The economics of wholesale credit

Setting profitable credit limits and managing trade receivables risk

Traditionally, corporate trade credit limits have been set based on customer size, an internal orexternal credit score, and a qualitative sense of risk appetite. These limits have been effective inminimizing write-offs, principally because they are conservative.

If robust, more precise, probabilities of default can be obtained, those credit limits can beadjusted to yield higher margins and extend credit where economically justified. The result candrive higher volumes to some existing customers and new sales to customers previously deniedcredit.

To do so, the credit limit setting process must focus on net value added, rather than on lossminimization. One side benefit of the approach is the ability to credit-adjust the pricing of agiven transaction. This way, the cost of credit is explicit and can be added to the requiredmargins.
 

Forgone margins and growth

More than $500 billion dollarsin unsecured trade credit are currently extended in the UnitedStates by wholesale sellers, marketers, or traders. The average is roughly $200 million inAccounts Receivables (AR) per firm. Many of the mid-size to larger marketing entities hold AR books larger than regional banks. Yet they manage these AR positions with the primary goal oflimiting write-offs.

Inevitably, firms turn away profitable customers, and sales to non-investment gradecounterparties are highly restricted. The seller’s return on capital becomes limited, and totalsales and profits suffer, all needlessly. Rethinking how credit limit tables are constructed canalleviate these restrictions for many customers, directly increasing both sales margins andrevenues
 

Traditional risk management of trade receivables

Many wholesale marketer/traders have traditionally measured and monitored credit risk by assigning credit limits to counterparties. They have also tracked limits and usage for individual customers and the portfolio as a whole. Credit limit assignments are made based on pre-approved tables. These tables show that a maximum suggested limit for customers of a given size (sales or net worth) and creditworthiness (internal or external scoring). These tables typically suggest higher limits for larger customers and lower limits for heightened credit risk customers, sometimes allocating zero unsecured credit to non-investment grade counterparties.

This approach abstracts from other potentially useful information, such as customer margins or target returns, achieve high levels of simplicity and usability. Employing such a table lets the firm focus on its primary business model and still maintain a low level of write-offs by setting credit limits conservatively. This strategy is doubly important when internal scoring methods are inconsistent, when there is a lack timely customer information, or when external credit scores are stale. With a history of low defaults or write-offs, sellers become comfortable with a conservative schedule of maximum limits. They can also see how it effectively limits their overall credit risk exposure.
 

An economic approach to limit setting

Credit risk aversion and conservatism are understandable for firms in lines of business other than banking. However, it can be worth evaluating just how much conservatism is inherent in the standard corporate approach to credit limits. The benchmark to assess is the “risk-neutral” economic maximum credit limit, or that limit calculated with an extra risk premium to account for credit risk aversion. The maximum economic credit limits (MECL) are reached when higher limits would not result in any further expected profit (see the Appendix). The intuition behind the economic limits is straightforward. The margin on the trade flow must cover the expected loss due to non-payment. Also, it must cover the cost of capital for being able to withstand that potential loss.

The overall implications of a traditional credit limit table are best seen in a side-by-side comparison of MECL results (Table 1). Here, we use an actual credit limit table used by an industrial wholesaler as an example of a typical structure. The next two columns show the maximum economic credit limits for variously scored customers. We first assume a 3.5% profit margin on sales, a 7% cost of capital, and a 10% excess return on capital as a credit risk aversion premium. The sales level and AR level were chosen so that the Aaa-rated customer limits match those in the industry example.

The last column in Table 1 shows the MECL results, holding all assumptions in the 3.5% margin calculations constant, but increasing the sales margin to 5.5%.

The calculated MECL columns show higher limits for better credit risk customers. That unsecured credit may be denied to some high credit-risk customers, just as we see in the Industry Example limits. But there are also three important differences in the MECL results:

• Limit reduction with credit deterioration. Simplified credit limit structures, such as the Industry Example, reduce the maximum credit limit by 70% for a rating reduction of Aaa to Ba. The MECL approach reduces the maximum limit by less than 5%. The MECL reduction is small, because the default probability of the Aaa-rated customer and the Ba-rated customer are both low (0.002% and 0.9%, respectively) and similar.
• Limit restriction with credit deterioration. The Industry Example limits deny any unsecured credit to potential customers rated B or below. The B credit rating category has a historical default rate of about 3.4% per year. This rate creates the distinct likelihood that some credit losses might be experienced from these buyers. Yet the MECL results show that even a modest margin of 3.5% is enough to make transactions with this customer class profitable.
• Margin effects. The impact of higher customer profitability clearly increases the maximum economic credit limits and makes extending credit to higher risk customers much more viable. The additional profits offset expected losses and capital charges, and can result in significant revenue and profit growth.
   

