Ask someone who does not work in securitisation to describe a securitisation and they will probably start with assets being transferred to an SPV and securities being issued to investors. Ask a securitisation lawyer and they are more likely to start with tranching.

That distinction matters.

Significant risk transfer transactions, or SRTs, make the distinction particularly clear. Banks can achieve SRT by using synthetic securitisation to transfer a tranche of the credit risk on a portfolio while retaining the underlying exposures on their balance sheet. The risk transfer may be funded or unfunded. In an unfunded transaction, it may be provided through a guarantee, credit derivative or insurance arrangement.

If the transaction achieves significant risk transfer, the originator may calculate its capital requirements on the retained securitisation positions rather than the underlying exposures.

So far, so familiar. But the growing participation of insurers in unfunded SRT, and the EU’s plans for a new category of “resilient” securitisation, bring another question into sharper focus.

If the assets remain with the bank, no securities are issued and the transaction does not provide funding, why does this belong in securitisation rather than derivatives or insurance?

A securitisation that looks like a derivative

On paper, an unfunded SRT can look much more like a credit derivative than a conventional securitisation.

The bank identifies a portfolio of loans but continues to own them. Credit protection is written over a specified tranche of that portfolio. If losses reach the protected tranche, the protection provider is required to pay. Where the provider is an insurer, the contractual form may look rather different from the capital markets documentation traditionally associated with a securitisation.

No asset sale is required. No funding need be raised. There may be no securities.

Look only at the instrument used to transfer the risk and the conclusion seems obvious. This looks rather more like a derivative or insurance arrangement than a securitisation.

But that overlooks what has to happen before a promise to absorb losses can become an effective SRT.

The reference portfolio must be identified. The credit risk within it must be analysed and divided into tranches. The parties must determine where the protection attaches and detaches, how losses are calculated and allocated, and which risks remain with the originator. They must decide how the structure will operate as the portfolio evolves. The protection must be capable of responding when the relevant losses arise. And the transaction must achieve the regulatory capital outcome for which it was established, not only at closing, but throughout its life.

That is securitisation structuring

The derivative, guarantee or insurance contract is the instrument through which the credit risk is transferred. Securitisation is the architecture around that transfer.

Securitisation is more than funding

We have written before that securitisation is better understood as a technique than as a market in its own right. That matters here. If it is a technique, there is no particular reason why funding must always be its objective.

A traditional securitisation may use that technique to raise funding as well as transfer risk. An SRT uses it differently: the assets stay where they are, while a defined tranche of their credit risk is transferred elsewhere.

In both cases, the essential exercise is familiar: identify the relevant pool of risk, determine how that risk should be divided and allocate it between parties willing and able to bear it.

That does not mean that every credit derivative is a securitisation. The defining feature of a synthetic securitisation is not simply that one party has agreed to protect another against a credit loss. It is the tranching and allocation of the credit risk associated with a portfolio of exposures.

In a traditional securitisation, securities are generally the means by which that allocation is funded and distributed. In an unfunded SRT, a guarantee, credit derivative or insurance arrangement may perform the risk-transfer function. Different instruments are being used, but the structuring question is recognisably the same: what risk is moving, where is it going and does the transaction work as intended?

What makes a risk transfer “resilient”?

This question is particularly relevant following the provisional political agreement reached by the Council of the European Union and the European Parliament on the EU securitisation reforms. The agreement introduces a new “resilient” category, intended to identify well-structured, lower-risk transactions that can benefit from lower capital requirements. It also expressly contemplates the participation of insurers and reinsurers in the synthetic securitisation market. The detailed amendments remain subject to technical finalisation and formal approval.

It would be easy to treat “resilient” as another regulatory label and focus on the capital treatment that comes with it. That may be where the immediate economics lie, but it is not the most interesting question.

The more interesting question is what resilience means when the transaction is unfunded.

In a funded synthetic securitisation, the protection is supported by collateral. In an unfunded transaction, the bank relies on the protection provider performing in the future.

The form of protection therefore changes the risks within the structure. The transaction must address not only losses in the reference portfolio, but whether the provider can meet its obligations, whether the circumstances triggering payment are sufficiently clear and whether payment will be made when it is due. These kinds of counterparty, legal and settlement risks were among the concerns reported as being considered in the EU debate around unfunded protection.

This is where resilience stops being a label and becomes a structuring question.

