Anderson Sweeney Williams Camm Cochran Fry Ohlmann

An Introduction to
Management Science, 15e
Quantitative Approaches to Decision Making

© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Chapter 16: Markov Processes

16.1 – Market Share Analysis

16.2 – Accounts Receivable Analysis

2
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Markov Processes (1 of 2)

Markov process models are useful in studying the evolution of

systems over repeated trials or sequential time periods or
stages.
Markov process models can be used to describe:
• the probability that a machine that is functioning in one period
will continue to function or will break down in the next period.
• the probability that a consumer purchasing brand A in one
period will purchase brand B in the next period.
• the health status probabilities for persons aged 65 and older.
Such information was helpful in understanding the future need
for health care services and the benefits of expanding current
health care programs.

3
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Markov Processes (2 of 2)

An in-depth treatment of Markov processes is beyond the scope

of this text, so the analysis in this chapter is restricted to
situations consisting of a finite number of states, the transition
probabilities remaining constant over time, and the probability of
being in a particular state at any one time period depending only
on the state in the immediately preceding time period.

Such Markov processes are referred to as Markov chains with

stationary transition probabilities.

4
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (1 of 12)

Suppose we are interested in analyzing the market share and

customer loyalty for Murphy’s Foodliner and Ashley’s
Supermarket, the only two grocery stores in a small town. We
focus on the sequence of shopping trips of one customer and
assume that the customer makes one shopping trip each week to
either Murphy’s Foodliner or Ashley’s Supermarket, but not both.

5
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (2 of 12)

We refer to the weekly periods or shopping trips as the trials of

the process. Thus, at each trial, the customer will shop at
either Murphy’s Foodliner or Ashley’s Supermarket. The
particular store selected in a given week is referred to as the
state of the system in that period. Because the customer has
two shopping alternatives at each trial, we say the system has
two states.

State 1. The customer shops at Murphy’s Foodliner.

State 2. The customer shops at Ashley’s Supermarket.

6
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (3 of 12)

Suppose that, as part of a market research study, we collect

data from 100 shoppers over a 10-week period. In reviewing
the data, suppose that we find that of all customers who
shopped at Murphy’s in a given week, 90% shopped at
Murphy’s the following week while 10% switched to Ashley’s.
Suppose that similar data for the customers who shopped at
Ashley’s in a given week show that 80% shopped at Ashley’s
the following week while 20% switched to Murphy’s.

7
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Transition Probabilities (1 of 2)

Transition probabilities govern the manner in which the state

of the system changes from one stage to the next.

These are often represented in a transition matrix.

8
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Transition Probabilities (2 of 2)

A system has a finite Markov chain with stationary transition

probabilities if:
• there are a finite number of states,
• the transition probabilities remain constant from stage to
stage, and
• the probability of the process being in a particular state at
stage n + 1 is completely determined by the state of the
process at stage n (and not the state at stage n – 1).
This is referred to as the memory-less property.

9
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (4 of 12)

Transition Probabilities
Next Weekly Shopping Period
Current Weekly
Murphy’s Foodliner Ashley’s Supermarket
Shopping Period

Murphy’s Foodliner 0.9 0.1

Ashley’s Supermarket 0.2 0.8

p ij
 probability of making a transition from state i
in a given period to state j in the next period

p p 12  0.9 0.1
P   11  
 p 21 p 22   0.2 0.8 

10
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (5 of 12)

State Probabilities

11
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (6 of 12)
State Probabilities for Future Periods Beginning Initially with a
Murphy’s Customer
Period (n)
State
0 1 2 3 4 5 6 7 8 9 10
Probability

π1(n) 1 0.9 0.83 0.781 0.747 0.723 0.706 0.694 0.686 0.680 0.676

π2(n) 0 0.1 0.17 0.219 0.253 0.277 0.294 0.306 0.314 0.320 0.324

State Probabilities for Future Periods Beginning Initially with an

Ashley’s Customer
Period (n)
State
0 1 2 3 4 5 6 7 8 9 10
Probability

π1(n) 0 0.2 0.34 0.438 0.507 0.555 0.589 0.612 0.628 0.640 0.648

π2(n) 1 0.8 0.66 0.562 0.493 0.445 0.411 0.388 0.372 0.360 0.352

12
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Steady-State Probabilities (1 of 3)

• The state probabilities at any stage of the process can be

recursively calculated by multiplying the initial state
probabilities by the state of the process at stage n.
• The probability of the system being in a particular state after
a large number of stages is called a steady-state probability.

13
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Steady-State Probabilities (2 of 3)

Steady state probabilities can be found by solving the system

of equations P   together with the condition for
probabilities that  i  1.

