Cash flow models for non-performing loan securitisations...9 September 2020 Burkhard Heppe , PhD Chief Technology Officer NPL Markets Ltd. burkhard.heppe@ nplmarkets.com +44 (0) 20

9 September 2020 Burkhard Heppe, PhD Chief Technology Officer NPL Markets Ltd. burkhard.heppe@ nplmarkets.com +44 (0) 20 3984 6288

Cash flow models for non-performing loan securitisations

Table of Contents

Overview 1

Outright sale versus structured NPL transactions

1

Investor due diligence requirements in securitisation regulation

3

The role of cash flow models in securitisation regulation

5

Modelling the underlying asset pool performance

9

Structural elements of liability cash flows

10

Building cash flow models for structured NPL transactions

12

Conclusions 14

References 15

Overview We describe the transaction structures currently seen in European securitisations of non-performing loans (NPL) with a particular focus on the liability cash flow waterfalls. Asset-backed securities (ABS) backed by NPL have been issued mostly in Italy thus far with a few deals in Portugal, Spain, Ireland and Greece. The current downturn caused by Covid-19 is expected to increase the volumes of NPL securitisation as an important tool for banks to transfer larger portfolios to investors to reduce NPL from their balance sheets. Given the current economic pressure in many countries, we expect that NPL securitisation will become more important outside of Italy. We explain the relevant aspects of the European securitisation regulation for NPL ABS and the prevalence and use of different structural details highlighting some of the differences to performing loan ABS. This article complements our previous articles on the valuation and reporting challenges of NPL. The ESMA securitisation disclosures, which will enter into force on September 23, are a particular challenge for NPL ABS as discussed in NPL Markets (2020a).

Outright sale versus structured NPL transactions Non-performing loans in Europe had been declining for several years prior to the Covid-19 outbreak in early 2020. Economic forecasts from the IMF, the OECD and many central banks predict a recovery from the dramatic decline in economic activity during the lockdown in March and April of 2020, but the speed of this recovery is uncertain given the possibility of further waves of Covid-19 infections. Many forecasters expect bank loan defaults to increase with a wave of new NPL once the extraordinary state aid programmes are phased out and once the economic contraction translates into higher

© NPL Markets Ltd. 2020 1

mailto:[email protected]

mailto:[email protected]

unemployment and borrower distress (NPL Markets 2020e). The currently observed or expected increase in NPL in many countries will require a range of solutions for banks to deal with distressed loans. Banks can either transfer resources internally and work out the loans on the balance sheet or alternatively transfer the loans to third-party investors in an outright sale or by means of a structured transaction. Table 1 provides an overview of the different transactions’ structures used to transfer NPL.

Table 1: Overview of structured NPL transactions. Working out the loan on the balance sheet incurs operating costs as well as increased capital costs when the NPL remain on the balance sheet for too long from the recently introduced prudential backstop (NPL Markets 2020f). Single loans and small and medium size portfolios of NPL up to a purchase price of around Euro 50 or 100 million are normally sold outright to one investor whereas larger deals are often split in two or more risk tranches in form of a structured transaction. While outright sales are more frequent, the NPL transaction market by volume is dominated by structured transactions. Table 2 shows the number and volume of public NPL securitisation issued in Europe from 2017 to June 2020. While several European NPL securitisations have been underperforming prior to Covid-19, as of June 2020 the majority of European deals are underperforming with several downgrades. Scope (2020) lists 7 out of 27 downgrades of the senior most tranche for Italian NPL ABS mostly on transactions issued prior to 2019. Irish and Portuguese NPL ABS have also been downgraded (DBRS 2020). Some transactions had their rating confirmed despite the underperformance relative to original servicer business plan projections given the available overcollateralization and credit enhancement. For instance, Moody’s recently confirmed the ratings for the large EUR 24bn GBV Siena NPL 2018 transaction despite underperformance in 3 out of 4 subpools (Moody’s 2020).


