Counter-Cycle Provisions System. Chapter 3.

Dear community members,

On our last post we commented the main difficulty for building a counter-cycle provisions system is measuring non-visible failed loans.

As we must evaluate non-visible failed loans, the provisions cannot be specific (we don’t know to what loan the provision belongs) but generic, trying to measure the failed loans in the portfolio without having to identify them univocally.

But determining the value of the generic provisions required is a difficult and risky task, as it can create arbitrary postings which affect the Bank result.

The International Financial Reporting Standards require to the Banks create provisions for covering the losses on their portfolio due to events which have already happened and will affect future cash flows.

Part of those losses described by the accounting rules are the non visible failed loans that we described at the beginning.

For determining the losses on those failed loans that we’re aware that exist in the portfolio but we haven’t detected yet, we have to evaluate the whole portfolio and adjust its value globally. That’s the main difference between a Generic and a Specific provision, as in the second we’ve identified the failed loan and it can be evaluated individually.

A very interesting approach for the evaluation of the generic provisions utilizes the IRB Approach of the Credit Risk Calculation (Basel II) as a basis for the determination of the generic provisions.

For the IRB Credit Risk Calculation we have to evaluate the several components; the Probability of Default (PD), the Loss Given Default (LGD), the Exposure at Default (EAD) and the maturity of the contracts (M).

Additionally, the IRB approach let us calculate the expected losses of the portfolio (EL), which is the expected loss for every loan that we can calculate with the following the formula:

EL=PD*LGD*EAD

But we must be careful as the IRB approach definition of Expected Losses is different of the concept of Incurred Losses of the IFRS.

The Expected Losses of the IRB approach is the average flow of losses that the internal rating calculation methods forecasts that is going to materialize in one year, while the Incurred Losses of the IFRS is the stock of existing losses of the portfolio at any given time, due to events in the past which will generate losses in the future.

Both, Incurred Losses and Expected Losses are different from the yearly manifested losses (flow of yearly defaults) and consequently the yearly flow of specific provisions.

Nevertheless, we can calculate the Incurred Losses according to the IFRS by estimating the yearly flow of expected losses and the time from the event which makes the loan failed and the time when the failed loan becomes visible. This period between both events is called Loss Identification Period (LIP).

For instance if the obligor losses his job and as a consequence he will be incapable of fulfilling his payment obligations 18 months later, the Loss Identification Period would be 18 months.

Consequently if we know both magnitudes (the Expected Losses and the Loss Identification Period) we can estimate the Incurred Losses multiplying both.

For example, if the calculated Expected Losses of our portfolio (IRB Approach) are 45 million dollars/year and the average Loss Identification Period is 2 years, that means the Incurred Loss in our portfolio is 90 million dollars.

Incurred Losses (IFRS) = Expected Losses (IRB Approach) * Loss Identification Period

On the expansion phase of the economic cycle the Loss Identification Period is longer due to the easiness for refinancing policies supported but the good economic conditions.

And according to the formula the longer Loss Identification Period will make the Incurred Losses higher during the expansion phase.

So we will get the counter-cycle behavior on the provisions for losses that we are searching.

What do you think?

Cheers.

Ferran