The topic of netting for repo, securities lending and prime brokerage has been alternately praised and attacked in financial markets. Proponents argue that it strengthens risk management by encouraging safe, balancing transactions by banks and broker-dealers. Critics say that netting hides risk and can be manipulated to suit the needs of the institution. Which arguments have merit and which are based on heresay and rumor?
A new Finadium research report, “Netting Rules for Repo, Securities Lending and Prime Brokerage“, investigates current netting practices and directions for their future evolution. Netting has been and will remain an important mechanism for freeing up balance sheet in securities finance transactions. However, its value is changing as Basel III and domestic rules shorten the time period of bank observations. Previously, banks would work to net their securities finance balance sheets once a quarter, a process that took weeks to arrange. Going forward, it is possible that netting will need to be done daily to meet daily average observation periods of stress tests. This would mean a different conception altogether of the importance and role of netting in financial markets.
Netting matters less when reporting is more frequent. When a repo book must report balance sheet absorption quarterly and the effort to reconstitute takes a week or two of concentrated effort to compress the book, including interruptions to client flow, then the trade off is considered a normal part of business. But if reporting is more frequent and the disruption caused by going through the netting exercise is nearly constant, then the cost/benefit may swing the other way. The negative optics of showing a rapidly rising and falling balance sheet to investors, regulators, or other stake holders, especially when the only difference might be whether books were actively netted for accounting purposes or not, should not be underestimated. One repo dealer reported to us that their own internal controls have been tightened to prevent massive swings in balance sheet use due to the application of netting. We suspect this has (or will) become common across other repo dealers too. Is this an implicit recognition that repo netting may not pass a “smell test”?
International Financial Reporting Standards (IFRS) differ from US GAAP rules in a number of areas. However, when it comes to repo netting, IAS 32 and Fin 39/41 are fundamentally parallel. One difference between GAAP and IFRS centers on intent. IFRS requires the intent to settle exposures on a net basis where GAAP is silent. Intent may be part of an overall business model but can be hard to prove on a more granular trade-by-trade level. Meanwhile, IAS 32 requires simultaneous net settlement of trades in order to qualify for netting. As with many accounting rules that need to be converted into actual practice, institutions interpret how to apply the directives differently.
Even in the post-Basel III world, netting still matters: our research finds that repo dealers believe that netting can compress High Quality Liquid Asset (HQLA) bilateral trading books by 60% to 80%. Non-HQLA books including corporate bonds, equities, or other less liquid assets do not compress as easily. Even so, the ability and intelligent application of netting rules means a major difference in securities finance capacity and profitability.
This report is valued reading for practitioners in securities finance, risk managers, Treasury managers and their advisors. More information and the table of contents can be found on the Finadium website.