Structured transactions include the sale to a special purpose vehicle where part of the acquisition cost is funded by a senior bank loan with limited recourse to the equity investor who acquires the junior risk tranche. Public or private securitisations are well-known structured transactions where the senior and mezzanine risk tranches are distributed to investors in the form of asset-backed securities. NPL ABS are often rated by one or more credit rating agencies. In Italy, 24 out of 27 NPL ABS benefit from a guarantee of the Italian government on the senior class note provided under the GACS scheme. GACS was originally introduced to ease the disposals of NPL and extended in 2019. In Greece, a similar government guarantee scheme called Hercules Asset Protection Scheme (HAPS) has seen the first transactions earlier this year with more transactions expected before year end.

Table 2. European NPL Securitisations. The gross book value of loans underlying public Italian NPL ABS amounts to EUR 75.7bn. Greece counts the Pillar and Cairo transactions. The senior guarantees in Italy are issued under GACS and in Greece under HAPS. Source: DBRS 2020, Scope 2020. Private outright sales or structured transactions are not included in this table. Other structured transaction types include the transfer of loans to an investor with an upside share agreement where the buyer agrees to pass some of the upside returns back to the seller. Another structured transaction is a hive down of larger volumes of NPL to a state-sponsored asset management company (AMC). The AMC or bad bank structure is currently considered by the European Central bank at the European level and have successfully been introduced in countries such as Italy, Spain, Germany and Ireland after the 2009 financial crisis. Investor due diligence requirements in securitisation regulation In Europe and many other countries, structured transactions with two or more stratified risk tranches reflecting different degrees of credit risk are regulated as securitisation transactions. Payments to the investors depend upon the performance of the specified underlying exposures, as opposed to being derived from an obligation of the entity originating those exposures. The tranched structures that characterise securitisations differ from ordinary senior/subordinated debt instruments in that junior securitisation tranches can absorb losses without interrupting contractual payments to more senior tranches.

The European Securitisation Regulation (SR) applies to securitisations that issue new securities from 1 January 2019 (ESMA 2020). Earlier securitisations are grandfathered and prior investor due diligence and risk retention requirements apply under the previously applicable regimes (i.e. CRR, AIFMR or Solvency II regimes for banks, funds and insurance companies, respectively). The SR sets out detailed


due diligence requirements including risk retention and other verification needs that must be conducted by European-regulated institutional investors before and whilst holding an exposure to a securitisation. Table 3 compares the securitisation regulation in different jurisdictions with regards to certain due diligence, disclosure and retention requirements, respectively. Previously, we looked at the detailed disclosure requirements and the use of the no-data option for European securitisations pointing out the particular challenges for NPL securitisations (NPL Markets 2020a, b). The detailed disclosure requirements will come into force on September 23, 2020 (ESMA 2020).

Table 3: Securitisation regulatory frameworks and disclosure requirements for loan-level data and cash flow model.

Loan-level disclosures are standard now for many securitised asset classes, however, to our 1

knowledge only the EU provides a dedicated data template for NPL securitisations. The same EU template was also adopted by the UK. Risk retention and other investor due diligence requirements apply in all jurisdictions considered here, but the details may differ. Institutional investors in many jurisdictions are required to perform due diligence before investing in ABS and while holding the investment.

For example, Article 5(3) of the SR states: “Prior to holding a securitisation position, an institutional investor, other than the originator, sponsor or original lender, shall carry out a due-diligence assessment which enables it to assess the risks involved. That assessment shall consider at least all of the following:

1. The risk characteristics of the individual securitisation position, 2. the risk characteristics of the underlying exposures, and 3. all the structural features of the securitisation that can materially impact the performance of the

securitisation position, including the contractual priorities of payment and priority of

1 We use the term loan-level or exposure-level disclosure synonymously as opposed to the less granular pool level disclosure.


payment-related triggers, credit enhancements, liquidity enhancements, market value triggers, and transaction-specific definitions of default; [...]”

Hence, the regulatory due diligence requirements demand both the analysis of the underlying assets as well as a detailed understanding of all structural features including the priority of payments to the different risk tranches. For a summary of recent regulatory changes in Europe that impact NPL securitisations see NPL Markets (2020g).