• Matrix P is the transition probability matrix

• Vector  is the vector of steady state probabilities.

14
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (7 of 12)

Steady-State Probabilities

Let  1  long run proportion of Murphy's visits

 2  long run proportion of Ashley's visits

Then,

0.9 0.1
 1  2  0.2 0.8   1  2


15
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (8 of 12)

Steady-State Probabilities
0.9 1  0.2 2   1 (1)
0.1 1  0.8 2   2 (2)
1   2  1 (3)

Substitute  2  1  1 into (1) to give:

 1  0.9 1  0.2(1   1 )  2 / 3  0.667

Substituting back into (3) gives:
 2  1/ 3  0.333

16
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Steady-State Probabilities (3 of 3)

Steady-State Probabilities
Thus, if we have 1000 customers in the system, the Markov
process model tells us that in the long run, with steady-state
probabilities  1  0.667 and  2  0.333, 667 customers
will be Murphy’s and 333 customers will be Ashley’s.

17
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (9 of 12)

Suppose Ashley’s Supermarket is contemplating an

advertising campaign to attract more of Murphy’s
customers to its store. Let us suppose further that
Ashley’s believes this promotional strategy will increase
the probability of a Murphy’s customer switching to
Ashley’s from 0.10 to 0.15.

18
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (10 of 12)

Revised Transition Probabilities

Next Weekly Shopping Period

Current Weekly
Murphy’s Foodliner Ashley’s Supermarket
Shopping Period
Murphy’s Foodliner 0.85 0.15

Ashley’s Supermarket 0.20 0.80

19
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (11 of 12)

Revised Steady-State Probabilities

0.85 1  0.20 2   1 (1)
0.15 1  0.80 2   2 (2)
1   2  1 (3)

Substitute  2  1  1 into (1) to give:

 1  0.85 1  0.20(1   1 )  0.57

Substituting back into (3) gives:
 2  0.43

20
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Example: Market Share Analysis (12 of 12)

Suppose that the total market consists of 6000 customers

per week. The new promotional strategy will increase the
number of customers doing their weekly shopping at
Ashley’s from 2000 to 2580.

If the average weekly profit per customer is $10, the

proposed promotional strategy can be expected to
increase Ashley’s profits by $5800 per week. If the cost of
the promotional campaign is less than $5800 per week,
Ashley should consider implementing the strategy.

21
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (1 of 25)

An accounting application in which Markov processes have

produced useful results involves the estimation of the
allowance for doubtful accounts receivable.

This allowance is an estimate of the amount of accounts

receivable that will ultimately prove to be uncollectible (i.e.,
bad debts).

22
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (2 of 25)

Let us consider the accounts receivable situation for

Heidman’s Department Store. Heidman’s uses two aging
categories for its accounts receivable:
(1)accounts that are classified as 0–30 days old, and
(2)accounts that are classified as 31–90 days old.

If any portion of an account balance exceeds 90 days, that

portion is written off as a bad debt. Heidman’s follows the
procedure of aging the total balance in any customer’s
account according to the oldest unpaid bill.

23
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (3 of 25)

For example, suppose that one customer’s account balance on

September 30 is as follows:

Date of Purchase Amount Charged

August 15 $25

September 18 10

September 28 50

Total $85

An aging of accounts receivable on September 30 would assign

the total balance of $85 to the 31–90-day category because the
oldest unpaid bill of August 15 is 46 days old.

24
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (4 of 25)
Date of Purchase Amount Charged
August 15 $25
September 18 10
September 28 50
Total $85

Let us assume that one week later, October 7, the customer

pays the August 15 bill of $25. The remaining total balance of
$60 would now be placed in the 0–30-day category because
the oldest unpaid amount, corresponding to the September 18
purchase, is less than 31 days old.
This method of aging accounts receivable is called the total
balance method because the total account balance is placed in
the age category corresponding to the oldest unpaid amount.

25
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (5 of 25)

Let us assume that on December 31 Heidman’s shows a total

of $3000 in its accounts receivable and that the firm’s
management would like an estimate of how much of the $3000
will eventually be collected and how much will eventually result
in bad debts.

The estimated amount of bad debts will appear as an allowance

for doubtful accounts in the year-end financial statements.

26
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (6 of 25)

First, concentrate on what happens to one dollar currently in

accounts receivable. As the firm continues to operate into the
future, we can consider each week as a trial of a Markov process
with a dollar existing in one of the following states of the system:

State 1. Paid category

State 2. Bad debt category
State 3. 0–30 day category
State 4. 31–90 day category

Thus, we can track the week-by-week status of one dollar by

using a Markov analysis to identify the state of the system at a
particular week or period.

