Speech Collateral and Liquidity: Striking the Right Balance
Deputy Head of Payments Policy (Financial System)
Global Investor/ISF Australia Conference
Good morning. It is a pleasure to address this gathering today and I would like to thank Global Investors and ISF Australia for the opportunity. The topic I would like to cover this morning is collateral and liquidity. In particular, I'd like to explore important changes in the way collateral is used and what this might imply for policymakers – and central banks in particular.
Regulatory developments and changes in financial institutions' risk preferences are fundamentally altering the way liquidity flows through the system and how institutions fund themselves. These include: the expansion of central counterparty (CCP) clearing; margining of non-centrally cleared derivatives; higher capital charges for uncollateralised exposures; an emerging investor preference for collateralised lending; margin segregation; increasing caution around rehypothecation and re-use of collateral … The list goes on. Given the experience of the global financial crisis, there are good reasons for all of these changes. At the same time, central banks and other regulators can help to smooth the transition to the eventual new steady state, including by ensuring appropriately measured implementation of new regulations.
An economist at the International Monetary Fund (IMF), Manmohan Singh, has written extensively on this topic. In a paper presented at the RBA conference on liquidity and funding markets last year, Singh emphasised the ‘financial lubrication’ role of collateral and the importance of well-functioning collateral markets for channelling liquidity within and between the banking and non-bank financial sectors (Singh 2013). He observed that the effectiveness of these markets may already have declined since the onset of the financial crisis. The broader implications of this had probably not yet been felt, however, because less effective circulation of collateral through the system had to some extent been offset by increased issuance of government debt and easy monetary policy.
But in time these changes seem likely to alter established market practices and reshape incentives. In the remainder of this talk I will discuss some of the key drivers and implications of these changes. I will also highlight where central banks or other regulators' responses could potentially help to strike the right balance.
Increased Demand for High-quality Liquid Assets
Much has already been written on the drivers of increased demand for eligible high-quality liquid assets. In an article co-authored with my colleague Alex Heath a couple of years ago (Heath and Manning 2012), we highlighted two key regulatory reform drivers of increased demand.
First, new liquidity standards introduced under the Basel III reforms. These standards include minimum requirements for the size and composition of banks' liquid assets: the Liquidity Coverage Ratio. In particular, banks will have to hold sufficient high-quality liquid assets to be able to withstand 30 days of outflows under stressed market conditions. For these purposes, the Australian Prudential Regulation Authority (APRA) has defined high-quality liquid assets to include cash, central bank reserves and government securities. Although these requirements do not formally take effect in Australia until the start of next year, we have already seen a marked change in Australian banks' liquid asset holdings. Since late 2006, the total quantum of liquid assets on banks' balance sheets has risen from around $100 billion to just over $300 billion. Adjusting for the overall growth in balance sheets over that period, this equates to an increase from around 7 per cent to 11 per cent of total assets. Within this, the share of short-term bank paper has declined from more than 50 per cent to just 15 per cent, with a particularly marked shift into government securities (Table 1).
|December 2006||December 2011||December 2013|
|$ billion||Share(a)||$ billion||Share(a)||$ billion||Share(a)|
|– Government securities(c)||10||10||72||28||126||41|
|– Short-term bank paper||54||52||64||24||44||15|
|– Long-term bank paper||9||9||74||29||60||19|
|Total bank assets||1,582||2,597||2,896|
(a) Share of total Australian dollar assets (per cent), subcomponents are
the share of liquid assets
Sources: ABS; APRA; RBA
Second, reforms that encourage or mandate CCP clearing of standardised over-the-counter (OTC) derivative trades that require collateralisation of non-centrally cleared trades will significantly increase the demand for collateral.
