Due Diligence

Due diligence is one of the cornerstones of the investment process of the PGGM Credit & Insurance Linked Investments (“CILI”) team.

Due to the blind pool nature of Credit Risk Sharing transactions that we invest in, it is essential to have a thorough understanding of the bank and the relevant processes and track record, both from a qualitative and quantitative perspective.

Key elements of PGGM’s philosophy are developing long-term partnerships with its partner banks and to share in the core businesses in which these banks have market leading positions. We provide capital to the banks by risk sharing in a representative piece of these core businesses, for which we believe our risk sharing partners are better positioned to originate and assess credit risk than ourselves. As a consequence, we do not assess the individual credit exposures in the underlying portfolio, but rather assess the quality and track record of our risk sharing banks in credit underwriting and risk management. In addition, we look in-depth at all risk characteristics.

We devote considerable time to understand the underlying exposures as well as the bank’s approach to the specific lending business. The track record of the bank in terms of credit risk management is very relevant as well. As a blind pool investor, one has to have a thorough understanding of all the relevant processes to be able to make a well-informed investment decision. This involves detailed due diligence on the nature of the underlying exposures, the relevant processes in the credit risk originating and risk management, as well as reviewing the steps and teams involved in the loan-life-cycle of the relevant credit book. This assessment is mostly qualitative in nature.

The assessment of the credit risk management track record of the bank is mostly quantitative in nature, and focuses on the internal credit rating and loss-given-default models and historical default data.. The outcome of the quantitative assessment allows us to form our own independent opinion of losses that can be expected in various economic scenarios and drive the assumptions we make in the transaction pricing analysis. An important input is the analysis of internal PD and LGD models and the historical track record. The historical track record covers at least an economic cycle and is based on single and multiyear default rates per rating category and the loss-given-default experienced for substantially similar exposures.

The qualitative and quantitative assessment converge, as it is important to understand to which extent past performance is expected to be representative for future performance. Due diligence findings influence pricing, as they either confirm the initial pricing assumptions or may cause an investor to tighten or widen these. Findings in due diligence also affect the choices made in the structuring process, for example via the eligibility and portfolio criteria or mitigating risks for adverse selection or moral hazard in the transaction governance. More information on the structural considerations can be found here. In the next section, the due diligence topics will be described in more detail, distinguishing between qualitative and quantitative aspects.

Qualitative Assessment

As investor we have to make a conscious decision to work with banks that have – next to a strong market position – a robust track record in origination and managing the credit risk exposures relevant for a given transaction. We therefore conduct an intensive due diligence to not only fully understand the specific credit risks and processes of the bank but also to ‘understand the underlying’.

“Understand the underlying” first of all and most importantly relates to a thorough understanding of the underlying credit risks:

I. The types of credit risk which are part of the reference portfolio, the specific characteristics of the underlying products and related loan contracts through which the credit risk emerges; and

II. The function of these products to the borrowers, the performance over time, the broader market of these products and the role of the relevant players over time.

The above forms a relevant basis for assessing the business strategy of the bank in relation to the relevant credit risk type(s), and the quality of people and processes involved. To that end we put significant emphasis on conducting an intensive due diligence to fully understand the specific approach of the bank, which in the end is reflected it its track record of realised losses and how these compare to predictions of loss expectations throughout time. The due diligence aims to obtain in-depth insight into the types of loan contracts and collateral structures, and learn what drives the risk and recovery profile of that specific lending segment. Relevant characteristics will then be incorporated in the deal criteria and taken into account for pricing.

Reference Registry

The reference registry is the core of any risk sharing transaction and therefore is the starting point of our analysis and due diligence. As an investor we require minimum reporting standards (see CRS template) covering required information a bank will need to provide on the loans in the portfolio they want to hedge. In short, the data requested includes amongst others: (i) unique obligor and group IDs; (ii) maturity date of the loan; (iii) the reference obligation notional (including currency and FX rate if applicable); (iv) the country of risk and industry group; (v) internal rating and LGD and the associated models used; (vi) details on type of loan and related security position; and (vii) applicable red flags such as watch list classifications. Depending on the specific loan book, additional details can be requested.

Strategy and Market Position

The strategy and market position of a partner bank is an important part of the philosophy of PGGM. We aim to share only in the core activities of a bank, as those give a bank a reason to exist and will most likely receive full attention to ensure ongoing high quality and successful risk management. By combining the requirement of ‘core’ with a strong and leading market position, a selection is made towards banks that have a successful long term track record, able to adapt to changing market circumstances over time and are likely to remain successful. Therefore, we assess what the strategy and market position of a bank is regarding the envisaged risk sharing book and how the proposed risk sharing transaction relates to the bank’s strategic objectives.

