction, the issuer would end up paying a floating rate of 115bps over Libor.
Swap spreads have historically been positive reflecting the fact that the swap rate is the benchmark rate at which a bank can borrow fixed-rate funds while the treasury rate is the rate at which the Federal government can borrow; the swap spread, therefore, has traditionally provided a measure of the incremental credit risk involved in term-lending to banks rather than the Federal government. However, given that swaps are subject to mandatory central clearing, they survive the default by a bank counterparty so that exposure in the swaps market to bank credit risk is for all practical purposes eliminated. The result is that swap spreads trade today more as a function of supply and demand than of market perception of bank default risk. (Instead, as noted in our research piece of 2/8 titled “Systemic Risk and Deutsche Bank”, the Libor-OIS spread is a preferred ‘barometer of fears of bank insolvency”).
What, then, are the supply-and-demand factors driving the swap spread? In broad terms, the natural demand for receive-fixed swaps is from corporates, pension funds, and insurance companies looking to hedge fixed-rate liabilities; this demand is met by a supply of pay-fixed swaps from asset managers and hedge funds looking to hedge or position portfolios; in particular, shorting the swap spread (by paying fixed in the swap market against a portfolio of treasuries) has been a traditional positioning for a “risk-off” environment when the swap spread tends to widen as swap rates increase with concerns around bank default and treasury yields decline in a flight to safety. However, these “swap-spread widener” trades their usual leveraged form depend on funding treasuries in the repo market and generating positive carry through the difference between the receive-floating leg of the swap, typically libor-based, and the overnight repo rate. Given this context, it is not surprising that the 25bps increase in this repo rate since last November has been accompanied by a commensurate decline in the swap spreads. This, in turn, increases the cost of hedging fixed-rate liabilities and has led to a reduction in natural demand for receive-fixed swaps from corporate borrowers with an adverse impact on FICC volumes.
Clearing the Repo Logjam
The root problem is that, for the purposes of calculating the supplementary leverage ratio, banks are not permitted to net repo trades with different counterparties and hence, in particular, are not permitted to net offsetting repo transactions a matched book; the natural solution is to move to comprehensive clearing, as exists in the futures and OTC derivatives markets, where a central clearing counterparty or CCP is accessed by end-clients through agency clearer “members”, in practice comprising securities-dealers, who commit to supporting the CCP through establishing a “default fund” to finance the purchase of a defaulted member’s portfolio. The challenge to this solution is that providing clearing for non-bank borrowers (i.e. securities lenders) meaningfully increases the exposure of the CCP and hence the size of member default-fund obligations; and these default-fund obligations, even if committed not funded, themselves contribute to exposure in the calculation of member SLRs.
A solution being discussed at the DTCC, which last year announced a default fund – referred to as the capped committed liquidity facility or CCLF – of $50bn for its systemically important financial market utility business, to comply with regulatory requirements, is to limit clearing to money market funds which act in the repo market only as cash-lenders and so do not increase clearing exposure. This reduces the SLR impact on member dealers of the default fund but lessens the opportunity for netting matched repo trades involving an MMF on one side and a non-bank borrower on the other. In short, dealer SLR is increased by repo transactions for non-bank borrowers whether because of the inability to net non-cleared transactions or the impact of default-fund obligations in cleared transactions.
Repo transactions with non-bank borrowers provide an example where regulatory reform has shifted risk (from securities-dealers to clearers), but not reduced it; reducing risk will require change in the underlying infrastructure to reduce settlement times and errors and this has motivated DTCC to work with Digital Asset Holdings on a blockchain solution for intra-day settlement and hence netting. The specific opportunity is that overnight repo transactions are typically rolled but, because settlement is at best T+1, it is not possible to net the expiring transaction against the replacement transaction. As Murray Pozmanter, representing the systemically important financial utility business at DTCC, remarks “if someone is doing the same transaction two days in a row, when the first trade comes off you should be able to offset it against the new trade that you’ve done; you can only do that if you have intra-day netting which the current technology does not allow”.
©2016, SSR LLC, 1055 Washington Blvd, Stamford, CT 06901. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein. The views and other information provided are subject to change without notice. This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results. The analyst principally responsible for the preparation of this research or a member of the analyst’s household holds a long equity position in the following stocks: JPM, BAC, C, WFC, and GS.