The first details of the Leverage Ratio were released by the Basel Committee on Banking Supervision in June 2013. The big shock was the exclusion of any form of netting.

In January 2014, however, Basel appeared to have relented and issued revisions to the Leverage Ratio that allow limited netting. These changes have been broadly welcomed. However, they may not be as generous as many commentators are assuming.

The one definitely helpful change to the Leverage Ratio rules is the decision to recognise the ‘final contractual exposure’ of a bank to a ‘qualifying’ central clearing counterparty (CCP) as the measure of gross SFT assets to be included in the exposure component of its Leverage Ratio. This is eminently sensible. Forcing banks to unravel a net exposure to a CCP into its original gross components would discredit Basel’s own policy of promoting greater use of CCPs. As over 70% of repos in the European market are cleared across CCPs and average netting rates are about 75%, the recognition of CCP netting is very significant. It is incredible that it was not recognised in the original draft of the Leverage Ratio rules.

Of course, once Basel allowed netting by CCPs, it could not maintain a blanket exclusion of legally-enforceable netting under bilateral agreements, such as the ICMA’s Global Master Repurchase Agreement (GMRA). But what Basel appears to have given with one hand, it may well have taken back with the other.

Basel has recognised bilateral netting for repos against reverse repos between the same counterparties but only of the cash payables against the cash receivables. This would have an important impact on Leverage Ratio exposures. Consider the following seller and buyer.

Seller of 10 in collateral v 10 in cash (repo)

assets		liabilities
collateral	10	liabilities	20
cash	10
LR   exposure	20

Buyer of 10 in collateral v 10 in cash (reverse repo)

assets		liabilities
receivable	10	liabilities	10
LR exposure	10

Now, consider the Leverage Ratio exposure of a party that does both of the above repos and reverse repos.

Repo of 10 in collateral v 10 in cash & identical reverse repo

assets		liabilities
collateral	10	liabilities	30
cash	10
receivable	10
LR exposure	30

If the repo and reverse repo were with the same counterparty, cashflow netting would allow the following very significant reduction of its Leverage Ratio exposure.

 

assets		liabilities
collateral	10	liabilities	10
LR exposure	10

The question is, why did Basel not also allow the netting of identical collateral securities between the same counterparties? The logic is the same as for netting cash. Netting opposite flows of the same collateral would entirely eliminate the Leverage Ratio exposure of matched repos and reverse repos. Perhaps, this was all too much for Basel?

However, not only did Basel not allow collateral netting but the netting of cash payables and receivables is subject to conditions:

1          The transactions must have the same explicit final settlement date.

2          The right to net cash payables and receivables must be legally enforceable, not just in the event of default, insolvency or bankruptcy, but also in the ‘normal course of business’, ie at any time, rather than being contingent upon a future credit or other event.

3          The counterparties must intend to settle net, settle simultaneously or the transactions must be subject to a settlement mechanism that results in the ‘functional equivalent’ of net settlement. Functional equivalence is defined as the cashflows being equivalent, in effect, as a single net amount on the settlement date. This requires offsetting transactions to be settled through the same settlement system and the settlement arrangements must be supported by cash and/or intraday credit facilities intended to ensure that settlement of all the transactions will occur by the end of the business day and that the linkages to collateral settlement do not result in the unwinding of the net cash settlement, ie fails must not stop payment.

The first and second conditions should be not obstacles to netting. The first condition is a standard requirement for netting and the second condition is satisfied, for those parties using the GMRA, by paragraph 6 (Payment and Transfer).[1] The issue is the third condition.

The intention to settle net that is required by the third condition can be satisfied by instructing settlement in the same batch-processing cycle. It can probably also be satisfied by the practice of ‘pairing-off’ matching trades with the same counterparty (agreed bilateral cancellation before settlement). Using a CCP might count as well, although not all CCPs may qualify. However, settling across the real-time gross settlement (RTGS) facility of a securities settlement system (SSS) is more problematic.

