This article analyzes recommendation #1 of this week’s Financial Stability Board publication, “Strengthening Oversight and Regulation of Shadow Banking: A Policy Framework for Addressing Shadow Banking Risks in Securities Lending and Repos.” This recommendation asks for improvements in regulatory reporting. We ask in turn, are governments really prepared to both collect and analyze the data across repo, securities loans and collateral? While it makes sense for governments to know what is going on, we don’t think this recommendation solves the problem. Here’s why.
The FSB’s intention in data collection is to solve two problems:
1) That a failure of one bank could destabilize another (systematic risk)
2) Preventing a liquidity shortage on the collateral side
Both of these are admirable goals, but we think it will take a lot more than governments collecting daily data to really spot problems before they occur. Here are the philosophical arguments against the recommendation:
– Modern risk management practices already look to head of counterparty concentration and liquidity risk problems across a host of asset classes, not just securities lending and repo. Counterparty exposure limits courtesy of Dodd-Frank and the long arm of the CFTC will add more roadblocks.
– On the liquidity side, there is a trust but verify element to the FSB’s proposals. Other FSB recommendations put limits around collateral investments to ensure stability. This recommendation seems like a double check rather than something truly helpful for market liquidity management.
As a practical matter, we don’t think that governments are up to the challenge. We know from primary experience that securities lending and repo datasets are difficult to work with in the best cases; even Markit and SunGard would agree that it takes an experienced eye to sort through and understand what the data are actually saying. While a big top level report will likely catch any outliers that could occur, we don’t think that regulators are going to get the right information at a granular level from the type of data that is being requested. The data are voluminous and can be outright complicated. It takes real time to understand the markets and dig into the data to understand what it is going on. This is time and, so far, detailed market knowledge that the majority of regulators do not possess.
Another argument against this recommendation centers on the collection of collateral data. Assessing what collateral is attached to which securities loans is straightforward; the data vendors have these data points already at least for cash vs. non-cash collateral. It is also a simple matter to ask large custodians that manage cash on behalf of their securities lending clients to report their holdings, as many do already through various US 2a-7 reporting formats. Securities lending agents also have all the data available on non-cash collateral holdings, even though sometimes it takes some prying to get details on what certain codes might represent. However, taking these data as individual points misses out a host of other potentially related factors in financial markets. Securities lending and repo collateral is increasingly intertwined with other financial market activities. Catching a buildup of one collateral type is not going to assist regulators in managing overall market liquidity risk. There needs to be a much more comprehensive understanding of market structure and interconnections in order for the FSB’s proposed monitoring efforts to occur.
There is no question that it is a good idea for governments to know what is going on in securities lending, repo and collateral. We support the idea of governments having regular access to data. However, we think is unrealistic to expect governments to collect large amounts of relevant data, reduce this all to a daily risk report, and come out with real results. We think this recommendation needs some philosophical and mechanical work before being finally adopted. We will publish our revised proposals once we have completed our analysis of the full FSB recommendations.