Collateral agreements in these markets are not new. Since well before the financial crisis, the proportion of non-centrally cleared trades that are subject to collateral agreements has been on the rise. Indeed, the most recent annual margin survey conducted by the International Swaps and Derivatives Association (ISDA) reports that around 90 per cent of transactions in credit, fixed-income and equity derivatives are subject to a collateral agreement. A smaller proportion of trades in foreign exchange and commodity derivatives are subject to such agreements, in part reflecting the greater participation of non-financial entities in these product markets (Graph 1).
Agreements to date have typically provided only for the exchange of collateral to cover current mark-to-market exposures; that is, variation margin. Looking ahead, however, OTC derivative trades, whether or not centrally cleared, will also be subject to collateral to cover potential future exposures should a counterparty default and the positions have to be closed out; that is, initial margin.
Initial margin requirements are already beginning to bite – at least among interdealer trades in the more standardised OTC derivative markets, such as that for interest rate derivatives. The majority of outstanding notional value of OTC interest rate derivatives is already centrally cleared and therefore subject to both initial and variation margin requirements. In some jurisdictions, such as the United States, this reflects that central clearing is already mandatory. In others, market participants have transitioned to central clearing either in anticipation of a mandate, or simply due to a commercial incentive as liquidity has shifted to the cleared market.
The most widely used CCP in the global OTC interest rate derivative market is the SwapClear service operated by LCH.Clearnet Limited (LCH.C Ltd). Since the global financial crisis, the outstanding notional value of OTC interest rate derivatives cleared through SwapClear has increased to more than US$200 trillion. At the same time, the initial margin posted by participants to support this activity has also risen significantly.
The transition to central clearing is less advanced in other OTC derivative product classes and with initial margining of non-centrally cleared derivatives not yet widespread the implications for collateral demand have yet to be fully felt on collateral demand. A number of studies have attempted to quantify the likely ultimate effect of both central clearing and initial margining of non-centrally cleared trades. These have reached a variety of conclusions, depending on assumptions around matters such as the volatility of the underlying products and the degree of netting efficiency. Several studies – including Duffie and Zhu (2011) and Heath, Kelly and Manning (2013) – observe that central clearing of multiple products through a single CCP is likely to deliver the greatest netting efficiency and as a result could temper the demand for collateral-eligible assets. Indeed, Duffie, Scheicher and Vuillemey (2014) estimate that central clearing of credit derivatives could reduce collateral demand by almost a third relative to a scenario in which these products remained non-centrally cleared but were subject to initial margin. Anderson and Jõeveer (2014), on the other hand, model conditions under which non-centrally cleared arrangements in regional markets could, by accommodating a wider range of local collateral assets, be less costly than clearing through CCPs.
The Australian authorities have closely monitored these developments, paying particular attention to the way regulatory changes are implemented. A case in point is mandatory clearing requirements for OTC derivatives. The regulators recognise the trade-off between the benefits of collateralisation in managing counterparty credit risk and the costs in terms of increased liquidity risk. Acknowledging that some market participants may face liquidity constraints, the regulators favour careful implementation of central clearing requirements. In their latest recommendations to government, the regulators observed that where small financial institutions and especially non-financial entities had restricted access to liquid assets to meet CCPs' initial and variation margin requirements, new sources of risk could emerge (ASIC, APRA and RBA 2014). Accordingly, the regulators concluded that mandatory clearing requirements should be imposed only on internationally active dealers. In particular, they noted that:
… for some non-dealers it is unclear whether either the private or public policy benefits will ever be sufficient to offset the costs. Given this, on the basis of currently available information, the Regulators would expect to give close consideration to a specific exclusion from any mandatory clearing obligation for certain non-dealers. ASIC, APRA and RBA 2014, p 4
Similarly, Australian regulators argued internationally for a narrower product scope for mandatory initial margining of non-centrally cleared derivatives that excluded foreign exchange derivatives and the currency component of cross-currency swaps. And, also recognising the limited supply of high-quality Australian dollar-denominated assets eligible to meet the Basel liquidity standards, APRA has permitted authorised deposit-taking institutions (ADIs) to access a committed liquidity facility made available by the Reserve Bank. I will return to this shortly.
