From the desk of John Arnesen,
Consulting Lead, Pierpoint Financial
In securities finance, collateral is taken as risk mitigation against the possibility that the counterparty does not return the lent security. Not only does the collateral received include a margin, which is typically between two, five or ten per cent higher than the current market value, but it is also marked-to-market daily to maintain that additional buffer. The question I am musing over is why is all of this necessary? What is the real counterparty risk an institution faces when lending securities and what additional revenue could be generated by eliminating collateral all together?
After all, it's been 12 years since the default of Lehman Brothers and that event was driven by a short-term liquidity crisis, and the regulation in subsequent years has been nothing if not obsessed with forcing banks to hold sufficient capital buffers, curb certain riskier activities and to strengthen banks against future systemic shocks.While the pandemic will have a detrimental effect on loan-loss provisions, the temporary relaxation of specific rules will allow banks to weather the storm in the short term, and the strong-arming by regulators over the suspension of bank dividend payments will add to capital reserves.
According to Global Finance Magazine, the 'world's safest top ten banks have a noticeable absence of participants most active in securities finance, and US banks are absent altogether. However, ZKB is an active participant, so why demand collateral from them?
Back in the 1980s and 1990s, the industry was dominated by US broker/dealers. Still, those institutions, usually operating as a subsidiary of an investment bank, look today nothing like they did back then from a risk perspective. The requirement to include a margin of 105% of market value for lent securities may have made a lot of sense back then. However, the same margin requirement is still applied today, regardless of the creditworthiness of the borrower. Why should ZKB, for example, post a 5% margin to an agent lender on behalf of a beneficial owner that is unrated?
I know the counter-argument: it's the way business is conducted, beneficial owners will be reluctant to lend without it, and it would affect the cost of capital for the agent lender in providing indemnification and blow up their credit lines. But are these objections worth more debate?
In the scenario described, would it not be sound and reasonable for ZKB to ask for margin from the beneficial owner? The question I'd like an answer to is how much would a non-collateralised, non-margined arrangement change the fee or rebate that an AAA-rated entity would be prepared to pay? Absent the opportunity to ask one directly about collateral and margining, I asked other participants. Here is what they said.
Naturally, to borrow securities unsecured would be a massive advantage from a risk perspective. The elimination of collateral costs, a reduction in Risk-Weighted-Asset (RWA) consumption and reduced use of VaR-based credit lines are all areas of benefit. However, ironically, this ideal scenario may create problems in itself. There is a reasonable chance that internal systems would not 'understand' the transaction and not be calibrated to operate this way, and it would implode under the confusion. Then, more realistically, a prudent borrower is likely to limit the amount of business it is prepared to do unsecured as it may not last. If the arrangement changed and there was suddenly the need to collateralise billions of loans it would create as many problems as it solved for at the outset. In other words, the risk reduction it offers the borrower establishes a future risk by the nature of the dependency on uncollateralised transactions.
The other telling issue is that fees are likely to only increase by the cost of sourcing collateral or the funding received by financing long positions. For example, the most expensive collateral to post is probably HQLA. If the borrower can save that cost (let's be generous and call it 25 basis points), then the fee or rebate of the loan could reflect part or all of that cost. That pick up, assuming it all is embedded in the fee will not be enough for beneficial owners and agent lenders that will need massive credit lines (at the notional level) to turn this into a reality. It must be recognised, however, that I'm describing a relationship between lender and borrower using creditworthiness as a reason to change the structure of the trade. There is a more prominent element that can't be ignored in that the borrower needs to borrow the securities, whereas the lender has the option to do nothing at all.
Given that position, the lender has more ability to set the terms of the transaction, particularly for assets with high scarcity value. Perhaps lenders could differentiate between borrowers' credit ratings by a relaxation in the acceptability of collateral? The higher-rated counterparties might be able to post less liquid or lower-rated collateral.
In the meantime, I can put that fantasy to one side for the time being and address the point I made about an unrated counterparty. I suggested it was reasonable for an AAA-rated entity to ask for margin from an unrated lender, knowing full well this is also a stretch. However, is it not feasible for a higher-rated counterparty to post a reduced margin to a lower-rated or unrated counterparty?
A triple-A rated institution is always going to be in a position where the majority of its counterparties are lower-rated than them. That has a cost to it and is reflected in the fee it will pay or, at least, it should. There will come a day, if it has not already happened, whereby a borrower, at the individual trader level, will need to know the legal counterparty (beneficial owner) and collateral it needs to post at the time of trading.
The type of collateral, be that cash or securities, the counterparty and its jurisdiction and the margin requirement all form part of a prescriptive calculation that determines the 'cost' of the transaction. That cost could range from 2% to 100% of the nominal amount of the loan, and these factors have a bearing on the price or fee the borrower is prepared to pay. Extrapolate this further, and you could have a snapshot of the same lent security with a range of loan fees due to these factors and every transaction perfectly reasonable and defensible from agent lenders' standpoint. This is one reason why using aggregated, vendor-based data to determine a fair value fee must be viewed with caution, or combined with other metrics, especially by regulators.
So is this happening in reality? I think it largely depends on to what extent the consumption of RWA is applied to an individual business line or even, to the individual trader or trade. There is plenty of evidence that for many firms, this is indeed the case. As that extends, all participants, not just borrowers, are going to apply approaches to securities finance to reduce their costs.
For agent lenders, cash collateral, which lost its appeal in Europe after the global credit crisis, shouldn't be discounted entirely. Reinvestment in money market funds is a relatively low risk and very liquid option with same-day subscription and redemption capabilities. It may suit UCITS funds, restricted by the inability to lend on a term basis and clients that are so restrictive in their collateral profiles (e.g. non-cash government debt) that it offers an alternative that yields higher than reverse repo in the same instruments. Cash collateral should also reduce the costs associated with indemnification given the absence of non-cash in the form of index equities and corporate debt that attract higher charges. Pledge, as an alternative to title transfer collateral, is a material benefit to borrowers, and it is the forward-thinking agent lenders that will adopt this sooner rather than later that will reap the benefits.
So collateralised transactions with haircuts and margin are here to stay. The other variables associated with a securities lending transaction are, on the other hand, going to change. The value you extract from securities lending will depend on how quickly these changes are embraced.