How will the European SR change the way that investors will analyse non-performing loan securitisations? On the asset side, the standardized ESMA exposure-level discloses will provide more transparency on the underlying assets and thus a better analysis of the so called risk layering where investors can see the cumulation of several risk factors and their prevalence in the portfolio (for examples, loans to borrowers in formal insolvency proceedings backed by industrial properties in a certain geographical region). The standardized historical collection data reported in the NPL Add-on Annex 10 of the ESMA disclosure requirements is important for monitoring the workout progress of the loans. However, the ESMA annexes miss a number of important data fields that investors in NPL securitisations rely on. The two most important deficiencies are the lack of servicer business plans and the lack of granularity on foreclosure expenses and servicing fees. Also, it is not currently clear how complete the ESMA annexes will be populated given the generous thresholds to use the no-data option for NPL securitisations (NPL Markets 2020b).

On the liability side, the EU, UK and Australia, require the arranger or originator to provide a detailed liability cash flow model for certain ABS to help investors understand and model the priority of payments and other structural elements like amortisation provisions, reserve funds, performance triggers or events of default. In the EU, the requirement to provide liability cash flow models applies to simple, transparent and standardized (STS) securitisations only and therefore exclude NPL securitisations together with most CMBS and CLO transactions.

Note that the European SR requires certain elements about the liability cash flows to be disclosed for all public or private securitisations including NPL, CMBS and CLO. Annex 12 of the disclosures, the pool-level investor report for non-ABCP securitisations, includes a section on tests, events and triggers as well as a section on cash flow information. For each cut-off date and reporting period the reporting entity must provide the cash flow item in the liability waterfall, the amount paid during the period and the available funds post application of the waterfall to the respective cash flow item.

The role of cash flow models in securitisation regulation Article 22(3) of the SR requires the publication of cash flow models for STS ABS that accurately describe the allocation of the cash flows produced by the underlying pool of exposures to the different liabilities of the securitisation special purpose entity (SSPE). The representation of the contractual relationships must be sufficiently accurate to allow investors to model payment obligations of the SSPE


and to price the securitisation accordingly. This may include algorithms that permit investors to model a range of different scenarios that will affect cash flows, such as different prepayment or default rates. The requirement for originators or sponsors of securitisations to publish cash flow models also exist in the UK and Australia. The Bank of England (2019) requires all securitisation transactions to make cash flow models available to investors free of charge. As with the European regulation, the cash flow model must precisely represent the contractual relationship between the underlying exposures and the payments flowing between the originator, sponsor, investors, other third parties and the SSPE. Cash flow models may be provided in either a spreadsheet format or on a website maintained by or on behalf of the transaction Originator or Arranger; or in a proprietary format on the platform of a third party provider. The model should enable the user to input key collateral data as well as output results using a recognisable spreadsheet format. The model should incorporate data from a pre-configured table of inputs to drive a cash flow model, provided by the model provider, and output the resulting cash flows for the expected life of the relevant bond.

Whilst inputs and outputs are bespoke to each transaction, at a minimum the Bank of England expects inputs to cover asset specific characteristics like principal and interest received, delinquencies and defaults and liability features like note balances, trigger breaches as well as interest and exchange rates. The output of the cash flow model should clearly show the payments in the waterfall, account and note balances for the life of the transaction. The model should incorporate all features of the transaction which are not open to change or interpretation like e.g. the note interest margins or the priority of payments in the waterfall.

In the US in 2010, in direct response to the problems with disclosures and business practices in the ABS market revealed during the 2009 financial crisis, the Securities and Exchange Commission (SEC) proposed substantial reforms to the regulations governing the issuance of ABS. Included among the proposed reforms was an unprecedented type of securities regulation: a requirement that the issuer of each ABS develop, and make publicly available, software that models the cash flows from the pool of assets and their waterfall structure. The SEC specified that the software must be written in the Python programming language. The cash flow model software development proposal met with fierce opposition from ABS issuers, other market participants, and the American Bar Association, and in August 2011 the SEC removed the requirement when it released the final rules (SEC 2014).

