Hi,
This is one of my responsibilities, and I made the changes. We always get lots of customer queries when it the coverage ratio trends downwards, so thought we'd at least write something if it was going to shift - in this case upwards.
There are two aspects to the expected loss calculation, (a) the monthly implementation of the output variables from the calculation, and (b) the methodology for the calculation.
The monthly (a) often causes small shifts as the application and default cohorts used in the calculation move forwards. Without being specific, these cohorts have to be old enough to have a fair representation of expected bad rates but new enough to be representative of current lending policy and economic conditions. Whilst always trying to eliminate seasonality. These variable updates are re-calculated monthly and implemented by myself as this cohort moves and new loans are used to determine the variables. Then the expected loss and coverage ratio on the web site is real-time automated, depending on what these variables for expected loss have been determined in in the monthly process. I don't want to get specific and we are not a bank, but those familiar with Basel II will have a vague idea of the ratings process we have implemented.
The methodology (b) to determine the variables rarely changes and in fact has only changed marginally 3 times in 18 months. We are moving from an environment where we had limited application and bad debt data to volumes that are statistically robust - and therefore able to change the methodology to be based on the dynamics of our portfolio as opposed to my experience. The original methodology as a result was very (some that have been through the data with me might say "ridiculously") prudent, which we have maintained but are becoming more data driven.
The recent change into methodology took into account the fact that mature loans default with lower average balances, which although obvious because the loan is mature and has amortised down, we now have more robust data that allows us to statistically determine exactly how much smaller. The methodology is still very prudent on a number of points, an aspect I am often challenged on internally and knowledgeable clients that visit. So I'm sure this will change again in the future.
I would however say that the expected loss and subsequent coverage ratio is only an estimate, based on one methodology using rear-view mirror data. The future could be worse or better, it's impossible to know. And therefore although internally we create a bunch of "scenarios" on where we might be in 5 years time (not stress testing, this is a load of b*****), we are essentially always aiming to grow the size of the fund and it's ratio to the fund under management. The next downturn will be upon us....
Lenders shouldn't get too fixated with the coverage ratio, I prefer to focus on the absolute monetary size. And those stressed or not sleeping, perhaps I should advise that you have miss-understood your personal risk appetite and you should invest in 100% safe and non-volatile returns. I personally don't want lenders that are uncomfortable with the low risk, but not no risk, offering that RateSetter provides.
westonkevRSlink.ratesetter.com/8Ls46js
www.linkedin.com/profile/view?id=19236219