‘New rating scale quantifies expected loss over life of debt instrument’

In February, credit rating agency Crisil announced its first rating of an infrastructure project under the new ‘expected loss’ (EL) scale for Purulia and Kharagpur Transmission Co Ltd, giving the Sterlite Power group company an Infra EL 1. In an interview with BusinessLine, Pawan Agrawal, Chief Analytical Officer, Crisil Ratings, explains the rationale for the new methodology and how both lenders and borrowers stand to benefit. Excerpts:

Can you give a background on how the new rating scale was created?

The new EL rating system was created after a consultation between the Ministry of Finance and rating agencies. The question was whether a new scale could be developed that would encompass the uniqueness of infrastructure projects. We’d been looking at infra projects, especially the PPP ones; there was stability of cash flows, where there was long-term visibility, and you have strong counter-parties in some cases, and we noticed that you don’t see NPAs here, even though the projects may not be paying on time.

When we looked at this in some more detail, we realised that the key differentiator is recovery. Recovery in India is constrained by weak bankruptcy laws and the legal system. In India, a default is a proxy for bankruptcy and this means there will be a long recovery process and (as a lender) you are uncertain around the time lines and you can’t factor that into your cash flows in any manner.

With the projects that we were looking at, the difference largely was that despite the legal system, these projects would continue to generate cash. Like toll roads, some times collections may be a little less than the projected or debt payments, but it doesn’t mean that there will be zero cash on a particular day. So similarly, if you have a windmill, a windmill in season will generate power and that will be supplied to the grid and eventually, with a lag, cash will come in. So the question we saw really here was if an investor was not worried about the time-line of payment and assuming that there is a delay, am I getting cash later to cure this delay? In the operational PPP infra projects, there’s a reasonable recovery that happens which, in our estimate, runs to 70-90 per cent.

How was the scale created?

In this working group, we had lenders like SBI, IIFCL, private financiers and the rating agencies provided data on the track record of cash flows coming in and in case of default, what were the chances of money coming in later.

In the last one year, we were able to validate our understanding with data and as a result, we were able to come up with a scale that was based on more than just our gut feeling.

The new scale quantifies expected loss over the life of the debt instrument and if the instrument is structured over the life of the project, it quantifies how much loss is possible at every stage of the instrument at every grade. Here for one set of loans to operational infra projects, we are able to see recovery trends that are different, and therefore, can factor that in and assign ratings on this scale.

Can you give more details on how to understand the new scale?

The new credit rating system is based on the ‘expected loss’ (EL) methodology. Which means, the rating will be an expert judge on EL over the life of the debt instrument by taking into account the two pillars of credit risk – the probability of default and the loss given default.

The ratings will be assigned on a scale from Infra EL1 to EL7, with EL1 having the lowest expected loss. With EL1, for instance, the indicative loss range is 0-1.25 per cent over the life of the instrument. Not all BBBs exhibit the same expected loss, but the ones with good recovery characteristics can be assigned an EL1 or EL2. This is an additional dimension to the existing scale.

Besides infrastructure, what are the other sectors this scale is suitable for?

The methodology can work for others too, but right now, we have data and recovery confidence only on infrastructure projects, particularly operational infra projects. And for may be under-constructions renewable projects with shorter construction periods.

For the rest, it should over a period of time when the bankruptcy law becomes more effective and you have more certainty on the recovery numbers and the loss given default.

How does this change pricing for borrowers?

Lenders now know the scale and the range of expected loss associated with it. So as a lender, once I build confidence on this scale,

I can use this expected loss as an input in my pricing. If there’s an EL1 rating, the range is 0-1.25 per cent possible loss on a 10-year instrument, then every year on an average my loss expectation is 1.25 per cent/10. Of course, there will be other factors like liquidity and credit, and these risks remain. But with this expected loss and the risk premium charged so far for this loss, lenders can see their price coming down.

We’re in an environment where banks are still risk-averse and are unwilling to fund riskier projects. So will the new methodology help?

What needs to be seen is if the financial ecosystem can build comfort with this scale. It may not happen overnight but you can see performance over time.

And when you look at the underlying, and look at any bank, look at the difference between NPAs in roads versus other sectors.

The defaults might be there in roads, but you won’t see the NPAs being very high and that is a very simple validation of this concept. And therefore, I think eventually they will build comfort lending on the new scale.

Also, I think for borrowers, their ability to go to the bond market increases when the ratings increase over a threshold with which the bond markets are comfortable.

COMMENT NOW