Allowance for Doubtful (Bad) Debts

Wayne,

Whenever this comes up, it is always a negotiation with the auditors.
1) You need stable history to use the historical method.
2) You need a good risk-analysis system to use the risk method (eg brought on a new risky client, you should weight that differently)
3) You need enough clients to have a normal distribution (thus the note in audits when you have a few, big clients).

My preference is to use History when it exists.
The fallback is Pareto; figure out the point risks and measure those, which works well when there are a few, big clients.
I've never been comfortable with my own risk-analysis systems as being better than history or pareto...they have been more binary; either the customer gets credit or they don't, or there is some limit. In the former, it is black and white. In the latter it feeds better into a Pareto.

Risk systems are great if you are investing in the data to get good output. Insurance, credit cards, etc are the gold-ish standard of "we actually have an actuary looking at this stuff." I'm simply more comfortable with the other methodologies, but that doesn't mean they're better (even when I do have decent data).

Cheers,

KP

Coming from a banking perspective....

Allowance for bad debts should be based on the CONTINUOUS evaluation (ability to pay or probability of default) of EACH and EVERY CLIENT that are past due.

Whether you establish a system of grading/tiers or the assignment of probability of collection/default according to the grading system, it isstill going to be on a PER CLIENT basis.

Furthermore, historical numbers may NOT address present conditions . One may however gleam some perspective as to how clients in a particular grade or tier eventually end up and the assignment a default percentage for this group.

I have always separated the specific reserve from a general reserve. The specific reserve is for those accounts where we see a real risk of payment. For the general reserve we have used the "black motor formula" or an off shoot of this that is based on historical write off percentages.

Patrick, can you explain the need for the general reserve when you have already evaluated all your accounts?

I evaluate ALL customers 90+ days past due to assign a collection risk % from 0% to 100%, plus any customers less than 90+ days past due if I am aware of a particular reason to. That's my specific reserve. To your point, Emerson, I don't particularly believe there is a need for a general reserve at this point, but external auditors just love the two pieces (must be their training); so like Patrick here, I add a general piece, which approximates 3 months worth of my historical write-offs. In the end, the general piece is less than 5% of my total reserve - not worth fighting over on multiple calls with Corporate, then different levels at our external auditor's firm.

A former Big 4 auditor myself...

Givenson, when you were a Big 4 auditor, what is YOUR (or your firm's) justification for the general reserve given that you say external auditors love the two pieces?

As a CFO, I (along with the Credit Team...bank setting) go through the process of evaluating each and every account past due to have a more accurate approximation of our eventual bad debts. Bank examiners (and there are several regulatory bodies) and our external auditors find the process acceptable....and we are in the credit business. We may argue/disagree on the percentages assigned on each classification/grade/tier, but not the process/concept. I find the addition of the "general" reserve to be self defeating (the process) even if it is just 5%. I would even venture to think that they WILL cite (favorably or unfavorably) us if we tack on an extra 5% on top of what the process has recommended.