SRT chronicle: part one

SRT chronicle: part one

Thursday 18 November 2021 10:04 London/ 05.04 New York/ 18.04 Tokyo

In the first of a three-part series on bank risk transfer transactions, Olivier Renault* takes a historical view of the market

The bank balance sheet securitisation market has enjoyed significant growth over the past five years and looks set to continue apace on the back of clearer regulation, an increasing number of banks having acquired the technology and economic incentives for banks to free up risk-weighted assets (RWAs) rather than raising capital. In this series of articles, we will use significant risk transfer (SRT) – which is the EU regulatory terminology – to denote these securitisations, which are also referred to as balance sheet CLOs, bank risk-sharing transactions or ‘reg cap’ trades (SCI 29 October 2021). Our objective is to provide a historical perspective of the development of the SRT market, give a snapshot of the state of the market as of end-2021 and explore its prospects for growth in the next 5-10 years. While the SRT market includes both cash and synthetic securitisations, we will focus on the latter for this series, as it is by far the largest and most diverse one.

A synthetic SRT transaction involves a bank contractually transferring the risk of a tranche of a loan portfolio to an investor while remaining the holder and servicer of the portfolio1. The transfer of risk is associated with a reduction in RWAs for the bank and therefore the transaction provides both risk and capital relief.

Historical perspective
Banks have used synthetic risk transfer to reduce RWAs since the 1990s, so the SRT market is not new. It has, however, evolved considerably over the past 30 years and the market has only existed in its current state (in terms of structures and nature of the risk transferred) since the implementation of Basel 2 from year-end 2007.

Under the previous regulatory regime (Basel 1), most transactions were supersenior hedges, where a counterparty sold protection to a hedging bank on the X%-100% tranche of a reference portfolio. The contract was typically a CDS and, due to restrictive definitions on eligible counterparties for RWA relief, many of the protection sellers ended up being other banks either intermediating the risk or keeping it in their trading books.

The Basel 2 regulatory standards - which are more risk-sensitive than Basel 1 - changed the banks’ incentives completely. Instead of having to hedge the senior risk, the banks got most of their capital benefits by protecting a junior tranche (first-loss or junior mezz) and retaining the senior unhedged with a low risk weight. The same remains true under Basel 3, albeit with a different methodology to calculate senior risk weights.

Figure 1 shows the number of transactions executed each year since 20102. The choice of the start date stems from: the fact that by that date, Basel 2 was implemented in most jurisdictions; and since 2010, transactions have all been real risk transfer trades with no bank-to-bank or short-dated year-end trades, while a few transactions executed during the 2008-2010 financial crisis lacked a genuine risk transfer element - which ended up tainting regulators’ perception of the asset class for many years. 

This chart only includes synthetic SRT transactions placed with private-sector investors; i.e. the many transactions executed by EIF/EIB and a few other public sector bodies are not included in the statistics.

While 2010-2014 saw a relatively small number of banks executing around 15 transactions per year on average, 2015-2020 saw a three-fold increase in the number of transactions and issuing banks. The inflection point in 2015 was to a large extent due to the ECB taking over banking supervision of the most significant financial institutions in the Eurozone from local regulators3. Prior to this date, local regulators in Italy, France and Spain were reluctant or even outright opposed to capital relief through synthetic risk transfer, while the ECB took an approach much closer to BaFin – the German regulator – which had long been comfortable with the technology.

In addition to more harmonised supervision, the clarification of a number of criteria by the EBA - most importantly related to what constitutes ‘significant risk transfer’ - was helpful in giving confidence to banks that they would execute transactions in the spirit of the rules set out by regulatory authorities.

The implementation of Basel 3 over the past decade has also given a lot of impetus to the SRT market, as banks were required to hold much more and higher quality capital. Consequently, freeing up RWAs became more economical for banks than raising capital on the liability side of their balance sheet.

