«Paul M. Collier Aston Business School, Aston University Accounting for Managers Accounting for Managers: Interpreting accounting information for ...»
Again using ROI, the impact of the additional investment can be seen in Table 13.2. Using ROI, Division A may not want its project to be approved as the ROI of 20% is less than the current ROI of 25%. The impact of the new investment would be to reduce the divisional ROI to 24% (£1.2 million/£5 million). However, Division B would want its project to be approved as the ROI of 13% is higher
than the current ROI of 10%. The effect would be to increase Division B’s ROI slightly to 10.14% (£2.13 million/£21 million). However, the divisional preference for B’s investment over A, because of the rewards attached to increasing ROI, are dysfunctional to Majestic. The corporate view of Majestic would be to invest £1 million in Division A’s project because the ROI to the business as a whole would be 20% rather than 13%.
The disposal of the asset can be considered even without knowing its value.
If Division A currently obtains a 25% ROI, disposing of an asset with a return of only 19% will increase its average ROI. Division B would wish to retain its asset because it generates an ROI of 12% and disposal would reduce its average ROI to below the current 10%. Given a choice of retaining only one, Majestic would prefer to retain Division A’s asset as it has a higher ROI.
The difﬁculty with ROI as a measure of performance is that it ignores both the difference in size between the two divisions and Majestic’s cost of capital. These issues are addressed by the residual income method.
Using residual income (RI), the divisional performance can be compared as in Table 13.3. In this case, we can see that Division A is contributing to shareholder value as it generates a positive RI, while Division B is eroding its shareholder value because the proﬁt it generates is less than the cost of capital on the investment.
Using RI, the impact of the additional investment is shown in Table 13.4. Under the residual income approach, Division A’s project would be accepted (positive RI) while Division B’s would be rejected (negative RI).
Similarly for the asset disposal, Division A’s asset would be retained (ROI of 19% exceeds cost of capital of 15%), while Division B’s asset would be disposed of (ROI of 12% is less than cost of capital of 15%).
The main problem facing Majestic is that the larger of the two divisions (both in terms of investment and proﬁts) is generating a negative residual income and consequently eroding shareholder value.
A further problem associated with measuring divisional performance is that of transfer pricing, which was introduced in Chapter 8.
Transfer pricing When decentralized business units conduct business with each other, an important question is what price to charge for in-company transactions, as this affects the proﬁtability of each business unit. However, transfer prices that are suitable for evaluating divisional performance may lead to divisions acting contrary to the corporate interest (Solomons, 1965).
For example, consider a company with two divisions. Division A can produce 10,000 units for a total cost of £100,000, but additional production costs are £5 per unit. Division A sells its output to Division B at £13 per unit in order to show a satisfactory proﬁt. Division B carries out further processing on the product. It can convert 10,000 units for a total cost of £300,000, but additional production costs are £15 per unit. The prices B can charge to customers will depend on the quantity it wants to sell. Market estimates of selling prices at different volumes (net of
variable selling costs) are:
Volume Price 10,000 units £50 per unit 12,000 units £46 per unit 15,000 units £39 per unit The ﬁnancial results for each division at each level of activity are shown in Table 13.5. Division A sees an increase in proﬁt as volume increases and will want to increase production volume to 15,000 units. However, Division B sees a steady erosion of divisional proﬁtability as volume increases and will seek to keep production limited to 10,000 units, at which point its maximum proﬁt is £70,000.
The company’s overall proﬁtability increases between 10,000 and 12,000 units, but then falls when volume increases to 15,000 units. From a whole-company perspective, therefore, volume should be maintained at 12,000 units to maximize proﬁts at £112,000. However, neither division will be satisﬁed with this result,
ACCOUNTING FOR MANAGERSTable 13.5 Divisional ﬁnancial results
as both will see it as disadvantaging them in terms of divisional proﬁts, against which divisional managers are evaluated.
For Division A, variable costs over 10,000 units are £5, but its transfer price is £13, so additional units contribute £8 each to divisional proﬁtability. A’s average costs reduce as volume increases, as Table 13.6 shows.
However for Division B, its variable costs over 10,000 units are £28 (transfer price of £13 plus conversion costs of £15). The reduction in average costs of £2.50 per unit is more than offset by the fall in selling price (net of variable selling costs), as Table 13.7 shows.
PERFORMANCE EVALUATION OF BUSINESS UNITS 203
There are several methods by which transfer prices between divisions can be
ž Market price: Where products/services can be sold on the outside market, the market price is used. This is the easiest way to ensure that divisional decisions are compatible with corporate proﬁt maximization. However, if there is no external market, particularly for an intermediate product – i.e. one that requires additional processing before it can be sold – this method cannot be used.
ž Marginal cost: The transfer price is the additional (variable) cost incurred. In the above example, the transfer price would be £5, but Division A would have little motivation to produce additional volume if only incremental costs were covered.
ž Full cost: This method would recover both ﬁxed and variable costs. This has the same overhead allocation problem as identiﬁed in Chapter 11 and would have the same motivational problems as for the marginal cost transfer price.
ž Cost-plus: This method provides a proﬁt to each division, but has the problem identiﬁed in this example of leading to different management decisions in each division and at corporate level.