The problem is that the vast bulk of modern securities settlement is delivery-versus-payment (DvP), in other words, the simultaneous and mutually-conditional exchange of cash and securities. In other words: no security, no cash. Basel must be aware of that. Is this condition just a spoiling tactic? Or is it just confused thinking?

The fact is that, even in RTGS, most securities transactions are not settled one by one on a truly gross basis. In practice, cash payments for securities settlement tend to be made in batches, on a net basis. And in the event of a lack of cash, the securities settlement systems (SSS) at the International Central Securities Depositories (ICSD) and many domestic Central Securities Depositories (CSDs) are supported by facilities that can provide credit secured against holdings of other securities or auto-collateralisation mechanisms. So DvP is not necessarily a problem. Credit facilities and auto-collateralisation mechanisms would appear to be the sort of safeguards that Basel wants. They would certainly prevent an inadvertent shortage of cash from stopping settlement.

But is a cash shortage what Basel is really worried about? They seem to want cash to flow despite failures to deliver securities. Cash or credit facilities would not do anything to stop fails. What is actually needed is auto-borrowing facilities of the sort provided by the ICSDs and some CSDs. It would help if Basel would elaborate. Unfortunately, that is something that they have not been inclined to do to date.

The big losers from Basel’s restrictive rules on netting are those banks who trade frequently with the same counterparties. This means the largest banks. However, it is not clear how large such back-and-forth repo business is, even for the largest banks.

Where did the Basel netting conditions come from? They resemble the netting rules of the International Financial Reporting Standards (IFRS), specifically, International Accounting Standard (IAS) 31. However, the third Basel condition has a crucial difference. In 2011, the International Accounting Standards Board (IASB) clarified what they meant in IAS 31 by an ‘intent’ either to settle net or simultaneously. The IASB said that, although an entity may still realize the asset and settle a liability separately, if it can settle amounts in a manner such that the outcome is, in effect, equivalent to net settlement, then it would meet the IFRS net settlement criterion.

Equivalence to net settlement in IAS 31 will occur if, and only if, the gross settlement mechanism has features that eliminate or result in insignificant credit and liquidity risk, and will process receivables and payables in a single settlement process or cycle. For example, a gross settlement system that has all of the following characteristics would meet the net settlement criterion:

1          Financial assets and financial liabilities eligible for set-off are submitted at the same point in time for processing;

2          Once the financial assets and financial liabilities are submitted for processing, the parties are committed to fulfil the settlement obligation;

3          There is no potential for the cashflows arising from the assets and liabilities to change once they have been submitted for processing (unless the processing fails—see below);

4          Assets and liabilities that are collateralized with securities will be settled on a securities transfer or similar system (for example, delivery versus payment), so that if the transfer of securities fails, the processing of the related receivable or payable for which the securities are collateral will also fail (and vice versa);

5          Any transactions that fail, as outlined above, will be re-entered for processing until they are settled;

6          Settlement is carried out through the same settlement institution (for example, a settlement bank, a central bank or a central securities depository); and

7          An intraday credit facility is in place that will provide sufficient overdraft amounts to enable the processing of payments at the settlement date for each of the parties, and it is virtually certain that the intraday credit facility will be honoured if called upon.

Note the fourth characteristic: ‘if the transfer of securities fails, the processing of the related receivable or payable for which the securities are collateral will also fail’. This is strict DvP and the opposite of the Basel condition that ‘linkages to collateral settlement do not result in the unwinding of the net cash settlement’. In other words, IFRS want DvP but Basel wants payment, even in the event of a failure to deliver securities. This is not going to happen, as it would re-introduce settlement risk. No one (least of all the other Basel regulators down the corridor at the CPSS) is going to allow a regression away from DvP.[2]

But would the IFRS rules actually make netting easier for Leverage Ratio exposures? Probably not. In practice, IAS 31 is notoriously restrictive about netting, especially compared with the US GAAP rules on netting in the US Financial Accounting Standards Board (FASB) FIN 41. So European banks may not have much to lose. But US banks do.

 

---