. 49
( 100 .)


a numerical credit scoring system to identify borderline applicants, who are then the
subject of a full-blown detailed credit check. Other applicants are either accepted or
rejected without further question.


The Credit Decision
You have taken the first three steps toward an effective credit operation. In other words,
you have fixed your terms of sale; you have decided whether to sell on open account or
to ask your customers to sign an IOU; and you have established a procedure for esti-
mating the probability that each customer will pay up. Your next step is to decide on
credit policy.
If there is no possibility of repeat orders, the credit decision is relatively simple. Fig-
Standards set to determine
ure 2.9 summarizes your choice. On the one hand, you can refuse credit and pass up the
the amount and nature of
sale. In this case you make neither profit nor loss. The alternative is to offer credit. If
credit to extend to
you offer credit and the customer pays, you benefit by the profit margin on the sale. If
the customer defaults, you lose the cost of the goods delivered.

The decision to offer credit depends on the probability of payment. You
should grant credit if the expected profit from doing so is greater than the
profit from refusing.

Suppose that the probability that the customer will pay up is p. If the customer does
pay, you receive additional revenues (REV) and you deliver goods that you incurred
costs to produce; your net gain is the present value of REV “ COST. Unfortunately, you
can™t be certain that the customer will pay; there is a probability (1 “ p) of default. De-
fault means you receive nothing but still incur the additional costs of the delivered
goods. The expected profit8 from the two sources of action is therefore as follows:
Refuse credit: 0
Grant credit: p PV(REV “ COST) “ (1 “ p) PV(COST)
You should grant credit if the expected profit from doing so is positive.

If you refuse credit, you make REV COST
neither profit nor loss. If you
offer credit, there is a
Customer pays (p)
probability p that the
customer will pay and you
will make REV “ COST; Offer credit
Customer defaults (1 p)
there is a probability (1 “ p)
that the customer will default
and you will lose COST. COST

Refuse credit


8 Notice that we use the present values of costs and revenues. This is because there sometimes are significant
lags between costs incurred and revenues generated. Also, while we follow convention in referring to the “ex-
pected profit” of the decision, it should be clear that our equation for expected profit is in fact the net pres-
ent value of the decision to grant credit. As we emphasized, the manager™s task is to add value, not to maxi-
mize accounting profits.
Credit Management and Collection 237

The Credit Decision
Consider the case of the Cast Iron Company. On each nondelinquent sale Cast Iron re-
ceives revenues with a present value of $1,200 and incurs costs with a present value of
$1,000. Therefore, the company™s expected profit if it offers credit is
p — PV(REV “ COST) “ (1 “ p) — PV(COST) = p — 200 “ (1 “ p) — 1,000
If the probability of collection is 5/6, Cast Iron can expect to break even:
Expected profit = 5/6 — 200 “ (1 “ 5/6) — 1,000 = 0
Thus Cast Iron™s policy should be to grant credit whenever the chances of collection are
better than 5 out of 6.

In this last example, the net present value of granting credit is positive if the proba-
bility of collection exceeds 5/6. In general, this break-even probability can be found by
setting the net present value of granting credit equal to zero and solving for p. It turns
out that the formula for the break-even probability is simply the ratio of the present
value of costs to revenues:
p — PV(REV “ COST) “ (1 “ p) — PV(COST) = 0
Break-even probability of collection, then, is

What is the break-even probability of collection if the present value of the revenues
Self-Test 3
from the sale is $1,100 rather than $1,200? Why does the break-even probability in-
crease? Use your answer to decide whether firms that sell high-profit-margin or low-
margin goods should be more willing to issue credit.

What effect does the possibility of repeat orders have on your credit decision? One of
the reasons for offering credit today is that you may get yourself a good, regular cus-
Cast Iron has been asked to extend credit to a new customer. You can find little in-
formation on the firm and you believe that the probability of payment is no better than
.8. If you grant credit, the expected profit on this order is
Expected profit on initial order = p — PV(REV “ COST) “ (1 “ p) — PV(COST)
= (.8 — 200) “ (.2 — 1,000) = “$40
You decide to refuse credit.
This is the correct decision if there is no chance of a repeat order. But now consider
future periods. If the customer does pay up, there will be a reorder next year. Having
paid once, the customer will seem less of a risk. For this reason, any repeat order is very

Think back to earlier material, and you will recognize that the credit decision bears
many similarities to our earlier discussion of real options. By granting credit now, the
firm retains the option to grant credit on an entire sequence of potentially profitable re-
peat sales. This option can be very valuable and can tilt the decision toward granting
credit. Even a dubious prospect may warrant some initial credit if there is a chance that
it will develop into a profitable steady customer.

Credit Decisions with Repeat Orders
To illustrate, let™s look at an extreme case. Suppose that if a customer pays up on the
first sale, you can be sure you will have a regular and completely reliable customer. In
this case, the value of such a customer is not the profit on one order but an entire stream
of profits from repeat purchases. For example, suppose that the customer will make one
purchase each year from Cast Iron. If the discount rate is 10 percent and the profit on
each order is $200 a year, then the present value of an indefinite stream of business
from a good customer is not $200 but $200/.10 = $2,000. There is a probability p that
Cast Iron will secure a good customer with a value of $2,000. There is a probability of
(1 “ p) that the customer will default, resulting in a loss of $1,000. So, once we recog-
nize the benefits of securing a good and permanent customer, the expected profit from
granting credit is
Expected profit = (p — 2,000) “ (1 “ p) — 1,000
This is positive for any probability of collection above .33. Thus the break-even prob-
ability falls from 5/6 to 1/3.