The fundamental role of collateral is to manage counterparty credit risks in wholesale markets. There are a number of reasons why collateral is an effective tool for this:
- Collateral overcomes problems of informational asymmetry; a collateral receiver need only monitor the quality of the collateral, not its counterparty.
- Collateral offers pre-funded credit protection, since the collateral receiver has the assets in hand at the time of default.
- Secured lending and other forms of collateralisation of exposure overcome difficulties in writing complex contracts that can anticipate all possible market and economic circumstances; legal agreements that underpin collateral arrangements by contrast are relatively simple and can rely on standardised documentation.
Given its role and purpose, collateral typically takes the form of high-quality assets that are subject to limited credit and market risk and easy to monitor. Collateral-eligible assets typically must also be liquid, since in the event of a counterparty default, the holder of collateral will aim to liquidate the assets on a timely basis.
For instance, repurchase agreements – or repos – are typically written on a defined set of high-quality assets. In Australia, most repos are contracted against Commonwealth Government Securities or securities issued by the states and territories. As collateral becomes more scarce, however, market participants may begin to accept a broader set of collateral assets than solely high-quality liquid assets.
The CPSS-IOSCO Principles for Financial Market Infrastructures (the Principles) set international standards for CCPs and other financial market infrastructures (CPSS-IOSCO 2012). The Principles state a clear preference for collateral assets ‘with low credit, liquidity and market risks’, and set similar expectations around the assets in which a CCP reinvests any cash collateral received. Other assets may be eligible ‘if an appropriate haircut is applied’. In the centrally cleared space, some CCPs already accept a relatively wide range of collateral assets to meet margin obligations (Table 2). For instance, in addition to cash and government bonds, CME and Eurex accept various private debt securities. While in most cases, cash and government bonds still predominate, this flexibility is valuable to some participants.
|Money-market fund shares|
(a) Various currencies/governments
Sources: CCP websites
For non-centrally cleared transactions, cash remains the most common form of collateral. ISDA reports that around 75 per cent of collateral is currently in cash form, with around 15 per cent in government securities and 10 per cent in other fixed-income securities, including corporate bonds (ISDA 2014). As initial margin is more frequently exchanged in these markets, non-cash collateral is likely to become more important. Similar to the Principles, new standards established by the Basel Committee on Banking Supervision (BCBS) and International Organization of Securities Commissions (IOSCO) require that collateral assets ‘be highly liquid and should, after accounting for an appropriate haircut, be able to hold their value in a time of financial stress’ (BCBS-IOSCO 2013, p4). BCBS-IOSCO (2013) provides an illustrative list of eligible collateral, which extends well beyond traditional high-quality liquid assets.
Clearly, there's a balance to be struck. And that's where the central bank comes in. A central bank, uniquely, can use its balance sheet to transform financial assets of differing characteristics into cash balances. One way to increase the potential size of liquidity operations is to expand the set of securities that are eligible for repo to the central bank. Not only does this directly contribute to liquidity, but it also does so indirectly since collateral eligibility criteria in the private sector typically consider the assets that the central bank is willing to accept in its operations. For example, since access to liquidity from the Reserve Bank is an important consideration in the liquidity frameworks of Australian CCPs, the range of collateral accepted by the Bank is a determinant of both their collateral eligibility criteria and their decisions around reinvestment of cash collateral. In addition to government securities, the Reserve Bank will accept private securities issued by ADIs, as well as a range of asset-backed and corporate securities that meet minimum criteria for credit quality.