Loan Life Cycle

The entire loan life cycle of a bank is covered during due diligence, from origination to workout. First, the entire origination process is assessed. The goal is to understand how new loans are initiated, what the underwriting strategy is and which steps it follows in the banks processes. This involves thorough understanding of how a new credit application comes about, who is involved and signs-off on it and key areas covered. In addition, having insight into the pricing mechanism for loans and how credit risk and the cost of capital are taken into account will provide insight to what extent there is alignment between the business appetite and risk appetite of the bank.

Next, the approval process for granting new credit is examined to get a detailed overview of the departments involved in the approval process and the division of responsibilities, often part of the three lines of defence model of a bank. This assessment includes insights into the composition, level of authority and voting mechanism of each relevant credit committee, and the specific role of credit risk, ESG and the portfolio management departments in the approval process. Particular attention is paid to the credit culture of the organisation and whether the bank has a clear separation of roles between the business line, risk and portfolio management.

The rating and loss-given-default models and related model governance is also part of our review. Here we look at the applicable rating models, whether or not they have been internally created, and how often these are reviewed by the relevant supervisor. As a blind pool investor, we are relying on the bank’s internal rating and loss-given-default data for our pricing methodology. Therefore, the quality of the rating and loss-given-default models is an important part of our due diligence. Aspects covered during our assessment include monitoring of the models, the frequency of updates and validation of model calibration. Quality of the models should also be confirmed when looking at its predictive power in terms of rank ordering and volatility in outcomes of the models and related use of overrides. Understanding the relation between the internal rating and external rating allows us to assess overlap between the two which can be used as confirmation of our quantitative analysis. Model governance is related to the processes and responsibilities to assigning and updating internal ratings, whereby particular attention is paid to potential conflicts of interest.

Once a client has been onboarded the relationship will require constant monitoring and dialogue by the bank. It is important to understand how this is set up and which departments are responsible for the different steps in the monitoring process. Areas covered include the frequency of monitoring, systems used, common triggers that activate red flags and related timelines for action, including potential downgrades. The use of a watch list system is also discussed, including information on the related processes and responsibilities. The course of action of the bank for credit exposures that show negative developments is expected to have an impact on credit migration and ultimately potential loss amounts. Historical credit migration and loss amounts are part of the quantitative assessment and will need to confirm our qualitative findings on the monitoring process of the bank.

Finally, there will always be clients of the bank that face economic hardship. Getting a good overview on how a bank identifies and manages problem loans – including provisioning - is important for the risk sharing transaction and will allow us to substantiate the historical loss data. Here the entire workout process is reviewed, starting from identification and transfer of the loans to - and intake by - the recovery team, the strategies most often chosen with regard to the relevant lending book and the related decision authorities. This way we assess the resources available, separation of responsibilities and potential conflicts of interest that could harm the risk sharing transaction.

Credit Portfolio Management

The credit portfolio management team manages the credit risk of the bank and is generally responsible for executing CRS transactions on the bank’s side. Amongst others we want to understand the mission of portfolio management (for example managing credit risk, P&L volatility, concentration risk or setting prices), the role within the organisation and which hedging tools are used. Further, for the risk sharing transaction it is important to know how loans will be selected for the reference portfolio, whether this is done in an automated and non-discretionary way, and how the portfolio is optimised from the bank’s perspective (also depending on whether the ultimate goal is capital relief and/or limit relief). Taking into account structuring aspects such as eligibility criteria, the envisaged portfolio will be compared to the total underlying loan book of the bank to ensure that we are sharing in a fair representation of the core book.

Personnel and Culture

Core to all departments functioning well is the right personnel and culture. During the onsite sessions with the various teams of the bank, information will be requested on the background of the people working in the team, the average experience and number of employees. Headcount turnover and movement between the teams is also considered of relevance here.

Environmental, Social and Governance

It is a core belief of PGGM Asset Management that sustainable development is necessary in order to generate good and stable investment returns in the long run. This is embedded within the PGGM Responsible Investment Policy in Credit Risk Sharing. As part of our core beliefs, PGGM integrates environmental, social and governance (“ESG”) factors in the due diligence. During the assessment of the banks origination strategy and quality of the risk management process, we investigate which ESG policies the bank has implemented, how it ensures these are being adhered to and impact the bank’s decisions, and to what extent the convictions behind these policies are part of the bank’s culture. We seek to incorporate specific ESG criteria in the transaction itself as to ensure the lowest chance of our client having undesired exposures, as reflected through the CRS Exclusion List. Lastly, reputational risk is also analysed in relation to the type of clients a bank onboards, whereby at a minimum, sanction lists will need to be adhered to.