If one sale may lead to profitable repeat sales, the firm should be inclined to
grant credit on the initial purchase.

How will the break-even probability vary with the discount rate? Try a rate of 20 per-
Self-Test 4
cent in Example 4. What is the intuition behind your answer?

Real-life situations are generally far more complex than our simple examples. Cus-
tomers are not all good or all bad. Many pay late consistently; you get your money, but
it costs more to collect and you lose a few months™ interest. And estimating the proba-
bility that a customer will pay up is far from an exact science.
Like almost all financial decisions, credit allocation involves a strong dose of judg-
ment. Our examples are intended as reminders of the issues involved rather than as
cookbook formulas. Here are the basic things to remember.
1. Maximize profit. As credit manager your job is not to minimize the number of bad
accounts; it is to maximize profits. You are faced with a trade-off. The best that can
happen is that the customer pays promptly; the worst is default. In the one case the
firm receives the full additional revenues from the sale less the additional costs; in
Credit Management and Collection 239

the other it receives nothing and loses the costs. You must weigh the chances of these
alternative outcomes. If the margin of profit is high, you are justified in a liberal
credit policy; if it is low, you cannot afford many bad debts.
2. Concentrate on the dangerous accounts. You should not expend the same effort on
analyzing all credit decisions. If an application is small or clear-cut, your decision
should be largely routine; if it is large or doubtful, you may do better to move
straight to a detailed credit appraisal. Most credit managers don™t make credit deci-
sions on an order-by-order basis. Instead they set a credit limit for each customer.
The sales representative is required to refer the order for approval only if the cus-
tomer exceeds this limit.
3. Look beyond the immediate order. Sometimes it may be worth accepting a relatively
poor risk as long as there is a likelihood that the customer will grow into a regular
and reliable buyer. (This is why credit card companies are eager to sign up college
students even though few students can point to an established credit history.) New
businesses must be prepared to incur more bad debts than established businesses be-
cause they have not yet formed relationships with low-risk customers. This is part of
the cost of building up a good customer list.

Collection Policy
It would be nice if all customers paid their bills by the due date. But they don™t, and,
since you may also “stretch” your payables, you can™t altogether blame them.
Slow payers impose two costs on the firm. First, they require the firm to spend more
resources in collecting payments. They also force the firm to invest more in working
capital. Recall that accounts receivable are proportional to the average collection period
(also known as days™ sales in receivables):
Accounts receivable = daily sales — average collection period
When your customers stretch payables, you end up with a longer collection period
and a greater investment in accounts receivable. Thus you must establish a collection
Procedures to collect and
monitor receivables.
The credit manager keeps a record of payment experiences with each customer. In
addition, the manager monitors overdue payments by drawing up a schedule of the
aging of receivables. The aging schedule classifies accounts receivable by the length of
Classification of accounts
time they are outstanding. This may look roughly like Table 2.11. The table shows that
receivable by time

TABLE 2.11
Customer™s Less than More than
An aging schedule of
Name 1 Month 1“2 Months 2“3 Months 3 Months Total Owed
A $ 10,000 $ 0 $ 0 $ 0 $ 10,000
B 8,000 3,000 0 0 11,000
• • • • • •
• • • • • •
• • • • • •
Z 5,000 4,000 6,000 15,000 30,000
Total $ 200,000 $40,000 $15,000 $ 43,000 $ 298,000

customer A, for example, is fully current: there are no bills outstanding for more than a
month. Customer Z, however, might present problems, as there are $15,000 in bills that
have been outstanding for more than 3 months.
When a customer is in arrears, the usual procedure is to send a statement of account
and to follow this at intervals with increasingly insistent letters, telephone calls, or fax
messages. If none of these has any effect, most companies turn the debt over to a col-
lection agency or an attorney.

Suppose a customer who buys goods on terms 1/10, net 45 always forgoes the cash dis-
Self-Test 5
count and pays on the 45th day after sale. If the firm typically buys $10,000 of goods a
month, spread evenly over the month, what will the aging schedule look like?

There is always a potential conflict of interest between the collection department and
the sales department. Sales representatives commonly complain that they no sooner win
new customers than the collection department frightens them off with threatening let-
ters. The collection manager, on the other hand, bemoans the fact that the sales force is
concerned only with winning orders and does not care whether the goods are subse-
quently paid for. This conflict is another example of the agency problem introduced ear-

Good collection policy balances conflicting goals. The company wants cordial
relations with its customers. It also wants them to pay their bills on time.

There are instances of cooperation between sales managers and the financial man-
agers who worry about collections. For example, the specialty chemicals division of a
major pharmaceutical company actually made a business loan to an important customer
that had been suddenly cut off by its bank. The pharmaceutical company bet that it knew
its customer better than the customer™s bank did”and the pharmaceutical company was
right. The customer arranged alternative bank financing, paid back the pharmaceutical
company, and became an even more loyal customer. It was a nice example of financial
management supporting sales.



. 49
( 100 .)