With the benefit of hindsight, the original SEC proposal from 2010 to request Python code looks progressive and we think it is regrettable that it was not introduced. Arguments from industry players and vendors at the time were that the securitisation industry was not using Python, that the waterfall was only a small part of the code required to model assets and liabilities, and the claim of the SEC that creating the Python code would take around two hours per deal was seen as unrealistic. Vendors like Intex or Bloomberg argued that the subprime mortgage losses of the financial crisis were not the result of failures in the cash flow models. We disagree based on our experience with analysing numerous defaulted subprime CDOs after 2010. Rating agencies and commercially available cash flow models at


the time did not account for certain important transaction features like the portfolio liquidation after an event of default. While raters and investors focussed on asset risk characteristics such as mortgage default, loss severity and prepayment vectors driving the asset performance which were supported by the vendor models, it was not the realised loan losses that wiped out the CDO investors but the asset liquidation at depressed market values (i.e. unmodelled market value risk rather than the modelled credit risk). To our knowledge, commercially available cash flow models at the time did not offer a market valuation of the underlying assets and the ability to model the liquidation in a forced sale. The risk of a forced liquidation after an event of default remains an important risk factor in today’s securitisations especially for NPL ABS.

In 2017, Deloitte published a study about the potential use of blockchain technology to advance the securitisation industry into the digital age. Deloitte (2017) gave the liability cash flow waterfall as an example for a smart contract, i.e. a software code, that could be executed automatically on the blockchain and thus increase security and efficiency. As of today, we understand that the use of different cash flow models by different participants in the same deal gives rise to operational uncertainty and unnecessary interpretation and reconciliation problems. The idea of complementing several hundred pages of legal documentation with a concise computer code remains very appealing and smart contracts may receive a wider adoption in structured finance in future.

Whereas parts of the securitisation market are still based on Excel cash flow models, the availability of large loan-level data sets has encouraged participants in the ABS industry to adopt an enterprise standard IT environment with code based queries of external or internal data warehouses for due diligence, reporting and valuation. Python and the statistical programming language R are now standard for coding complex data science problems and enjoy a widespread use in the financial industry. The use of code as a smart contract to communicate the transaction structure has great merit, but just publishing a snippet of code with the liability waterfall will not be useful and sufficient.

Other solutions exist to provide market participants with consistent, immutable and easy-to-use information with the ability to execute cash flow calculations automatically that do not require a blockchain infrastructure. Securitisations are regulated and trusted custodians exist (such as trust service providers, securitisation repositories, central bank and supervisory websites) neutralising one benefit of the blockchain. To be truly useful to all participants a comprehensive solution is required that addresses data consistency and integration, user interaction and the ability to run different asset side scenarios in addition to the liability waterfall. A solution could be to add an executable container to the regulatory disclosures that users can run on their own IT infrastructure or in the cloud. Such containers are transferable, version controlled, and agnostic to the computer language. They are self-sufficient software packages that do not have to be installed and can be isolated from other software. Importantly, such containers are already in widespread use in IT departments and cloud deployment is available. The container solution would be friendly to end users and IT departments offering


consistency and immutability without the disadvantages of Excel spreadsheets or standalone code snippets.

Figure 1: Data lifecycle for structured NPL transactions.

The liability cash flow model is only one of several important components in managing the information flow of securitisation transactions. Figure 1 displays the full life cycle of data related to a securitisation starting with the extraction of the data from the originator or data provider up until the ongoing reporting at loan-level to investors and securitisation repositories to meet the disclosure requirements of the European SR.

Data extraction, transformation and validation for completeness, consistency and accuracy is a critical first step for data providers benefitting from dedicated software for ETL (extract, transform, load) and validation. A loan valuation engine offers the user the ability to run a set of assumptions in different economic scenarios governing the workout of the loans, the probability of different workout scenarios, the timing and amount of future recoveries given the characteristics of the loan, borrower, collateral, legal process and prior collections.