As previously noted, the Reserve Bank has taken a similar stance in responding to the structural shortage of high-quality liquid assets to meet ADIs' requirements under the Basel liquidity standard. In particular, ADIs may establish a committed liquidity facility with the Reserve Bank to help meet these requirements (APRA 2011). Under such a facility, the Reserve Bank would commit, in exchange for a fee, to making available a pre-agreed amount of liquidity under repo. All securities eligible in the Reserve Bank's normal operations will be eligible for the committed facility. It should be noted that, in the case of CCPs, access to liquidity from the Reserve Bank has not been formalised in the same way as for ADIs.
The basic conclusion of a number of recent studies is that there is no globalised shortage of high-quality liquid assets (CGFS 2013). That may well be true today, particularly given the rapid expansion of government debt in most jurisdictions in recent years (Graph 2). However, the focus of most analyses to date has been on outstanding securities issued, rather than the ‘effective’ supply: that is, the outstanding supply of securities on issue, adjusted to reflect how much is actually available to meet collateral needs. This requires an adjustment for how much is locked away in long-term portfolios and not made available for loan, and a further adjustment for the ‘velocity’ of collateral. Collateral velocity refers to how many different purposes are satisfied, on average, by re-using a single line of collateral. Most re-use activity is concentrated in securities lending and repo markets. Standard collateral agreements, such as the Credit Support Annexes that support ISDA documentation, generally include a right of re-use unless this is expressly removed.
Kirk et al, researchers from the Federal Reserve Bank of New York, describe a number of dealer activities that rely on efficient rehypothecation and re-use of collateral (Kirk et al 2014). These include, for instance, intermediation of offsetting repo or derivative transactions between clients. In both cases, considerable efficiencies – and hence cost advantages to the clients – can be gained by passing through the collateral received from one client to the other, rather than relying on own-sourced collateral. In the absence of efficient rehypothecation and re-use, it is likely that the costs of such intermediation – which is integral to market funding and risk management – could rise materially.
There are of course important risk considerations associated with rehypothecation and re-use. These were revealed by the severe deleveraging that occurred following the failure of Lehman Brothers in 2008, and the difficulties experienced by many institutions in recalling securities that had passed through long re-use chains. Accordingly, some investors in high-quality liquid assets have become less inclined to make these assets available in lending programs, and there is increasing caution among collateral providers around rehypothecation and re-use. Central clearing also naturally curtails this activity, since CCPs are not permitted to re-use posted collateral. As a result of such factors, Singh (2013) estimates that collateral velocity in the major financial centres declined by around a third between 2007 and 2012. Kirk et al (2014) present similar evidence.
Looking ahead, some regulatory restrictions on this activity will be introduced, including in the derivative margining regime established by BCBS and IOSCO. And some policymakers favour more wideranging restrictions. While acknowledging the risk considerations, the Reserve Bank has argued strongly in international forums that imposing tighter restrictions on collateral rehypothecation and re-use would be counterproductive. Such restrictions could materially raise the cost of intermediation and increase the risk that the demand for high-quality collateral exceeded supply. After extensive consideration of these issues, the Financial Stability Board (FSB) concluded last year that the focus should instead be on greater transparency of securities lending and repo activity and better disclosure to clients around the extent to which their collateral would be rehypothecated (FSB 2013). The Bank supports the FSB's conclusions.
Recognising emerging constraints on their collateral, financial institutions are increasingly focusing on how to ensure efficient use of collateral. Many institutions are taking steps to improve the flow of information on collateral holdings across business units and geographical locations, in some cases assisted by third-party collateral management services. We will no doubt hear about some of these in the next session. Many such services also provide tools that assist in optimising collateral use, including by facilitating re-use. These typically apply algorithms that identify the collateral that is cheapest to deliver to meet a particular collateral need, given the collateral receiver's eligibility criteria.
We have also heard a lot about collateral transformation, whereby one party exchanges low-quality or illiquid assets with another for high-quality assets that meet some collateral eligibility criteria. There does not yet appear to be very active use of these services, neither domestically nor internationally, but there remains some prospect that these arrangements will become more important and more widely used over time.