ž Negotiated prices: This may take into account market conditions, marginal costs and the need to motivate managers in each division. It tends to be the most practical solution to align the interests of divisions with the whole organization and to share the proﬁts equitably between each division. In using this method, care must be taken to consider differential capital investments between divisions, so that both are treated equitably in terms of ROI or RI criteria.
In practice, many organizations adopt negotiated prices in order to avoid demotivating effects on different business units. In some Japanese companies it is common to leave the proﬁt with the manufacturing division, placing the onus on the marketing division to achieve better market prices.
Transaction cost economics
A useful theoretical framework for understanding divisionalization and the transfer pricing problem is the transactions cost approach of Oliver Williamson (1975), which is concerned with the study of the economics of internal organization.
Transaction cost economics seeks to explain why separate activities that require
ACCOUNTING FOR MANAGERSco-ordination occur within the organization’s hierarchy (i.e. within the corporate structure), while others take place through exchanges outside the organization in the wider market (i.e. between arm’s-length buyers and sellers).
The work of business historians such as Chandler (1962) reﬂects a transaction cost approach in explanations of the growth of huge corporations such as General Motors, in which hierarchies were developed as alternatives to market transactions.
It is important to note that transactions take place within organizations, not just between them. For managers using accounting information, attention is focused on the transaction costs associated with different resource-allocation decisions and whether markets or hierarchies are more cost effective.
Transactions are more than exchanges of goods, services and money. They incur costs over and above the price for the commodity bought or sold, such as costs associated with negotiation, monitoring, administration, insurance etc.
They also involve time commitments and obligations, and are associated with legal, moral and power conditions. Understanding these costs may reveal that it is more economic to carry out an activity in-house than to accept a market price that appears less costly but may incur ‘transaction’ costs that are hidden in overhead costs.
Under transaction cost economics, attention focuses on the transaction costs involved in allocating resources within the organization, and determining when the costs associated with one mode of organizing transactions (e.g. markets) would be reduced by shifting those transactions to an alternative arrangement (e.g. the internal hierarchy of an organization). The high costs of market-related transactions can be avoided by specifying the rules for co-operative behaviour within the organization.
The markets and hierarchies perspective considers the vertical integration of production and the decision about whether organizations should make or buy. Both bounded rationality and opportunistic behaviour are assumed in this perspective (see Chapter 6 for a discussion of this in relation to agency theory) and transaction costs are affected by asset speciﬁcity, when an investment is made for a speciﬁc rather than a general purpose. Transaction costs are also affected by uncertainty and the frequency with which transactions take place.
Seal (1995) provided the example of a make versus buy decision for a component.
In management accounting, a unit cost comparison would take place. (Relevant costs for make versus buy decisions were described in Chapter 9.) A transaction cost approach would consider whether the production of the component required investment in specialized equipment or training, a problem of asset speciﬁcity.
This approach raises the problem that an external contract may be difﬁcult to draw up and enforce because the small number of organizations bargaining may be hindered by opportunistic behaviour. It may therefore be cheaper to produce in-house due to contractual problems.
Williamson (1975) argued that the desire to minimize transaction costs resulting from bounded rationality leads to transactions being kept within the organization, favouring the organizational hierarchy over markets. Markets are favoured where there are a large number of trading partners, which minimizes the risk of opportunistic behaviour.
PERFORMANCE EVALUATION OF BUSINESS UNITS 205Recurring, complex and uncertain exchanges that involve substantial investment may be more efﬁciently undertaken when internal organization replaces market transactions. The efﬁciency of a transaction that takes place within the organization depends on how the behaviour of managers is governed or constrained, how economic activities are subdivided and how the management accounting system is structured.
However, decision-makers may themselves indulge in opportunistic behaviour that causes the beneﬁts of internal transactions to be reduced. Therefore, the management accounting system can be used to ensure that these internal transactions are conducted efﬁciently.
Rather than reﬂecting a concern with utility maximization (the assumption of agency theory), the transaction cost framework is more concerned with bounded rationality. While an agency perspective ignores the power of owners and also that of employees, who can withdraw their labour, transaction cost theory gives recognition to power in the hierarchy that is used to co-ordinate production.
Conclusion: a critical perspective In this chapter we have described the divisionalized organization and how divisional performance can be evaluated. We have discussed the controllability principle and the transfer pricing problem.
The divisional form is a preferred organizational structure because it allows devolved responsibility while linking performance to organizational goals through measures such as ROI and RI that are meaningful at different organizational levels, particularly when these support shareholder value methods such as the link between RI and EVA.
However, research by Merchant (1987) concluded that the controllability principle was not found in practice and that managers should be evaluated ‘using all information that gives insight into their action choices’.
Managers are often critical that the corporate head ofﬁce fails to distinguish adequately between controllable and non-controllable overhead. The point has already been made in Chapter 2 that determining a risk-adjusted cost of capital can be a subjective exercise.
Relationships between business units frequently cause friction, particularly in some organizations where the number of business units has been increased to a level that is difﬁcult to manage. Transaction cost economics, a rational markets/hierarchies approach like agency theory described in Chapter 6, provides
a useful though limited perspective. For example, Child (1972) concluded:
When incorporating strategic choice in a theory of organization, one is recognizing the operation of an essentially political process in which constraints and opportunities are functions of the power exercised by decision-makers in the light of ideological values. (p. 16) The political process inherent in transfer pricing between divisions is also evidenced in many multinational corporations, where transfer pricing is more