Thursday, December 9, 2010

More Thoughts on Risk

In my last post I suggested that that the pricing of receivables purchases had to cover three things:

1) The Buyer’s cost of funds,

2) The Buyer’s operating costs, and

3) An appropriate return for the risk assumed.

The issue of appropriate pricing for risk assumed is also, in my opinion, a function of three principal elements. It’s been a while since I addressed that topic but current bidding dynamics make me think it’s time to at least remind MYSELF of those elements.

The CAPACITY of the Account Debtor to pay is one of the three elements of pricing risk. It is an important one; but it is only one.

It appears that some Buyers are being persuaded in their bidding and pricing decisions by this element only and that, I think, can be perilous.

The fact that an auction includes invoices presumably payable by a Debtor with an unquestioned CAPACITY to pay loses value if that Debtor either does not ACTUALLY have an obligation to pay or, almost as importantly, does not BELIEVE it has the obligation to pay.

The two additional elements critical to the pricing decision are:

a) The CAPACITY of the SELLER to pay if the Debtor does not, and

b) The character of the RELATIONSHIP between the Seller and the Debtor.

Under the terms of a Seller’s agreement with the Exchange, the Seller is obligated to pay if the Debtor does not. If it does not appear that the Seller has the capacity to make good on the obligation, it is clear that the level of risk is elevated. But the ability of a Buyer to assess the Seller’s capacity is impaired by the Buyer’s inability, under Exchange rules, to have direct contact with the Seller.

The Buyer, therefore, is limited in its assessment of the Seller’s capacity to an analysis of information available either via the Exchange or via publicly-available credit and other data.

TRE is doing a much better job these days at getting timely updates of financial statements from Sellers, but it remains the case that most Seller statements are internally-generated without any third-party review. They cannot be treated as if they were audited or even independently reviewed. And TRE itself makes no representations regarding the accuracy of the financial information provided.

Since 99% of TRE Sellers are privately-owned, the value of information available from credit rating/reporting agencies is also limited. Most often the data available from those sources relates to a Seller’s payments of typical operating expenses. Do they pay their utility, credit card and equipment leasing bills, for example?

That sort of information has SOME value but it does not really address the capacity of the Seller to make good on an unpaid invoice; especially if that invoice is very large in the context of the Seller’s resources.

Many Sellers are in businesses that require very little equity capital to either establish or operate. Many have little or no net working capital and little or even negative equity.

That doesn’t mean, necessarily, that they could not access the funds needed to make good on a re-purchase obligation. But where the Seller’s capacity to re-purchase is not beyond question (and certainly where that capacity is clearly questionable) there is an element of Seller risk that SHOULD be compensated regardless of the payment capacity of the Debtor.

Why is that?

Because there IS a risk, no matter who the Debtor is, that the Debtor is NOT actually liable to pay. And there is a risk that even if it IS liable to pay, it will ASSERT that it is not, in which case there might be a difficult and potentially expensive fight to prove liability and to collect.

That brings in the third issue that has to be considered in pricing transaction risk i.e. the relationship between the Seller and the Debtor.

The “relationship” issue I’m referring to here has two elements:

a) The contractual relationship regarding the obligation to pay a specified invoice, and

b) The history between the Seller and the Debtor specifically involving the provision of and payment for similar goods or services.

I’ve written in the past about the fact that the TRE model is not a “full notification” model. With few exceptions the Exchange does NOT receive assurance from the Debtor that the goods or services represented by the invoice have actually been provided in a manner that meets the contractual requirements for payment.

The Exchange DOES verify that an invoice exists in the Debtor’s AP system that matches the one posted for sale; however, after a pattern is established between Seller and Debtor, that verification process includes a statistical sampling element allowing for less than 100% of invoices to be verified.

There are reasons that the TRE process falls short of a “full notification” standard and those reasons are easily understood.

However, it’s just a fact that sometimes errors are made in billing or that goods are damaged in transit or that goods or services fail to meet specification.

Challenges to payment liability can arise for a variety of reasons that legitimately cause actual payments to differ from invoiced amounts. The “verification” process is not likely to bring these to the surface.

The full notification process is, itself, not without risk or flaw, but it does provide a significantly stronger assurance to a buyer of invoices than does the verification alternative.

That being said, it’s fair to acknowledge there are many Sellers with longstanding relationships with Debtors whose goods or services carry a relatively low risk of failing to meet specification.

For example: a staffing company that provides relatively unskilled personnel to a client. The question is: did employee Smith work 40 hours at a given hourly rate during a certain week?

If employee Smith is moving boxes from a truck to a warehouse the question of whether he properly performed his task will be an easy one to answer. And, if the documentation provided with that invoice includes a countersigned time card on which the Debtor verifies that Smith actually did work 40 hours moving boxes, there is relatively little risk that the Debtor will subsequently challenge the invoice.

But if a staffing firm provides a skilled technician in a research lab, or a computer programmer whose work has to properly interface with that of a dozen others, for instance, a failure to perform the tasks properly might not be immediately apparent.

So the character of the goods or services provided does (or should) affect the assessment of risk. It's part of the "relationship" question.

Now, if the Seller has posted and the Debtor has paid 25 similar invoices for similar services in the past, then the assessment of risk is further mitigated.

But the converse is also true: the more limited the experience and the more specialized the goods or services, the greater the likelihood of disputed invoices.

Why do I come back to some issues I’ve written about in the past?

Because it is my perception that there are Buyers who are pricing transactions today as if there is essentially neither Seller risk nor relationship risk. If the Debtor is a well-known and financially sound company it appears that some Buyers will pay nearly any price asked by a Seller.

And because it appears to me that some Buyers are not even looking at the back-up documentation made available by the Exchange either on the Sellers or their invoices. The auctions are selling too quickly in many cases for a Buyer to have even calculated the likely duration of the transaction.

I suspect that there will be those who will disagree with my suggestion that risk is being mispriced in some current and recent TRE auctions.

They might argue that the pricing does not reflect a lack of attention to risk but rather the ability to access extremely low-cost funds.

I actually wish that were true. But I don’t believe it is.

And the process of re-pricing to more rational levels can be difficult for all involved.