Optimal limit setting

A maximum credit limit, derived by setting the economic value to zero, is not an optimal limit. However, it does give some excess return on capital driven by the risk-aversion premium. The economically optimal limit is the maximum Account Receivable for that customer, because in this case, no capital costs are wasted on unused limits. But to set the credit limit at the level of the AR, the seller must be able to curtail sales immediately at invoice non-payment. In reality, some level of headroom above the maximum expected AR might be required to allow for late payments and other issues related to the terms and conditions of the transaction.

Credit concentration can also be important in setting trade credit limits, and two kinds of concentration are sometimes considered. First is a revenue concentration limit that prevents becoming dependent on a single customer that is “too large.” These limits are typically stated as a percentage of the seller’s total revenue. Their goal is to prevent seller bankruptcy driven by the loss of a single customer. The second type of concentration limit is normally stated as a percentage of the buyer’s net worth. These limits are intended to ensure that the buyer can afford the level of purchases being considered. 

The MECL calculation can satisfy the economic restriction by requiring the annual transaction margins to be greater than the expected losses and capital charges. Yet transaction volumes can vary across billing cycles in some cases, and the deal pricing can also be volatile. All of these considerations create the need for a reasonable buffer, above the estimated AR level, in setting a customer’s overall trade credit limit.

The ability to set appropriate and economic credit limits can significantly expand the potential customer pool and the margins those transactions generate. But the arithmetic can also be inverted to calculate the margin necessary for a given counterparty and deal size. Often, the marketing team will find this “inverted limit setting” highly useful. For example, a customer might propose a purchase that would create an AR of $40 million, well above its traditional limit. Verifying the economics using the MECL approach might show that the credit cannot be extended when the margin is 3%. However, the deal is attractive and the credit is viable at a 6% margin.
 

Credit-adjusted pricing

Determining the margin required to support sales to a customer with imperfect credit quality is a straightforward extension of the economic credit limit calculation discussed earlier. Here, the margin becomes the output. Inputs include the customer probability of default (PD), the credit limit headroom above the maximum expected AR, and the number of invoices outstanding at the time the customer’s supply is cut off. Figure 1 shows a typical calculation of the additional margin or “Adder” that should be charged to cover the cost of extending unsecured trade credit.

Note, this margin Adder is in excess of standard profit margins. This amount would “pay for” expected credit losses and the capital charge and risk premium required to take on a non-core business risk. The necessary Adder increases as credit quality deteriorates and as the exposure at default (EAD), or number of invoices at default, increases. Interestingly, the additional margin required changes only slightly for counterparties in the Aaa to Ba ratings range (flat portion of the graph). The margin Adder change is minimal in this range because the PDs are so low that the expected losses are nearly immaterial for these ratings. Also, the calculated margin adder is almost solely a function of the cost of capital and the risk premium applied to that cost.
 

Broad implications and caveats

A more rigorous, accurate, and economic credit limit setting process has some major implications for a wholesale seller. The most tangible are increased sales and margins. 

Not every customer or potential buyer will be affected by a change in the maximum credit limit available. In fact, many customers may never become aware of the change. But some buyers source goods from multiple sellers and manage their supply chains to minimize cost and ensure sufficient reliability. Offering more unsecured credit, or initial credit where none was previously available, will be very appealing to that fraction of the customer base. Where Letters of Credit, collateral, credit insurance, guarantees, and other costly forms of trade credit support can be avoided, sales volumes will increase commensurately. Some customers will gladly pay slightly higher prices for that privilege and create measurable margin enhancements as a result. 

The business-process aspects of using a fact-based, analytically driven set of maximum credit limits are universally positive: 

• Credit tables can be used in the same way they are currently, requiring no major changes for marketing or trading teams. 
• Marketing or trading will have the new capability to price deals using “what-if” cases and identify margins that allow sufficient credit to enable the transactions. 
• MECL tables reflect risk premium adjustment changes. Such changes can create compelling conversations within the management team regarding risk and return and the firm’s value proposition to its customers. 
• The quantitative underpinnings of the customer scoring can reduce workloads at the credit-analyst level and enable significant gains in automated data sourcing. 
   

Successfully pursuing this more quantitative approach does require high-quality, timely data to ensure accurate analyses that reflect the current commercial environment.
 