Quantitative Assessment

Mapping

There is a limit to how much historical data a bank can share, largely due to having used different systems and different data definitions over time. Given that we desire to have as much historical data as possible, we found a solution by using the historical data of Moody’s or S&P. This requires ‘translating’ or mapping as we call it. Hereby, we incorporate our view on the expected performance of a risk sharing portfolio in terms of defaults via the mapping of the bank’s internal ratings to the public credit ratings of Moody’s or S&P on a ‘through-the-cycle’ basis. We compare realised default rates for a substantially similar portfolio, based on exposure and rating profile, from the data provided by the bank to what could be expected based on Moody’s or S&P realised default rates. In this analysis we take into consideration the nature of the underlying exposure and the relevant geographies where relevant. We always ask to include enough historical data such that it covers at least one full credit cycle.

Recovery Rate Analysis

Based on an analysis of the LGD models that are relevant for the underlying credit exposure and historical loss data of the loan book, a quantitative assessment of expected recovery rates for the underlying portfolio can be performed. More specifically, we look at realised recovery rates of the underlying loan book, preferably on a loan-by-loan level, and analyse how this relates to the exposures in the risk sharing portfolio and the way recoveries will be determined for the CRS transaction. It may matter which type of credit events will be covered. The assessment includes checking how recoveries in the dataset are calculated, what the average expected LGD of the dataset is compared to the risk sharing portfolio, at which level recoveries are determined, and how these developed through time and different stages of the economic cycle. Further analysis can be done at, for example, country or regional level when this is relevant and to see whether there are notable deviations. Lastly, a comparison is made to the regulatory LGDs of the bank in order to benchmark our assumptions.

Simulation of Expected Losses

Based on the selected rating mapping and recovery rate assumptions, and taking into account the underlying portfolio characteristics (including eligibility and portfolio criteria set), a Monte Carlo simulation is performed to estimate the distribution of expected losses for the risk sharing transaction. This is used to determine the required credit spread. Unquestionably this quantitative assessment does not stand on its own, findings in the qualitative due diligence will need to confirm our pricing assumptions.

Data Quality

An important part of the quantitative analysis includes data quality checks. As the pricing assumptions and structuring considerations are based on the historical data provided by the bank, it is crucial that this data has a certain minimum quality and is reflective of the credit risk shared in the transaction.

The most important aspects are:

(i) The historical data should be representative for the underlying portfolio when looking at rating models, countries and type of borrower, amongst others;
(ii) The historical data should fit with the default definition of the CRS transaction (at a minimum the PD and LGD data should be based on the same definition of default);
(iii) The historical data should go back as far as possible, and cover at least a full credit cycle in order to assess behaviour of default in times of economic hardship;
(iv) The historical data should include cumulative multi-year default rates by cohort (rather than vintage), split out per internal rating. This allows us as an investor, going to be exposed to the underlying credit risk over several years, to adequately assess multi-year credit migration taking into account the specific rating profile of the underlying portfolio. LGD data should furthermore be as granular as possible, preferably at loan level.

As part of the data quality exercise we will also perform several back tests such as comparing our PD assumptions to both historically realised default rates and the bank’s regulatory PDs, examining whether rating models did an accurate job of predicting defaults and whether observed recovery rates are in line with modelled recovery rates.

Where qualitative and quantitative converge

Risk sharing transactions are about taking first loss risk, and therefore thorough due diligence is necessary in order to understand and get comfortable with the underlying risk that is going to be shared. In the end, the quantitative assessment of the bank’s credit risk management track record needs to coincide with the approach of the bank in terms of credit underwriting and monitoring that is examined as part of the qualitative due diligence. Here, the qualitative and quantitative assessments converge. Firstly, it is important to be able to qualitatively support the findings from data analysis rather than relying on data alone. Secondly, it is crucial to understand to which extent past performance is expected to be representative for future performance. A change in origination process, client focus, loan terms, or models used can lead future performance to be different than the data suggests. Ultimately, the outcome of due diligence and the resulting findings will affect the pricing assumptions, transaction structure and governance.