The loan valuation should include an impact analysis of the balance sheet of the lender taking into account current and expected operational and capital expenses. For example, the prudential backstop with future calendar provisions will lower the break-even price at which a bank should be willing to sell an NPL today taking into account expected future capital expenses. The valuation analysis allows the seller to select an optimal portfolio that maximizes the economic benefit of the transaction. The projected asset cash flows are then fed into the liability waterfall for a structured deal to value and assess risk tranches of different seniority and to assess the impact on the internal rate of return on the junior risk tranches of different structural features. The liability cash flow model produces the current and future bond cash flows to conduct a rating and risk analysis under different scenarios and to calculate bond metrics like the price/yield, the weighted average life and expected maturity. Once a


transaction has been executed, the data provider will produce a series of reports including collection updates and changes to the loan-level data for investors and as part of the regulatory disclosures.

Modelling the underlying asset pool performance

While the modelling of liability cash flow waterfalls is a necessity which should be based on a legal analysis and hard facts, the modelling of the asset cash flow remains highly subjective and different users employ different assumptions and valuation techniques. For details of the valuation methodology for non-performing and unlikely-to-pay loans deployed by us on the marketplace and valuation tool please see NPL Markets (2020d). At the core of a valuation tool for non-performing loans is a model to generate loan-by-loan (and for secured loans collateral-by-collateral) recovery cash flows for each future calculation period net of workout expenses together with assumptions about servicing and transaction costs. The asset cash flows can be discounted for pricing an outright sale of the assets or can be fed into the liability cash flow model of a structured transactions or securitisations.

To assess the asset cash flows, investors and credit rating agencies will analyse: 1. the performance, experience and expertise of the special servicer, 2. the collateral pool including loan security where available, 3. the projection of future recovery cash flows by the servicer (the business plans), 4. historical performance data of the servicer or originator and 5. a cash flow scenario analysis using cash flow waterfall models. 2

Naturally, the analysis of recovery amounts and timing is more important for NPL than for performing loans. These two variables are a function of certain quantitative and qualitative factors:

1. the data and projections provided by the servicer or sponsor (or by the arranger on their behalf ) in the portfolio business plan,

2. historical servicer recovery data (provided on a loan-by-loan basis for secured exposures and on a static basis for unsecured NPLs),

3. exposure-level reviews of secured loans and/or larger loans or obligors, 4. rating agency’s economic outlook for the analysed jurisdiction, and 5. any expectation of the amount of NPL in the particular market that are likely to be worked out

over the same time period. Investors and rating agencies will start the analysis of NPL portfolios with a review of the servicer’s gross and net recovery expectations over time (DBRS 2020). For granular portfolios of unsecured loans, such business plans are often provided on an aggregated basis and are accompanied by historical recovery data of similar portfolios. For secured loans and/or larger exposures, investors expect to

2 The list shows that the mandatory ESMA disclosures are insufficient to analyse an NPL securitisation as they lack the servicer business plans as well as the cash flow models.


receive the servicer’s detailed recovery expectations for each loan, obligor and collateral where applicable. Figure 2 shows the key elements of the asset and liability models that together allow investors to value their ABS holdings. Performance triggers for NPL differ from performing pool triggers. Most common is a test to compare actual cumulative recovery collections against those predicted by the servicer, the cumulative collection test. Another commonly used test is based on the profitability of those loans for which the workout has been resolved. In NPL Markets (2020c) we highlighted some pitfalls with the profitability ratio as a performance measure and the risk that overperformance in some subpools can mask the underperformance in other subpools. Investors should carefully analyse the performance versus original projections at loan level or at the level of subpools defined by the investor as required.

