Averting a Liquidity Crisis: The Road to More Resilient REPO Markets
© 2016 Fotolia
Loriano Mancini, Professor at the SFI@EPFL and his co-authors published a new paper in one of the most prestigious academic finance journals, which provides novel insights on the last financial crisis. Their findings were presented at the New York FED and at the European Central Bank, and have implications for revising financial regulation.
Problems in the US repo market were a catalyst for the Great Recession. Interbank lending faltered and many banks failed or were bailed out as the economic impact reverberated around the world. In the Euro countries, though, repo markets fared much better. Now a new study improves our understanding of these markets, reveals important differences between European and US repo practices, and highlights measures that can promote more efficient repo operations and help avoid a similar crisis in the future.
Repurchase agreements and repo markets are central to providing day-to-day funding for banks and other financial firms. If these trillion dollar markets fail to work efficiently, as the world discovered in 2008, credit dries up and this can quickly lead to systemic banking problems and a financial crisis. Yet despite their importance, many people in business, even in financial services, know very little about these trillion dollar markets.
Using data that has only been available relatively recently Loriano Mancini of the Swiss Finance Institute and EPFL, and his fellow academics Angelo Ranaldo and Jan Wrampelmeyer at the University of St.Gallen, have investigated the way that repo markets work. Their findings provide important insights into the elements necessary to create stable and efficient repo markets and avoid a similar credit crisis in the future.
Interbank lending
Repo is short for repurchase agreement. These are transactions where a "lender" commits to purchasing assets from a "borrower" at an agreed price, while the "borrower" agrees to repurchase those assets at a later date, usually at a premium. Transaction terms and structure vary, but the intention of repos is to facilitate a temporary transfer of funds between banks, secured against assets.
Banks often need to borrow money on a short term basis to fund their activities. They do this through both unsecured and secured lending. Unsecured lending is heavily dependent on trust. Consequently, at times of stress and turmoil in financial markets trust between financial firms, including banks, can be undermined and unsecured lending may disappear. At times like this the repo market should keep interbank lending alive and the banking system functioning. In 2008, however, the US repo market failed to compensate for a collapse in unsecured lending. Credit terms tightened, interbank funding became inaccessible. Some banks went bust, while others were bailed out by the state.
A lack of data made it difficult to understand what went wrong and why in the immediate aftermath of the credit crunch. Sensibly, the authorities began to require banks and financial institutions to report repo trading data such as lending volumes and activities. Mancini and his colleagues have used the increased availability of data, and their connections with European repo trading platform Eurex Repo, to study the European repo markets.
Their findings reveal important differences between European and US repo markets including: the role of a central counterparty clearing house (CCP); anonymous trading; the mechanism for dealing with default; the nature of the assets used for securities; and the phenomenon of "unwinding". These differences may account for the different fortunes of the two markets during the crisis. In the European repo market trading volumes increased in the crisis, while rates remained relatively stable. Far from grinding to a halt, the European repo market appeared to act as a shock absorber, providing a reliable source of liquidity at times of greater market turbulence, volatility and risk.
In Europe some 60% of repo transactions happen via a central counterparty clearing house (CCP), with some 30% occurring bilaterally and the balance via a tri-party arrangement (the main method in the US). The CCP governs the market, matching banks on both sides of the transaction, and facilitating anonymous trading; banks only see the CCP, not each other. This helps eliminate counterparty risk, as the banks need only trust the CCP. In the US, although the majority of repos are facilitated by a third part agent, banks have visibility of both the counterparty bank as well as the tri-party agent. Consequently, any systemic or specific mistrust in the general financial position of banks can contaminate the repo market.
If a borrower defaults the CCP has a series of defences to protect the lender and the market. The CCP may: close down the borrower's positions using any available funds to repay the lender; liquidate the collateral and repay the repo loan from the proceeds; use the borrower's clearing fund contributions to pay the lender (banks admitted to the market must contribute towards a clearing fund); if further funds are required have recourse to the clearing fund more generally, drawing on the contributions of all market participants.
In the US repo market no similar series of defences exists. If a borrower defaults the lender has the often difficult task of valuing and selling the collateral. If it quickly dumps the collateral on the market that will negatively impact the price of those assets, adversely affecting other banks and investors.
A second difference is the type of asset used as collateral. In Europe, the CCP provides reassurance to the market by specifying a list of acceptable assets. If an asset's risk profile changes the list is updated. In the US, increasingly risky asset backed securities were used as collateral in the bilateral repo market. As compensation for the risk, lenders required loans to be over-collateralised. As the risk increases, the difference - the haircut - between asset value and loan amount obtainable increases, making it more difficult to obtain adequate liquidity.
Unwinding
Another key difference between European and US repo markets is the practice of unwinding. Take a typical US tri-party repo contract. The lender advances the money, the borrower puts up collateral, both via a tri-party agent. But regardless of duration, the position is unwound at the beginning of each day. The collateral is returned to the borrower. The tri-party agent returns the loan amount to the lender, but the tri-party agent does not recoup the loan amount from the borrower until later in the day when it rewinds the position. Until then the tri-party agent is advancing large amounts of intraday credit.
Historically the reason for this arrangement was to provide flexibility to both borrower and lender. However, in practice it created a potential vulnerability in the banking system. As only a few tri-party agents are involved in the US repo markets – two, the Bank of New York Mellon and JP Morgan as of the end of 2015 - the effect was to allow a small number of firms to shoulder the risk of funding a large part of the shadow banking system, extending trillions of dollars of intraday credit spread across numerous banks. In Europe, although there is no comparable unwind and rewind process, there is still flexibility. A borrower trading via a CCP can provide any asset as collateral from a predefined basket of assets, it can also substitute that asset for another in the basket if it wishes.
The research by Mancini and his colleagues highlights the mechanisms which make these vitally important repo markets resilient in the face of enormous market stresses. In particular it emphasises the importance of having a CCP and anonymous central clearing, using safe collateral, and avoiding the daily unwinding and rewinding of repo contracts, other than when the contract reaches maturity. Given these conditions, an interbank repo market can act as buffer, a shock absorber in times of market stress and threats to market liquidity.
These findings should inform ongoing efforts to make repo markets more resilient. This is true in the US, for example, where current reforms leave scope for a fire sale of collateral, an unwind mechanism remains albeit temporarily (as implementation of reforms has taken longer than hoped for), and the availability and use of collateral could be further improved. If appropriate steps are not taken in the US and elsewhere, a crisis precipitating credit crunch originating in the repo markets may not be far away.