Does a default matter? 

Tue, 06 Feb 2018 09:17:13 +0000

An overview of the Loss given default


 IN my quest to provide information that eases the implementation challenges of IFRS 9 for financial institutions and those entities with significant financial assets meeting the amortised cost classification. 

I have written three articles in the last month, to provide a simplified overview of the International Financial Reporting Standard 9 (IFRS 9) expected credit loss (ECL) model.

In this vein, I discussed the formulae for calculating the expected credit losses under the general approach, as the Probability of Default (PD) x Loss Given Default (LGD) x Exposure at Default (EAD). In the first two articles, I gave an overview of IFRS 9 while the third article focused on how to introduce forward-looking metrics in the estimation of the credit losses.

Suffice to say, there are various models that are used to determine the individual components of PD, LGD and EAD for financial assets in calculating the expected credit losses.

A number of firms including EY do assist entities to develop those models.

In this article we focus our discussion on one of the components of the expected credit loss model, the LGD, introduced above, however let me first digress by addressing the provision matrix, an operation simplification introduced by the standard.

The developers of IFRS 9 were sensitive to the inherent difficulties in applying the expected credit loss model (i.e. among others, determining the individual components of PD, LGD and EAD as a basis to determine the expected credit losses) as such, it introduced some operational simplification guidelines with respect to receivables, thus the provision matrix.

The idea behind a provision matrix is that an estimate of the expected credit losses can be made for receivables based on the age of receivables or other risk classification.

The premise is that it is easier to use the provision matrix to calculate the expected credit losses as long as the relationship between the age of the receivables (or other risk classifications) and the potential risk of non-payment of the receivable balance is established.

The idea is to use the loss incurred in the past as a percentage of the credit exposure at the reporting date and then sanitise those credit losses by the expected future conditions that will have an effect on expected credit losses.

It is important to note that the provision matrix that only considers past incurred losses, might not generate the losses expected in the future, thus the incurred losses must be adjusted by expectations that are sensitive to current conditions and those subsisting in the future, including risks specific to the assets.

So what are the typical steps to take in developing a provision matrix in the determination of lifetime credit losses?

First, the receivables must be segmented into appropriate groups according to the age categories to determine the potential non-recoveries in each group.

Groupings can also be analysed by customer type/characteristics, geographical region or collateral. Subsequently for each of the aged receivables (or other risk category) by looking at the historical incurred statistics a percentage of the values of receivables that were previously collected as well as those that were written-off can be calculated to determine the historical loss ratios to attribute to each age-band.

These historical loss ratios are then adjusted if necessary by current economic factors, customer-specific conditions and specific credit management policies to determine the applicable loss ratios to apply to specific age categories of receivables (or other risk classifications).

It is necessary to note that although a simplified model is provided for in the standard, IFRS 9 does not prescribe how an entity should estimate the lifetime expected credit losses when using this Model, however, the use of a provision matrix is specifically cited as one of the practical expedients.

Assuming a provision matrix is utilised to calculate credit losses for a portfolio of K250 million receivables as depicted in table 1, we can see that the calculated credit losses are ZMW4.1 million based on a single percent of credit losses for each segment of receivables determined using historical losses adjusted for current conditions.

ZMW4.1Million would be the amount recorded as a credit loss provision in the financial statements at the reporting date.

Having noted the practical expedient above, let me revert back to the LGD discussions.

In the earlier articles, I had briefly introduced the Loss Given Default (LGD) as the credit losses incurred as a consequence of default after the deduction of the value of collateral and is expressed as a percentage of the ‘full loan value plus an expected conversion of the commitments of an entity during the life of those commitments, the exposure.

Expressed differently, the LGD can be calculated by taking the full credit losses divided by the Exposure at Default (EAD) or taking the full credit losses divided by the unsecured exposure at default.

Put another way, the LGD tries to quantify the percentage of the full value of the expected credit losses relative to an entity’s exposure, sometimes referred to as the Exposure at Default (EAD). The LGD tries to express the value of financial assets that can be wiped out if a borrower defaults.

Petr Jakubík and Jakub Seidler (2009 ) noted that “LGD is usually defined as the percentage loss rate suffered by a lender on a credit exposure if the obligor defaults.

In other words, even if the counterparty defaults (fails to repay the amount owed), the lender will usually succeed in recovering some percentage of the current amount owed in the process of workout or sale of the obligor’s assets.

This percentage is termed the recovery rate (RR), i.e. the following relation holds RR = 1 – LGD. The LGD can be estimated on the basis of historical data on realised losses.”

The historical LGD can then be extrapolated to the future, perhaps by regression analysis and then adjusted by the point in time economic factors.

Because of this, the LGD tends to be facility focused as the specific characteristics of the transactions, such as the collateral that is a part of the contractual arrangement entered into by entity, influences the extent of the risk of expected credit losses and therefore the estimated credit losses.

Of specific note is that there are many approaches for calculating an appropriate loss given default rate to be applied to each credit risk exposure.

At the minimum, a Loss Given Default Model must evaluate the value and quality of collateral an entity has for its debts. A high value of collateral translates into a low LGD.

Because of this, this is most likely to adversely impact lending behaviour in countries without formalised or advanced Obligors credit scoring systems.

This is more so for corporate facilities. For retail facilities, Tony Bellotti and Jonathan Crook (2009) in their article LGD models for UK retail credit cards put it more succinctly when they noted that “In general, for retail credit, there are five categories of circumstances that will affect the amount an individual repays on a defaulted loan and can be used to build models of LGD:

(1)        “Individual details, some of which can be collected at time of application such as age, income, employment, housing status and address;

(2)        Account information at default: date or age of account at default and outstanding balance;

(3)        Changes in personal circumstances of an obligor over time;

(4)        Macroeconomic or business conditions on date of default, or possibly with a lag or lead on date of default;

(5)        Operational decisions made by the bank, such as the level of risk they were willing to accept on the credit product and the process they use to follow up bad debt.”

This information can be used to determine the portfolio LGD which must be appropriately segmented rather than determining LGD for each retail facility for the purpose of simplicity.

Having noted the above, how can the function of LGD be determined? As noted above RR expressed as a formulae is 1- LGD, therefore LGD can be defined as 1-RR with RR being the value of Collateral divided by the value of the Loan or receivable or 1- (Collateral / value of loan or receivable including estimated commitments converted). This formulae expresses the percentage loss rate suffered by a lender on a credit exposure if the obligor defaults.

Further the Basel 2 Guidance on Paragraph 468 of the Framework Document provides the following measures of computing the recovery rate used in the determination of the LGD as follows:

  • “Discounting the stream of recoveries and the stream of workout costs by a risk-adjusted discount rate which is the sum of the risk-free rate and a spread appropriate for the risk of the recovery and cost cash flows,
  • Converting the stream of recoveries and the stream of workout costs to certainty equivalent cash flows and discounting these by the risk-free rate, or
  • By a combination of adjustments to the discount rate and the stream of recoveries and the stream of workout costs that are consistent with this principle.”

The article’s focus was on the overview of LGD, however the LGD calculation can be complex when other variables such volatility factors etc., are added. This is where the implemented systemic models come in handy.

The author is an Assurance and Advisory professional and can be contacted on +260-976377484.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button

Adblock Detected

Please consider supporting us by disabling your ad blocker