Figure 2: Asset and liability cash flow model elements. Some elements like loan characteristics, pool composition and historical collections on the asset side or the priority of payments on the liability side are factual whereas other elements are assumptions that require user inputs. Structural elements of liability cash flows Aggregate recovery cash flows net of work-out costs collected in each period are distributed to securitisation noteholders based upon the priority of payments (the liability cash flow waterfall) established in the transaction documents. Unlike performing asset securitisations where principal and interest collections may be accounted for separately and then subjected to separate payment waterfalls, in NPL securitisations, recoveries are normally aggregated to create a single pool of total available funds, which are then subject to a payment waterfall. Recoveries typically first pay for recurring transaction expense items like trustee, liquidity facility, senior servicing, guarantee (if any) or other transaction management fees, and then pay noteholders interest and principal. Figure 2 shows the main elements of the liability model and Figure 3 displays an example of the detailed priority of payments for an Italian NPL ABS with three classes of notes (Senior, Mezzanine and Junior).


The waterfall is adjusted after an interest subordination event or an acceleration event. The payment of interest on the mezzanine notes is subordinated to the payment of principal on the senior notes in case the portfolio underperformance e.g. if the cumulative collections received drop below a given percentage of the collections expected in the servicer business plan. An acceleration event occurs at legal maturity or can be triggered by an event of default or an optional redemption of the notes (the latter is only possible if the notes can be repaid in full). Performance thresholds or triggers can mitigate the effects of a deteriorating economy and poor collateral performance. The performance triggers are designed to provide for an increase in credit enhancement levels beyond what is initially included in the transaction, thereby enabling the transaction to withstand a lower-than-expected recovery performance; for example, by shutting off cash paid to mezzanine or junior notes or by switching from pro rata to sequential priority of payments. The example in Figure 3 follows a strict sequential pay structure after an interest subordination event, i.e. funds are first applied to the senior class of notes until it is fully repaid and then directed to the next class of notes in the waterfall. Alternative structures that are more beneficial for the mezzanine and junior investors may see some pro-rata amortisation of principal of the subordinated notes prior to full repayment of the senior notes as long as the portfolio performance is equal or better than expected.

Figure 3: Liability cash flow waterfall for an Italian NPL securitisation (Futura NPL 2019). The figure shows the pre-acceleration waterfall of a structure without GACS guarantee. Interest payments on the Mezzanine Class B notes can be subordinated upon an Interest Subordination Event (the curved arrow). After an acceleration event, the waterfall is amended by excluding the positions 5 and 6 (light blue). To achieve an investment grade rating on the senior class of notes, NPL securitisations benefit from structural elements that provide liquidity support and credit enhancement. Recovery cash flows are volatile and the higher the expected volatility of periodic cash flows and the less granular the portfolio, the greater the need for liquidity support from cash reserves or a committed liquidity facility from creditworthy counterparties. The liquidity support will bridge cash flow timing mismatches and will mitigate the risk of cash flow shortfalls as a result of cash collections underperforming the business plan with regard to the amount and timing of recoveries.


The required amount of credit enhancement to support a desired target rating on the senior class of notes depends on a number of factors, including the overall amount and timing of recoveries and potential disruptions in collections. Credit enhancement may be provided through a combination of subordination, a reserve fund and overcollateralization.

● Overcollateralization consists of the amount by which the asset collateral amount exceeds the total amount of securities issued as liabilities.

● Subordination is created by tranching and assigning a lower priority of payments to subordinated classes of notes.

● Cash reserve accounts provide liquidity to address recovery shortfalls from the pool of NPL. Such accounts may be funded at issuance or filled by trapping funds to a pre-specified amount after paying interest on the senior notes and before funds are allocated to other payments.

