By David Simpson
•
November 7, 2022
Following last month’s so called 'mini' budget and the consequent dramatic sell-off in Gilts and the eventual intervention of the Bank of England, the industry has been discussing the way Pension Funds (PFs) manage their assets and liabilities. Particularly, derivative contracts and the underlying margin & collateral management i.e. the credit support annexes (CSAs). We would like to explore the background and how increasing data transparency might help ease some of the pinch points and save money in the process. For initially sound reasons, following the GFC, ISDA in conjunction with the Regulators and Banks changed from so called ‘Dirty CSAs’ to ‘Gold Standard’ or ‘Clean’ CSAs. Under Dirty CSAs, PFs and Insurers could post a wide range of collateral. However, for several reasons, including the demise of Libor and the interrelationship of the value of the derivative and the collateral received under the CSA, the industry moved to Clean CSAs. Originally the plan was to include government bonds/gilts within the types of ‘eligible collateral’ listed under Clean CSAs. However, bonds do not net with Bank derivative exposures for leverage balance sheet (LBS) purposes. Use of such scarce capacity effectively increases bank opportunity costs and capital charges and so the banks insisted on cash as collateral. In addition, banks were penalised heavily for valuation disputes in the value of the derivative/collateral terms by draconian capital provisions. Thus, the Gold Standard CSA for banks became to accept only cash as collateral. When PFs sought to hedge their gilt and interest rate exposures with derivatives, they were unable to post non-cash assets (the gilts) as collateral. In theory, the PFs could have posted assets under repo facilities in exchange for cash but often these were restricted to cash and gilts as well. (PFs that had repo facilities could raise cash by posting their gilts for cash, but as gilt prices became volatile, the amount of gilts needed to support the cash borrow increased. In addition, for technical and systems reasons, it was difficult to finance the index linked gilts on repo. Raising the gilt valuation issue again). Ultimately many Funds were forced to sell gilts and to raise cash as their repo lines proved inadequate. This was a liquidity rather than a solvency problem, as PFs, ultimately, could reduce the pay out to stakeholders or -ask the sponsor corporate for increased contributions. Nevertheless, it was a systemic liquidity problem and a significant funding crisis emerged in a gilt market thought almost immune from such runs. We understand from banks that, in some cases, margin calls were made and not met, primarily because of settlement fails in the market as it was difficult for Funds to sell the gilts as liquidity drained from the market. Ironically, had Pension Funds been allowed to post liquid assets such as government bonds as margin, the sell-off may not have been as extreme, meaning the downward spiral that ensued, would have been reduced. From the PF counterparty Bank’s perspective surely one lesson to be learned was it would be better to receive some type of collateral than non at all? This highlights the issue of procyclicality, the times when additional margin is required is when liquidity is scarcest. As in previous crises, the need to rapidly sell assets in order to provide liquidity to maintain margin requirements leads to a drop in value of the assets being liquidated. In addition, the manual nature of T+2 settlement, with statements only calculated at end of day (often NY time) and delivered at some point the next morning, the valuation, reconciliation and settlement issues this creates (c.f. the hold up in processing margin calls at some US custody banks, apparently due to settlement areas being overwhelmed with trade volumes) with the market trying to determine whether gilts had actually been delivered, at what price and in sorting the inevitable fails in a rush for the exits, exacerbated the situation. If the Pension Funds and their counterparties were able to view and agree on near real-time pricing of these assets and liabilities, we believe the process would have been simpler. Counterparties who could replicate their own margin situation rather than waiting for their counterparty or custodian’s calculations can be far more proactive in ensuring the correct margin is posted, reducing fails elsewhere in the system. With financial technology, such as Siman Systems Margin & Cost Optimisation Module, businesses can now collate, monitor & process the real-time pricing of their assets and liabilities and can predict (and stress test) the margin requirements. Such transparency is essential if the requests for a broader range of ‘eligibility criteria’ for liquid non-cash assets are to be posted as collateral once again. Of course, of itself, this does not solve the banks LBS issues. However, we would suggest the following three practical lessons could be learned from the impacts following the notorious 'mini' budget. (In addition to improving resilience, the algo will add transparency saving the Pension Fund money). (1) Consider adding government bond collateralization provisions in CSAs (with perhaps a multiplier for bonds, to encourage cash provision in ‘normal’ times, would help reduce the volatility and volume of market sales and purchases in extremis. (2) Improving PF transparency and reducing two day data lags by using an off the shelf system, such as Siman Systems Margin Module, will enabling real time valuations of assets, liability and margin and would reduce settlement disputes and ease correspondence between Banks and Counterparties as both parties can compare consistent data. (3) Adding contingent repo lines or agent lending facilities -allowing participants to exchange asset types more efficiently reducing pinch points.