A risk transfer is not effective merely because the contractual promise exists or because the transaction satisfies a regulatory test at closing. The protection must remain capable of responding when it is actually needed.

Designing for the difficult period, not the closing date

That through-the-life question is easily underestimated.

At closing, the protection provider may be financially strong, the transaction may be economically attractive and the regulatory analysis may be clear. But an SRT is designed to deal with the possibility that the credit environment will deteriorate. That is not an incidental risk to the structure. It is the very risk the transaction has been designed to transfer.

That is precisely when the protection becomes more valuable. It may also be when replacement protection is more difficult, or more expensive, to obtain.

A genuinely resilient structure therefore has to do more than work on day one.

It needs to contemplate what happens if the protection provider’s position changes, whether there are effective mechanisms for dealing with deterioration and whether protection could realistically be replaced if required. Where several providers participate, the structure also needs to allocate responsibility between them and deal with the consequences if their circumstances do not move in lockstep. And throughout, there needs to be sufficient certainty around the losses that qualify for protection and the mechanics and timing of payment.

These are not peripheral issues. They go to whether the credit risk has actually been transferred in substance or merely re-described in the documentation.

They also illustrate why the legal analysis cannot sensibly be separated from the economics of an SRT. A bank may have modelled an attractive capital benefit, but that benefit depends on the structure continuing to meet the applicable requirements and on the protection operating as intended. Attachment and detachment points, amortisation, credit-event definitions, termination rights and loss-settlement mechanics all influence where the risk truly sits.

The purpose of the documents is not simply to record an agreed economic bargain. It is to turn that bargain into a durable allocation of risk.

The resilience paradox

There is, however, a tension at the heart of the debate.

The more safeguards that are attached to unfunded protection, the closer the transaction may come to addressing some of the risks inherent in relying on an unfunded protection provider. The safeguards attributed to a Commission expert opinion illustrate the point: minimum financial-strength criteria for insurers, more than one protection provider, replacement following a sufficiently serious downgrade and greater scrutiny of concentration risk.

But safeguards of that kind also bring cost and complexity.

Replacement provisions are of limited value if there is no suitable provider willing to step in when they are triggered . Requiring more than one protection provider may reduce reliance on a single counterparty, but it also raises practical questions about how the protection is divided and operates between them. And financial-strength criteria may increase confidence in the protection provider, while also limiting the pool of insurers able to participate.

At some point, safeguards intended to make unfunded protection more robust may begin to erode the flexibility and efficiency that made the unfunded structure attractive in the first place.

That does not mean the safeguards are unnecessary. Nor does it mean unfunded protection should receive favourable treatment simply because additional SRT capacity is commercially desirable.

It means that the success of the resilient framework cannot be assessed solely by asking whether its criteria are prudent. Those criteria also have to be capable of being incorporated into transactions that banks and protection providers are willing and able to execute.

A resilient structure that cannot operate in practice does little to further the EU’s objective of freeing bank capital to support increased lending. Equally, a flexible structure that does not hold up when losses arise defeats the point of the exercise. The difficult question is where to draw the line between the two.

Perhaps that is why it belongs in securitisation

An unfunded SRT may look, instrument by instrument, more like a derivative or insurance arrangement than a traditional securitisation. Viewed as a structure rather than as an instrument, the picture changes.

The parties are not simply entering into a bilateral contract under which one agrees to compensate the other for a loss. They are identifying a portfolio of credit exposures, isolating a defined layer of risk and allocating that risk between an originator and one or more protection providers. They must establish how losses move through the transaction, what remains with the originator and whether the transfer will continue to be effective as the portfolio and the parties’ circumstances change.

The synthetic element tells us how the credit risk is transferred. The securitisation analysis tells us which risk is transferred, how it is allocated and what that transfer means for the transaction as a whole.

Perhaps, then, the growth of unfunded SRT does not blur the boundaries of securitisation so much as expose what has always sat at its core. If securitisation is a technique rather than a market, there is no reason to define that technique solely by the issuance of securities or the provision of funding. Securities are one way of transferring and allocating credit risk, but they are not the only one. As Europe’s SRT market develops, the value of securitisation expertise may lie increasingly not in the instrument used to transfer the risk, but in understanding the risk being transferred, where it lands and whether the structure will continue to work when that risk crystallises.

If that is right, the interesting thing about the rise of unfunded SRT is not how far it has moved away from traditional securitisation, but how clearly it reveals what securitisation can do.

Contact