Table 4 compares elements of securitisation models for performing loan and NPL securitisations. Table 4 does not provide a comprehensive list of performing loan cash flow model features which depend on the asset class, but rather highlights some important differences for NPL securitisations. Building cash flow model software for structured NPL transactions The NPL Markets cash flow model builder software follows the guidance on cash flow models by the Bank of England and EBA and generates the required reporting items on triggers and cash flows for the ESMA securitisation disclosures under Annex 12. It combines the advantages of a code based model with rigorous data formats, version control, automated testing etc. while producing an easy to understand Excel output file that non-programmers can understand. The software allows the user to establish the deal structure by setting a series of deal input parameters in a configuration file. The software generates the asset and liability cash flows for download as an Excel file for further testing and better understanding of the deal structure. Asset scenarios can either be fed as collection vectors generated by the user’s own tools, by using the NPL Markets valuation tool or through a set of parameters that describe the recovery curves for each of several subpools. The user can design a new transaction by testing different transaction structures by setting user input parameters (e.g. two or three note classes, a fast and a slow pay A1/A2 senior tranche, different rules regarding the subordination of interest on the mezzanine notes or subordination of servicing fees in the case certain performance triggers are breached). The user can test what happens in case of an early acceleration or enforcement event should the junior note holders decide not to provide additional funds through the optional liquidity facility and rather allow for an early liquidation of the assets through a portfolio sale. The Excel output allows users to understand all deal features without the need to understand the code. Using the software means that sponsors, arrangers or investors can create a full asset and liability cash flow model for European NPL securitisations in less than the 2 hours in line with SEC’s timing estimate in 2010.


Table 4: Comparing cash flow model elements for performing and NPL securitisation transactions.


Conclusions We explain the regulatory requirements for due diligence and disclosure including the provision for cash flow models. Cash flow models are a necessary component of analysing, stress testing and valuing structured transactions and securitisations. Cash flow models for NPL securitisations share many similarities with performing loan securitisations but several aspects are unique to the NPL asset class such as collection and profitability based performance triggers, the availability of senior tranche government guarantee schemes in some countries and the importance of special servicing fees across different positions of the liability cash flow. Cash flow models are required to translate stressed scenarios on the asset performance into liability cash flows for valuation or risk analysis. Cash flow models are also required for a derecognition analysis in case the seller of the NPL retains a residual interest in the performance of the assets either through an vendor financing arrangement and/or an upside share agreement.

Additional disclosure of detailed loan-level data and liability cash flow models will only increase market efficiency if these disclosures are accurate, complete and accessible i.e. can be used by the market with acceptable financial or operational effort. As long as the currently used non-standard investor report for NPL securitisations contain unique and necessary information like the servicer business plans or the details of the workout and servicing expenses, then the investor will have increased costs and effort to analyse and reconcile the existing disclosures with the new exposure-level ESMA disclosures that might include important additional insight in risk layering. Liability cash flow waterfall models are currently not part of the disclosure requirements for NPL securitisations in the EU and the US and thus investors will either develop these models themselves or rely on commercial vendors. As long as there is no smart contract with a precise coding of the cash flow allocation it is possible that arrangers, agents, or investors will encounter hard to reconcile differences.

At NPL Markets we welcome additional regulatory disclosures and endeavour to integrate all available information in a consistent and easy to use data and model infrastructure for deal structuring, valuation and reporting. We aim to provide all market participants from arrangers, advisors, loan sellers and buyers with seamlessly integrated tools for a fast and accurate analysis of all aspects of a non-performing loan transaction covering the entire data lifecycle of outright or structured transactions.


References DBRS Morningstar 2020. Methodology Rating European Non-Performing Loans Securitisations. European Securities and Markets Authority (ESMA) 2020. Securitisation regulation and disclosures. https://www.esma.europa.eu/policy-activities/securitisation Moody’s Investor Services 2020. Moody's takes rating actions on five Italian NPLs deals (July 2,2020). https://www.moodys.com/viewresearchdoc.aspx?docid=PBS_ARFTL427770 NPL Markets 2020a. The importance of collecting historical recovery data for NPL transactions. https://www.nplmarkets.com/news/article/the-importance-of-collecting-historical-recovery-data-for-npl-transactions NPL Markets 2020b. The No Data Option in Securitisation Disclosures. https://www.nplmarkets.com/news/article/the-no-data-option-in-securitisation-disclosures NPL Markets 2020c. Heterogeneity matters: the benefits of loan-level investor reports for non-performing loans. https://nplmarkets.com/research/article/heterogeneity-matters-the-benefits-of-loan-level-investor-reports-for-npl-transactions NPL Markets 2020d. How to value bank loans in a crisis. https://nplmarkets.com/research/article/how-to-value-bank-loans-in-a-crisis NPL Markets 2020e. Forecasting NPL ratios after Covid-19. https://nplmarkets.com/research/article/forecasting-npl-ratios-after-covid-19 NPL Markets 2020f. Impact of prudential backstop on bank balance sheets. https://nplmarkets.com/research/article/impact-of-the-prudential-backstop-on-bank's-balance-sheet NPL Markets 2020g. Regulatory Update for NPL Securitisations in Europe. Scope Ratings 2020. Italian NPL ABS: collection data improves but uncertainties still linger. Securitisation Exchange Commission 2014. Asset-Backed Securities Disclosure and Registration. Final Rule. https://www.sec.gov/rules/final/2014/33-9638.pdf