2019 was the record year for the SRT market so far on all counts: highest volume issued (over US$12bn of junior tranches), largest number of transactions (around 60) placed with private investors4 and largest number of bank issuers (25). Several reasons explain this surge in issuance, including a very strong credit market with large demand at tight spreads making it very attractive for banks to issue and a temporary grandfathering of the Basel 2 supervisory formula for transactions executed before end-2019, which allowed banks to free up more capital than under the Basel 3 ‘SEC-IRBA’ formula.

The following year - which was marred by the beginning of the Covid-19 crisis - was a remarkable experience for the SRT market and should help consolidate the reputation of this product as a resilient and dependable capital management tool for banks and their regulators, and as a reliable source of investment opportunities for investors. Although volumes were down versus 2019 - which was an outlier for the reasons explained above - 2020 ended up delivering the second largest tranche volume and number of transactions on record, despite dramatic swings in structured credit markets and high uncertainties about credit fundamentals.

Perhaps counterintuitively issuance volumes in 2020 were actually limited by a lack of bank supply rather than investor demand. As part of Covid capital preservation measures, a number of regulators imposed dividend distribution restrictions to the banks while at the same time encouraging them not to take a too conservative approach to provisioning (in particular, for loans subject to payment moratoria).

Bank capital ratios therefore increased materially in 2020, making the need for capital relief less than in a normal year. Some banks decided therefore to reduce their SRT issuance for that year. In addition, the European Investment Bank group (EIB/EIF) executed a record number of transactions in 2020 (not included in our statistics), such that the gap in total volume between 2019 and 2020 was actually much lower than what is implied in Figure 1.

2021 marks a return to growth for the SRT market and the year is expected to close with issuance volumes second only to 2019 (at least US$11bn) across 50-plus transactions and around 25 issuing banks (see Figure 2), comprising traditionally active banks and a record number of first-time issuers, which bodes well for the continued growth of the market. A characteristic of 2021 is the extension of the Simple, Transparent and Standardised (STS) framework to SRT transactions, which will make it a lot easier for standardised banks to achieve an attractive cost of capital relief (see Box A for a discussion of STS).

In the second part of this series on bank risk transfer transactions, to be published next week, we will provide a snapshot of the state of the market as of end-2021.
 

Box A – The Simple, Transparent and Standardised securitisation framework

In the example above, we show the day-one cost of capital for a standardised bank operating with a core equity Tier 1 ratio (CET1) of 12%, which is hedging a portfolio with 100% risk weight. We assume that the expected loss on the pool is 2% during the lifetime of the transaction and that the bank keeps the 0%-2% tranche as well as the supersenior.

In the first column, the bank considers a non-STS transaction. In order to maximise capital relief, it needs to place the 2%-25.5% tranche, which it splits into a 2%-8% junior mezz and a 8%-25.5% senior mezz respectively, with 10% and 3% spreads. The cost of capital in this scenario is over 12%, which is usually deemed too high for a bank to execute.

If, instead, the bank can execute an STS transaction on the same portfolio (second column), the bank can achieve maximum capital relief by placing a much smaller 2%-15% tranche. This is because the formula used to calculate the risk weight on the retained senior tranche is much more lenient for STS transactions. The bank splits the 2%-15% into a 2%-8% and an 8%-15%, with 10% and 4% spreads respectively, which leads to a circa 9% cost of capital.

In this example, which will be the case for many transactions considered by standardised banks, having or not having STS will determine whether a transaction is economical or not.

Notes

  1. More details of the structures are provided in Box B (see part two).
  2. As transactions are mostly private, our database is probably incomplete. But we are confident that it covers over 90% of the market, including bilateral transactions.
  3. The regulation on the Single Supervisory Mechanism (SSM) entered into force in August 2014.
  4. 55 transactions are reported in Figure 2, but we expect our database to cover around 90% of total transactions.

*Olivier is a veteran of the SRT market and one of the leading originators and structurers on the asset class


×