The central bank again has a role to play in improving collateral efficiency. At the time ASX was developing its collateral management service, for example, market participants made it clear that, as a major counterparty in the Australian repo market, the Reserve Bank's participation was important. The Reserve Bank has, accordingly, joined the service as a collateral receiver.
In its financial stability role and as overseer of key financial market infrastructure, the Reserve Bank will keep a close eye on the evolution of collateral management services. The Bank will also monitor whether, over time, collateral transformation evolves as a core financial market activity. As these arrangements become more widespread, the Bank will want to be aware of any new interlinkages and dependencies and will want to understand how they may be altering market participants' behaviour.
In conclusion, the way that collateral is used is changing rapidly. Any interruption to collateral markets, or material decline in their efficiency, could have important implications for the cost of a range of intermediary activities and, more broadly, capital allocation and risk transfer in the financial markets.
Central banks and other financial regulators need to ensure that the implementation of regulation in this space strikes the right balance. In this talk, I have cited the importance of the appropriate exercise of regulatory discretion and a considered approach to implementing regulatory reforms – such as OTC derivative reform and Basel liquidity standards. Greater transparency around how collateral markets function is also important. For instance, good data on collateral rehypothecation and re-use could help central banks and other financial regulators to respond effectively if there was some disruption to these markets, or if in adjusting to the new steady state new risks emerged.
Thank you for your attention.
I would like to acknowledge the assistance of Angus Moore and Belinda Cheung in the preparation of this speech, as well as helpful comments and contributions from a number of colleagues. Any remaining errors are my own. [*]
Anderson R and K Jõeveer (2014), ‘The Economics of Collateral’, Financial Markets Group Discussion Paper No 732.
APRA (Australian Prudential Regulation Authority) (2011), ‘Implementing Basel III Liquidity Reforms in Australia’, APRA Discussion Paper, 16 November.
ASIC (Australian Securities and Investments Commission), APRA and RBA (Reserve Bank of Australia) (2014), ‘Report on the Australian OTC Derivatives Market’, April.
BCBS-IOSCO (Basel Committee on Banking Supervision and Board of the International Organization of Securities Commissions) (2013), Margin Requirements for Non-centrally Cleared Derivatives, Bank for International Settlements, Basel.
CGFS (Committee on the Global Financial System) (2013), Asset Encumbrance, Financial Reform and the Demand for Collateral Assets, CGFS Papers No 49, Bank for International Settlements, Basel.
CPSS-IOSCO (Committee on Payment and Settlement Systems and Technical Committee of IOSCO) (2012), Principles for Financial Market Infrastructures, Bank for International Settlements, Basel.
Duffie D, Scheicher M and G Vuillemey (2014), ‘Central Clearing and Collateral Demand’, NBER Working Paper No 19890.
Duffie D and H Zhu (2011), ‘Does a Central Clearing Counterparty Reduce Counterparty Risk?’, The Review of Asset Pricing Studies, 1(1), pp 74–95.
FSB (Financial Stability Board) (2013), ‘Strengthening Oversight and Regulation of Shadow Banking’, August.
Heath A, G Kelly and M Manning (2013), ‘OTC Derivatives Reform: Netting and Networks’, in A Heath, M Lilley and M Manning (eds), Liquidity and Funding Markets, Proceedings of a Conference, Reserve Bank of Australia, Sydney, pp 33–73.
Heath A and MJ Manning (2012), ‘Financial Regulation and Australian Dollar Liquid Assets’, RBA Bulletin, September, pp 43–52.
ISDA (International Swap Dealers Association) (2014), ‘ISDA Margin Survey 2014’, April. Available at <http://www2.isda.org/functional-areas/research/surveys/margin-surveys/>.
Kirk A, J McAndrews, P Sastry and P Weed (2014), ‘Matching Collateral Supply and Financing Demands in Dealer Banks’, Federal Reserve Bank of New York Economic Policy Review, 20(2), pp 1–25.