SRTs are pricier but more effective

The capital adequacy treatment of SRTs is also harder to pin down with precision, unlike that for SCLOs, which were completed pursuant to Basel 1 or Basel 2, in which rules regarding capital relief were relatively straightforward and quite uniform from country to country. For modern SRTs, differences in capital treatment can emanate from whether the originating bank is a standardised approach bank or an IRB bank; whether the bank is US-based, UK-based or Eurozone-based; whether the transaction is structured to meet the so-called STS criteria; whether the risk that is transferred to third parties is transferred by means of an unfunded CDS contract, a funded or margined CDS contract, or a credit-linked note; and from a number of additional factors.

The failure to transfer the first loss tranche in the past resulted primarily from the virtual absence, 25 years ago, of credit investors which could analyse and assume the high risk/high return of this tranche. Investors were also reluctant to underwrite the risk of a portfolio if the originator no longer had ‘skin in the game’. They felt that, if all the portfolio risk was assumed by third parties, the originating bank would no longer monitor and pursue its borrowers with the care and diligence that would be expected if the bank was still exposed to the underlying loans.

Finally, given the extreme difficulty of predicting the likelihood and extent of losses on the first loss tranche (which, recall, is exposed to the first several portfolio loans that default), potential investors who even considered investing in this tranche requested yields that were simply uneconomical for the originator.

The private credit revolution of the past few years has turned this situation on its head. Not only are there now meaningful numbers of institutional investors willing to assume first loss risk, but they probably outnumber the deals that offer such a tranche, causing spreads to compress significantly in recent times. That said, for portfolios of high-risk exposures, such as leveraged loans or in some cases SME loans, the investor yield can often reach low double digits, even today, and this is without counting the SOFR or Euribor component that supplements the yield when the risk transfer takes the form of a credit-linked note.

And what of the senior tranche? As mentioned above, it was common for originators in the days of SCLOs to purchase protection on this tranche, at a relatively low price in the single digits, via a senior CDS contract, from a highly-rated independent bank. This would reduce the tranche’s risk weight from 100% to 20% — and was done despite the fact that this tranche could be shown to be of triple A quality on a stand-alone basis given its significant subordination (logically it should have qualified for a largely reduced risk weight on this basis alone).

In the case of the SCLO transaction analysed above, we saw that a total subordination of almost 8% was needed to generate a triple A rating on the senior tranche, even though the portfolio in question was of investment grade quality. So the thinking went as follows: if banks’ investment grade portfolios are incurring loss rates of 8%, the rest of their portfolios are presumably incurring even steeper losses, possibly taking the bank’s total credit losses up to the mid-teens. But banks’ capital is usually only around 10% of their assets, so they would be wiped out under such a scenario.

This meant that just when the originating bank needed its counterparty most desperately to cover losses above the 8% threshold, the counterparty would have exhausted its own capital and would be unable to honour the senior CDS. In the financial crisis, this proved true of many banks, as well as monoline insurance companies such as MBIA, and multiline insurers such as AIG. In short, the originating banks have now realised that the premium paid for senior tranche protection, although modest, was wasted. It was paid in return for support which would most probably not be available when needed.

If this is correct — and assuming the originating bank is required today to maintain a 12% minimum capital base under recent versions of Basel capital regulation — the capital requirement for the retained layer would shrink to:

This effectively delivers a 10% reduction in capital requirements from the original 12% applicable to the leveraged loan portfolio prior to any risk transfer.

The all-in annual cost of the transaction for German Bank is:

This compares to an annual cost of equity capital of:

(Remember that this 12% cost of equity capital is calculated pursuant to the capital asset pricing model as shown above.)

So prima facie, and unlike the SCLO, the cost of achieving capital relief through the SRT structure in this case is actually marginally higher than the cost of simply raising fresh capital in the equity capital markets. What might motivate German Bank to tolerate this extra cost?

rather than the 12% previously assumed, (ii) the annual cost of equity capital would now reach

This is higher than the 1.62% cost of achieving the capital relief using SRT.

In addition, the cost of the SRT protection is probably tax-deductible for the originator under German tax law, while the cost of raising fresh equity capital would not be.

Assuming a 40% tax rate for the originating bank, this reduces its after-tax cost of achieving capital relief to

At this point the advantages of the SRT became compelling.

Finally, if German Bank had merely issued fresh equity into the capital markets it would not have achieved any reduction in its risk exposure to the leveraged loan portfolio, whereas by using the SRT it has rid itself of all losses from the portfolio up to 18%. This is an additional feather in the SRT’s cap.