Litmus test

Is this worth pursuing for any given business? The credit-adjusted pricing obviously makes sense for businesses that regularly have some pricing flexibility and power. When to extend new or more trade credit to marginal counterparties is a more subtle decision. Still, some key questions may be useful in looking at the potential business case: 

• Does the current credit limit table show significant attenuation of limits as credit quality falls, even within the “investment grade” range? Customers in that credit range may be a ready source of sales growth under more economic limits. 
• Are historical bad debt write-offs well-below average national default rates (1.5% to 2.5%)? If so, such a scenario could indicate that the seller is screening out an excessive fraction of the market. 
• Are potential customers being turned away or granted only limited sales based on creditworthiness? These actions are the clear cases where a rigorous, data-driven evaluation can add immediate value. 
• Does Sales or Marketing regularly ask for credit limit exemptions? If the standard limit tables are out of date or a poor fit, recalibrating on an economic basis can noticeably expand sales to some buyers. Using economic credit limits will likely not affect most existing customer accounts. However, for the few where it can make a difference, the business case can be highly compelling.
   

A path forward

The principles discussed earlier are straightforward. The calculations can be readily worked out to suit any given set of invoicing and payment terms and conditions (see the Appendix). Legacy credit limit tables are an embedded part of the business processes of many wholesale sellers. They offer an easy, transparent mechanism that enables the marketing team to extend terms and to use a conservative set of credit controls. Without disrupting that workflow, this approach makes it possible to enhance that mechanism to better reflect a diverse and dynamic customer pool. The result is increased sales and/or margins.

The hurdles to expanding the margins and sales growth are typically not technical, but may involve moving beyond entrenched thinking and traditional practices. A way to overcome these hurdles is to try the approach on a portion of the marketing book or on a particular line of business. Figure 2 shows the framework for such a proof of concept program.

Using such an approach, the wholesale credit team can test the tangible enhancement of the firm’s bottom line. The team can still retain and improve the measurement and control of credit risk to the organization. Fundamentally, it is the higher-quality data and analytics applied to a legacy issue that make this change possible.
 

A practical implementation

Sourcing reliable and consistent PDs or EDF™ (Expected Default Frequency) credit measures that span the range of customer types (public, private, municipal, and so on) is the first key challenge. These EDF measures can be used directly to calculate maximum limits or the required minimum margins. Moody’s RiskCalc TM model offers current and, sometimes, early warning, EDF information for most of the corporate buyer portfolios. Within the existing workflow, it may be useful to also reproduce Credit Limit Tables, as shown in Figure 3. While this example employs a single EDF measure corresponding to an internal or external rating, it still reflects the maximum limits that can be extended to a customer, given that transaction’s margins.

The limits shown in this table decline with decreasing customer credit quality, but the decline is subtle for ratings in the “Investment Grade” range. This trend shows that the cost of offering trade credit to these buyers is primarily driven by the capital charges for the limit, not the expected losses. A second, potentially more useful way to manage trade credit issues for individual transactions is to examine the margin required to pay for the unsecured credit. Figure 4 shows one way to explore these creditadjusted margins.

With either approach, consistent and precise EDF credit measures for all counterparties remain the foundation for a trade credit management program. Such a program fully realizes the potential sales volumes and margins that may have been hidden by more traditional methodologies.
 

Appendix: Calculation details and example

The Maximum Economic Credit Limit (MECL) can be determined by first calculating the net economic value of the transaction and then setting this value to zero. This is the point at which the profits created are offset by the expected default losses and charges for use of the firm’s capital: 

0=Margin−E(Loss)−Capital Charge−Risk Premium 

0=MIF−EIN−RC−PC or C=I(MF−EN)/(R+P) 
 

C = Credit Limit ($) 

M = Margin on sales to this customer (%) 

I = Invoice amount expected for this customer ($) 

F = Frequency, expressed as the number of invoices per year (#) 

E = Expected Default Frequency (%) 

N = Number of outstanding Invoices at default (#) 

R = Rate of return on capital (%) 

P = Premium for undertaking credit risk (%)

For example: A 4% margin customer with an EDF value of 3.31% (~B rating) has an annual purchase of $360 million on 30-day invoicing terms ($30 million per month). This customer is likely to hold two invoices outstanding at default. If the firm’s cost of capital is 7% and its credit risk premium is 10%, the maximum economic credit limit is: 

0 = 4% x $30M x 12 – 3.31% x $30 x 2 – 7% x C - 10% x C = $30M x (0.04 x 12 – 0.0331 x 2) – (0.07 + 0.10) x C 

C = $30M x 0.414 / 0.17 = $73M

The customer in this example requires $60 million in credit to support two unpaid invoices of $30 million each. Note that the maximum credit limit calculated is not the optimal limit, as shown in Figure 5. This graph shows the impact of increasing credit limits when the buyer requires only a certain amount of credit. The annual Margin increases until the customer is buying all it needs. That line assumes that the customer is paying invoices on time. The optimal economic credit limit would require immediate payment of invoices, which minimizes the sellers exposure to credit losses.

The exposure to loss at the time of default is a critical parameter is in choosing an economic credit limit. It is determined by the invoicing cycle, terms of payment, and how quickly the sales can be terminated for nonpayment. In practice, these can vary across customers or products