Formal models in Budgeting and Incentive scheme part 2

Formal models in Budgeting and Incentive scheme part 2 - Transcript

FORMAL MODELS IN BUDGETING AND INCENTIVE CONTRACTS PART 2
CHAPTER 14 ADVANCED MANAGEMENT ACCOUNTING KAPLAN AND ATKINSON 3E

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS

1







Decentralized organizations allow managers to develop specialist information Information is simultaneously used in planning and also to evaluate the managers subsequent performance Central planners of the Soviet Union devised and apparently implemented a system of bonuses that rewarded both accurate forecasts and outstanding performances

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 2







The scheme provided penalties for managers who set output targets so low that the budget was easily achieved It provides incentive for managers after achieving the budgeted performance to put out even more effort to exceed the budget The scheme motivated people to disclose their information truthfully in the process of target setting and then once the target was set to work to achieve or better the target

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 3

Forecast incentive scheme Top management establishes A basic bonus pool Bo Three positive parameters m n o read alpha beta and gamma Manager first declare a budgeted output level Yh Which increases the bonus pool by nYh

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 4







This factor provides managers with an incentive to declare a higher rather than a lower budgeted output level If the actual output level Y exceeds the budgeted level Yh an additional bonus of m Y Yh is paid This bonus component motivates the manager to exceed the budgeted performance after the budget target has been established

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 5







m is set less than n so that output is going to be high it is better to declare this fact in the budget than to realize it by exceeding the budget If the actual output level y is less than the budget a penalty of o Yh Y is subtracted from the bonus In this case o is set higher than n so that manager receive no benefit from inflating the budget only to be disappointed later So incentive properties of this mechanism require that 0 m n o

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 6



B Bo nYh m Y Yh if Y Yh Bo nYh o Yh Y if Y Yh

Finally if B is the actual bonus paid to the manager the plan can be described by

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 7

Example Display bonus b as a function of Yh and Y when Bo 100 m 0 2 n 0 4 and o 0 6 B 100 0 4Yh 0 2 Y Yh if Y Yh 100 0 4Yh 0 6 Yh Y if Y Yh Highest bonus is achieved when Yh Y

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 8

truth inducing budget based contract
actual output y 50 60 70 80 90 100 110 120 50 120 122 124 126 128 130 132 134 60 118 124 126 128 130 132 134 136 70 116 122 128 130 132 134 136 138 80 114 120 126 132 134 136 138 140 90 112 118 124 130 136 138 140 142 100 110 116 122 128 134 140 142 144 110 108 114 120 126 132 138 144 146 120 106 112 118 124 130 136 142 148

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 9

Role of uncertainity




This model really makes little sense without some form of uncertainty If the controller believes that the manager knew with certainty what the result would be the incentive scheme would simply be to penalize the manager heavily for not disclosing during the planning stage what turned out to have actually happened

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 10

Certainty case is both trivial and unrealistic When manager doesn t know for sure what the outcome will be but have beliefs about what will happen then manager can express her certainty about the outcome Y in terms of cumulative probability distribution F Y


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 11

Managers expected return is maximized under this scheme when the target communicated has the following property F Yh n m o n


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 12

If we want mean of managers distribution of Y communicated as Yh then the parameters of the model will be chosen so that 0 5 n m o m or o 2n m


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 13

Continue with the example assume that manager believes that actual output can assume any value on the interval 50 to 120 with all outcomes equally likely F Yh n m 0 4 0 2 1 o m 0 6 0 2 2


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 14

That is the manager chooses a budget target Yh with the property that the probability is one half that the actual outcome will be less than or equal to the budget In the example manager choose 85 as the budget 50 0 5 120 50 85


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 15


Suppose we choose soviet rule of setting n 30 larger than m therefore n 1 3m now using rule o 2n m o 2 1 3m m 1 6m So if we set m 0 2 we have n 1 3 0 2 0 26 and o 1 6 0 2 0 32 Now F Yh n m 0 26 0 20 1 o m 0 32 0 20 2




MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 16

Value of Y h that would be chosen is 85 50 0 5 120 50 85 Same incentive to communicate the mean has been achieved but with parameters that are much closer together


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 17

Revised scheme of this model






Compensation experts appears to agree that a person s basic market wage should not be at risk Fixed part of the compensation should reflect the individuals market wage The fixed part of the compensation equals the Bo component and the portion of the BYh component

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 18

B Bo n Yh Yt m Y Yh if Y Yh Bo n Yh Yt o Yh Y if Y Yh
where Yt target output level Y output level Yh budgeted output level

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 19

Limitations






It provides an attractive but imperfect mechanism for eliciting realistic forecasts Costly mechanism since cash payments to managers are being transferred based on forecast Yh rather on the basis of actual output Problems in choosing the absolute level of Bo m n o

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 20



Another problem arises from the risk aversion on the part of the manager in the presence of uncertainty with risk aversion the budget chosen by the manager will reflect in addition to the parameters already discussed the managers attitude towards risk

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 21



Most important limitation is if real resources are to be transferred among divisions on the basis of forecasts an incentive still exist to misrepresent forecasts

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 22







When two or more divisions are competing for the firms common scare resources such as capital or computing facilities In this situation the central planner acquires a resource centrally and then allocates the capacity to the individual manager each period To undertakes the resource allocation the central planner relies on forecasts provided by the division managers of the return that they can earn from using the centrally supplied resource

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 23



In a system in which managers are rewarded based on divisional profitability and divisional profitability doesn t reflect the cost of centrally held resources they use the managers are motivated to overstate the return that the resource can provide in order to obtain as much of the centrally controlled resource as possible

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 24

Groves mechanism computes the performance measure for division I plus the forecasted profits of other divisions at their actual allocated resource level It has following advantages




Each division will attempt to maximize its actual profits since the groves measure is strictly increasing in the divisions profits Each divisions is best off sending an accurate forecast independent of what any other division sends or how it believes any other division is computing its forecast



MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 25

Each division s performance measure will be independent of the realized actual profits or operating efficiency of the other divisions Form of incentive scheme is REWARD a k Q R where a is a constant K is any constant lying between 0 and 1 Q is expected profit of all other divisions R is realized profit of managers divisions


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 26

Managers reward wage k firm s total expected profits variance in managers division If the division reward were based on realized division profits each division manager would be motivated to distort the communicated information about the value of allocating the common resource to himself Eliminating this tendency is the portion Q of the reward function relating to corporate profitability


MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 27

MODELS IMT 2008 QUANTITATIVE RELATIONSHIPS 28