27
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (7 of 25)

Using a Markov process model with the preceding states, we

define the transition probabilities as follows:
pij = probability of a dollar in state i in one week moving to state
j in the next week
Based on historical transitions of accounts receivable dollars, the
following matrix of transition probabilities, P, has been developed
for Heidman’s Department Store:
 p11 p12 p13 p14  1.0 0.0 0.0 0.0 
p p p p  0.0 1.0 0.0 0.0 
P   21 22 23 24 
 
 p31 p32 p33 p34  0.4 0.0 0.3 0.3
   
p
 41 p 42 p 43 p 44  0.4 0.2 0.3 0.1

28
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (8 of 25)

 p11 p12 p13 p14  1.0 0.0 0.0 0.0 

p p22 p23 p24  0.0 1.0 0.0 0.0 
P   21 
 p31 p32 p33 p34  0.4 0.0 0.3 0.3
   
 p41 p42 p43 p44  0.4 0.2 0.3 0.1
Note that the probability of a dollar in the 0–30-day category
(state 3) moving to the paid category (state 1) in the next period
is 0.4.
Also, this dollar has a 0.3 probability that it will remain in the 0–
30-day category (state 3) one week later, and a 0.3 probability
that it will be in the 31–90-day category (state 4) one week later.
Note also that a dollar in a 0–30-day account cannot make the
transition to a bad debt (state 2) in one week.
29
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (9 of 25)

An important property of the Markov process model for

Heidman’s accounts receivable situation is the presence of
absorbing states.

For example, once a dollar makes a transition to state 1, the paid

state, the probability of making a transition to any other state is
zero. Similarly, once a dollar is in state 2, the bad debt state, the
probability of a transition to any other state is zero. Thus, once a
dollar reaches state 1 or state 2, the system will remain in this
state forever.

We can conclude that all accounts receivable dollars will

eventually be absorbed into either the paid or the bad debt state,
and hence the name absorbing state.
30
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (10 of 25)

Whenever a Markov process has absorbing states, we do not

compute steady-state probabilities because each unit ultimately
ends up in one of the absorbing states. With absorbing states
present, we are interested in knowing the probability that a unit
will end up in each of the absorbing states.

For the Heidman’s Department Store problem, we want to know

the probability that a dollar currently in the 0–30-day age
category will end up paid (absorbing state 1) as well as the
probability that a dollar in this age category will end up a bad
debt (absorbing state 2). We also want to know these absorbing-
state probabilities for a dollar currently in the 31–90-day age
category.

31
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (11 of 25)

The computation of the absorbing-state probabilities requires the

determination and use of what is called a fundamental matrix.

The mathematical logic underlying the fundamental matrix is

beyond the scope of this text. However, as we show, the
fundamental matrix is derived from the matrix of transition
probabilities and is relatively easy to compute for Markov
processes with a small number of states.

32
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (12 of 25)

We begin the computations by partitioning the matrix of

transition probabilities into the following four parts:

1.0 0.0 0.0 0.0  1.0 0.0 0.0 0.0 

0.0 1.0 0.0 0.0  0.0 1.0 0.0 0.0 
P 
0.4 0.0 0.3 0.3  R Q 
   
0.4 0.2 0.3 0.1  

where
0.4 0.0  0.3 0.3
R  Q 
 0.4 0.2   0.3 0.1

33
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (13 of 25)

A matrix N, called a fundamental matrix, can be calculated

using the following formula:

N  I  Q
1

where I is an identity matrix with 1s on the main diagonal and

0s elsewhere. The superscript −1 is used to indicate the inverse
of the matrix (I – Q).

34
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (14 of 25)

The identity matrix I must be chosen such that it has the same
size or dimensionality as the matrix Q. In our example problem,
Q has two rows and two columns, so we must choose

1.0 0.0 
I  
 0.0 1.0 
Let us now continue with the example problem by computing
the fundamental matrix

1.0 0.0  0.3 0.3  0.7 0.3

I Q       
 0.0 1.0   0.3 0.1  0.3 0.9 

35
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (15 of 25)

Next, we calculate the inverse of (I – Q) using Appendix 16.1.

