In the decade since the collapse of Lehman Brothers, regulators around the world have taken steps which, they argue, have greatly strengthened the resilience of the financial system. Buoyant asset prices and rising bank shares suggest that investors largely believe them. Unfortunately, the effectiveness of these measures remains uncertain. That will only become clear in a real downturn, rather than a simulated stress test.
The focus on raising collateral requirements exemplifies the problem.
Collateral has long been used to secure borrowing. In derivative transactions, collateral is lodged to secure the current amount owed by a party. The method is used bilaterally between banks and between banks and their clients. It’s also used on a multilateral basis; for example, a significant proportion of derivative transactions are now routed through so-called Central Counter Parties, which use collateralisation to secure performance.
Since 2008, regulators have placed increasing emphasis on collateralisation to reduce risk, for instance by allowing banks to hold less regulatory capital when engaging in properly collateralised transactions. In a crisis, however, a host of factors will affect whether collateral will work as intended.
Originally limited to cash and government securities (though even these are no longer entirely risk-free), the range of securities accepted as collateral has expanded to riskier securities. In part, this reflects a shortage of high-quality securities, especially government bonds, because of central bank purchases. Some of these types of collateral have volatile prices, which reduces their utility as security.
While this is addressed by attributing lower value to them (known as a haircut), such adjustments may turn out to be incorrect. The correlation between the risk sought to be covered and the value of the collateral is critical. If the underlying risk increases at the same time as the value of the collateral decreases, the benefit of holding that collateral obviously shrinks.
Collateralisation also relies on models which aren’t foolproof. The underlying exposure as well as the value of the collateral must both be determined in a process that isn’t always scientific. As the world discovered in 2008, it’s hard to value less-liquid securities and derivatives, and there’s a not-negligible risk of manipulation of valuations. Models that use historical volatility to gauge the level of initial collateral required may prove inadequate in periods of stress. This is even truer in the case of portfolios involving a number of transactions, where past correlations used to calculate the overall exposure may not hold in future.
Relying so much on collateral assumes it can be sold off easily in case of default. Declines in trading volumes for many assets mean that creditors can’t take such levels of liquidity for granted.
There are operational and legal risks as well. The process assumes mark-to-market calculations are accurate, collateral is paid and received on time, collateral is monitored and control over the cash or securities is always maintained. Given the high volume of transactions across multiple time zones, operational accuracy and integrity is difficult to ensure. The legal validity of these arrangements, which involve a complex mix of domestic and international laws, isn’t assured. Cross-border enforcement may be difficult, with national courts failing to enforce foreign judgements.
Finally, collateralisation may alter the behaviour of market participants, creating new systemic risks.
The use of collateral shifts the emphasis away from the creditworthiness of a borrower or counterparty. Parties who would normally be ineligible to borrow or trade, if judged purely on their financial strength, might be able to enter into transactions if they can produce enough collateral. The rapid growth in debt levels, volume of derivative contracts and investment vehicles since 2008 has relied greatly on the fact that collateral has expanded the range of market participants.
Banks have an incentive to lower collateral levels in order to increase transaction volumes. This increases leverage but also risk. Banks have also used collateral to increase liquidity and leverage directly. So-called rehypothecation — where collateral received by a bank is then pledged to support other transactions — allows for an exponential expansion in leverage. Restricting such practices, on the other hand, would precipitate a rapid contraction in liquidity.
Collateral creates channels through which instability can be transmitted in a crisis, increasing the risk of contagion. In periods of stress, market participants will all seek more collateral or need to sell pledged securities. This will increase volatility, potentially fatally. At the same time, limited disclosure of collateral provisions and potential liquidity claims makes it difficult to assess the financial position of counterparties.
The goal of collateral — to reduce the risk of a financial crisis — is noble. In practice, collateral simply creates different risks. What regulators really need to do is address more fundamental problems: an overreliance on credit and leverage, an excessively large banking system, and unnecessary complexity and interconnectedness in the financial system itself. Ten years after the last crisis, those issues remain as unresolved as ever.