https://www.esma.europa.eu/policy-activities/securitisation

https://www.moodys.com/viewresearchdoc.aspx?docid=PBS_ARFTL427770

https://www.nplmarkets.com/news/article/the-importance-of-collecting-historical-recovery-data-for-npl-transactions

https://www.nplmarkets.com/news/article/the-importance-of-collecting-historical-recovery-data-for-npl-transactions

https://www.nplmarkets.com/news/article/the-no-data-option-in-securitisation-disclosures

https://nplmarkets.com/research/article/heterogeneity-matters-the-benefits-of-loan-level-investor-reports-for-npl-transactions

https://nplmarkets.com/research/article/heterogeneity-matters-the-benefits-of-loan-level-investor-reports-for-npl-transactions

https://nplmarkets.com/research/article/how-to-value-bank-loans-in-a-crisis

https://nplmarkets.com/research/article/forecasting-npl-ratios-after-covid-19

https://nplmarkets.com/research/article/impact-of-the-prudential-backstop-on-bank's-balance-sheet

https://www.sec.gov/rules/final/2014/33-9638.pdf

About NPL Markets

NPL Markets is an innovative marketplace for illiquid loan trading, operating throughout Europe. The platform is based upon four pillars: data preparation, marketplace execution and investor reach, valuation, and reporting. NPL Markets helps sellers of NPL to prepare and standardize transaction data and select the optimal transaction portfolio based on balance sheet impact. Our platform supports the entire transaction data lifecycle and assists sellers and buyers with loan valuation, deal structuring, bond valuation and reporting tools.

With the help of its data mapping and transformation tool NPL Markets assists financial institutions to map their data to several data data formats defined by EBA for NPL transactions, by EBA for the valuation in resolution, or by ESMA for securitisation disclosures. Once standardized and validated the loan-level data can be uploaded to the NPL Markets valuation tool to conduct a detailed discounted cash flow analysis using pre-populated pricing parameters in different macroeconomic scenarios. We support the transfer analysis of all major asset classes by outright sale or by means of structured transactions and securitisations. Disclaimer This paper contains confidential information about NPL Markets, current at the date hereof. This presentation is not intended to provide the sole basis for evaluating NPL Markets and should not be considered as a recommendation with respect to it or any other matter. This document and the information contained herein are not an offer of securities for sale in the United States and are not for publication or distribution to persons in the United States (within the meaning of Regulation S under the United States Securities Act of 1933, as amended). This presentation and the information contained herein does not constitute or form part of any (i) offer or invitation or inducement to sell or issue, or any solicitation of any offer to purchase or subscribe for, any securities or (ii) invitation or inducement to engage in investment activity within the meaning of Section 21 of the United Kingdom Financial Services and Markets Act 2000, as amended, nor shall any part of this presentation nor the fact of its distribution form part of or be relied on in connection with any contract or investment decision relating thereto.


Documents

Cash flow models for non-performing loan securitisations...9 September 2020 Burkhard Heppe , PhD Chief Technology Officer NPL Markets Ltd. burkhard.heppe@ nplmarkets.com +44 (0) 20