1.67 0.56 
N  I  Q
1
 
 0.56 1.30 
If we multiply the fundamental matrix N times the R portion of
the P matrix, we obtain the probabilities that accounts
receivable dollars initially in state 3 or 4 will eventually reach
each of the absorbing states. The multiplication of N times R for
the Heidman’s Department Store gives:

1.67 0.56  0.4 0.0  0.89 0.11

NR       
 0.56 1.30  0.4 0.2   0.74 0.26 

36
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (16 of 25)

1.67 0.56  0.4 0.0  0.89 0.11

NR       
 0.56 1.30  0.4 0.2   0.74 0.26 
• The first row of the product NR is the probability that a dollar
in the 0–30-day age category will end up in each absorbing
state.
• We see a 0.89 probability that a dollar in the 0–30-day
category will eventually be paid and a 0.11 probability that it
will become a bad debt.
• The second row shows the probabilities associated with a
dollar in the 31–90-day category; that is, a dollar in the 31–
90-day category has a 0.74 probability of eventually being
paid and a 0.26 probability of proving to be uncollectible.

37
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (17 of 25)

Let B represent a two-element vector that contains the

current accounts receivable balances in the 0–30-day
and the 31–90-day categories; that is,

B  b1 b2 

Total dollars in the Total dollars in the

0–30-day category 0–30-day category

38
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (18 of 25)

Suppose that the December 31 balance of accounts receivable

for Heidman’s shows $1000 in the 0–30-day category (state 3)
and $2000 in the 31–90-day category (state 4).

B = [1000 2000]

We can multiply B times NR to determine how much of the

$3000 will be collected and how much will be lost.

0.89 0.11
BNR  1000 2000    2370 630
0.74 0.26 

39
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (19 of 25)

0.89 0.11
BNR  1000 2000    2370 630
0.74 0.26 

We see that $2370 of the accounts receivable balances will be

collected and $630 will be written off as a bad debt expense.

Based on this analysis, the accounting department would set up

an allowance for doubtful accounts of $630.

40
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (20 of 25)

The matrix multiplication of BNR is simply a convenient way of

computing the eventual collections and bad debts of the
accounts receivable.

Recall that the NR matrix showed a 0.89 probability of collecting

dollars in the 0–30-day category and a 0.74 probability of
collecting dollars in the 31–90-day category.

Thus, as was shown by the BNR calculation, we expect to

collect a total of (1000)0.89 + (2000)0.74 = 890 + 1480 = $2370.

41
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (21 of 25)

Suppose that on the basis of the previous analysis Heidman’s

would like to investigate the possibility of reducing the amount of
bad debts.

Recall that the analysis indicated that a 0.11 probability or 11% of

the amount in the 0–30-day age category and 26% of the
amount in the 31–90-day age category will prove to be
uncollectible.

Let us assume that Heidman’s is considering instituting a new

credit policy involving a discount for prompt payment.

42
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (22 of 25)

Management believes that the policy under consideration will

increase the probability of a transition from the 0–30-day age
category to the paid category and decrease the probability of a
transition from the 0–30-day to the 31–90-day age category. Let
us assume that a careful study of the effects of this new policy
leads management to conclude that the following transition
matrix would be applicable:

1.0 0.0 0.0 0.0 

0.0 1.0 0.0 0.0 
P
0.6 0.0 0.3 0.1
 
0.4 0.2 0.3 0.1

43
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (23 of 25)

We see that the probability of a dollar in the 0–30-day age

category making a transition to the paid category in the next
period has increased to 0.6 and that the probability of a dollar
in the 0–30-day age category making a transition to the 31–90-
day category has decreased to 0.1. To determine the effect of
these changes on bad debt expense, we must calculate N, NR,
and BNR.

1.5 0.17  0.6 0.0  0.97 0.03

NR       
 0.5 1.17  0.4 0.2   0.77 0.23 

44
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (24 of 25)

1.5 0.17  0.6 0.0   0.97 0.03

NR       
 0.5 1.17  0.4 0.2   0.77 0.23 

We see that with the new credit policy we would expect only 3%
of the funds in the 0–30-day age category and 23% of the funds
in the 31–90-day age category to prove to be uncollectible.

45
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 Accounts Receivable Analysis (25 of 25)

If, as before, we assume a current balance of $1000 in the 0–

30-day age category and $2000 in the 31–90-day age category,
we can calculate the total amount of accounts receivable that
will end up in the two absorbing states by multiplying B times
NR.
0.97 0.03
BNR  1000 2000    2510 490
0.77 0.23
Thus, the new credit policy shows a bad debt expense of $490.
Under the previous credit policy, we found the bad debt
expense to be $630. Thus, a savings of $630 – $490 = $140
could be expected as a result of the new credit policy. Given the
total accounts receivable balance of $3000, this savings
represents a 4.7% reduction in bad debt expense.

46
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.
 End Presentation: Chapter 16

47
© 2019 Cengage. All rights reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part, except for use as permitted in a
license distributed with a certain product or service or otherwise on a password-protected website or school-approved learning management system for classroom use.