Public Sector Cost Of Capital: A Comparison Of Two Models.
As part of a broader pattern of micro and macro-economic reform in New Zealand since 1984, two significant initiatives concerned directly with cost of capital, and more fundamentally with rational use of capital in the public sector, have been undertaken. The first, dating from 1989 and called Capital Charging, was applied to government Departments (see Lally, 1995). It required them to make periodic payments to The Treasury determined by the prevailing level of capital invested in them and a rate of return compensating for time, risk and taxation. Departmental appropriations were increased to initially neutralise the required payments. However subsequent variations in departmental capital and/or output led to variations in the capital charge and/or the appropriation respectively. Thus neutrality was not guaranteed after introduction, so as to provide incentives to improve utilisation of capital.
The second initiative arose from the establishment of the State-Owned-Enterprises (SOEs) in the late 1980's. Some were subsequently privatised. Amongst those that were not, there has been continuing concern about the attainability of efficient operations without resort to privatisation. Some elements of private sector disciplines are present in these SOEs, in particular the preponderance of output sold to the private sector, the presence of debt and the labour market disciplines arising from assigning clear financial objectives to their managers. However the ultimate private sector disciplines are absent, in particular the ongoing performance assessment implicit in traded equity prices, the prospect of takeovers, and shareholders whose personal wealth is directly affected by the entity's performance. In these respects the SOEs are much like "investment" centres in private sector firms (large subunits whose managers possess substantial autonomy). Firms of this type are increasingly relying on a form of accounting performance measurement called "Residual Income" or "Value Based Reporting", in which accounting income is determined net of the costs of both debt and equity capital (traditional external accounting reports omit the latter and traditional internal reports often omit both). Following this private sector trend, of which Fletcher Challenge and Telecom are prominent examples, some SOEs have moved to report performance in the same way -- to date, Transpower and Airways Corp. This naturally leads to concerns about consistent methodology for measuring the cost of capital. As a consequence The Treasury has prepared a Handbook suggesting methodology in this area (see The Treasury, 1997). Although principally targeted at SOEs, it is also intended to apply to other public sector entities exclusive of Departments (denoted Crown Entities).
This paper describes these two processes for assessing and assigning cost of capital in the public sector, examines the differences, and recommends convergence in three areas.
2. Two Processes
In respect of Capital Charging, a Department's nominal cost of capital is calculated as a weighted average of the pre-company tax costs of debt and equity, with weights equal to the proportions of debt and equity observed in what are judged to be comparable private sector firms. This weighted average cost of capital is then converted to real terms by subtracting the expected rate of inflation (see Lally, 1995). The resulting real cost of capital, denoted k, is
(1) k = [[k.sub.e]/1-[T.sub.e]] (1-L) + [k.sub.d] L-i
where [k.sub.e] = cost of equity (expected rate of return to shareholders just compensating for risk)
[T.sub.e] = effective corporate tax rate, so that [k.sub.e]/(1 - [T.sub.e]) is the cost of equity on a pre-company tax basis
[k.sub.d] = interest rate on debt
L = leverage ratio (debt to total capital)
i = expected rate of inflation
Of these parameters, the expected rate of inflation is drawn from an average of forecaster's estimates, and is the subject of annual updating. The leverage ratio is drawn from what are judged to be comparable private sector firms or, in their absence, one of a set of default options such as the average over listed companies. Also, the interest rate on debt was set at 1% over the government borrowing rate and the effective company tax rate was estimated at .20. Finally, the cost of equity [k.sub.e] is determined using the standard version of the Capital Asset Pricing Model (or CAPM: see Sharpe, 1964, and Lintner, 1965):
(2) [k.sub.e] = [R.sub.f] + ([E.sub.m] - [R.sub.f] [[Beta].sub.e]
where [R.sub.f] = riskfree rate of interest (proxied by that on government debt)
[E.sub.m] = expected rate of return on the market portfolio
[[Beta].sub.e] = beta of the Department's equity (sensitivity of Department's equity value to market-wide returns)
Estimating this cost of equity involves estimating the riskfree rate, the market risk premium ([E.sub.m] - [R.sub.f]), and the equity beta. The first of these is observable but varies over time, and therefore is subject to annual updating. The second is not observable and has been estimated by The Treasury at .065 (see Lally, 1995, p19). The last parameter [[Beta].sub.e] is similarly unobservable, and potentially varies over Departments. Estimates are obtained from comparable private sector firms or, in their absence, one of a set of default options, such as the average beta for listed equities (see Lally, 1995). Having calculated the cost of capital in this way, a cash payment is made periodically to The Treasury to reflect this rate and the Department's level of capital for the relevant period.
By contrast, an SOE's nominal cost of capital is simply the cost of equity assuming zero debt financing (i.e. the unlevered cost of equity, denoted [k.sub.u]) and without "grossing up" to a pre-company tax basis. Furthermore [k.sub.u] is determined using a personal tax adjusted version of the CAPM (see The Treasury, 1997)
(3) [k.sub.u] = [R.sub.f](1 - [T.sub.c])[[E.sub.m]-[R.sub.f](1 - T.sub.c])][[Beta].sub.u]
where [T.sub.c] is the statutory company tax rate (currently 33%) and [[Beta].sub.u] is the beta in the absence of debt. The latter is estimated from comparable companies or, in their absence, one of two default options (see The Treasury, 1997). The parameter  is called the Market Risk Premium and has been estimated at .09 (see The Treasury, 1997). This formula is a special case of a more general tax-adjusted CAPM (see Lally, 1992), involving zero tax on equity returns (both dividends and capital gains), and interest taxed at the statutory company tax rate [T.sub.c]. The formula given assumes nominal values. It may however be converted to real terms -- the SOE methodology is not prescriptive on this question, although applications to date involve nominal costs of capital (see Transpower NZ Ltd, 1996, and Airways Corporation of NZ, 1996). Calculation of the real rate follows the Fisher equation and equals
(4) 1-[k.sub.u]/1+i - 1
As noted earlier, this SOE methodology is also intended to apply to Crown Entities. For cost of capital purposes the only relevant distinction is the fact that some of these Crown Entities are not subject to company tax. In that event their nominal cost of capital is determined as above and then converted to pre-company tax terms, to yield a nominal cost of capital of [k.sub.u]/(1-[T.sub.e]). The prescribed value for [T.sub.e] is 33%. If a real rate is required it is obtained from the Fisher equation (4).
These descriptions reveal two points of similarity and six of difference between the two processes. The first point of similarity is the use of models designed for estimating private sector cost of capital (although that for capital charging is then adjusted for corporate taxes). Before 1984, this approach was not employed in the New Zealand public sector, and is still uncommon in public sectors elsewhere (see, for example, Hartman, 1990, Kemp, 1990, and the Australian Department of Finance, 1987, for evidence on US, UK and Australian practice respectively, and Lally, 1995, for a review of the arguments in this area).
The second point of similarity is the choice of the CAPM for estimating the cost of equity. Although a range of alternatives exist (see, for example, Weston and Copeland 1991, p610-613), only Arbitrage Pricing Theory (see Ross, 1976) has proper theoretical foundation. However the latter poses even more severe parameter estimation problems than either version of the CAPM described here, and as a result has not been employed in cost of capital estimation in New Zealand. Alternatives to the CAPM which have been used in New Zealand have now largely given way to some version of the CAPM.
The six points of difference between the two processes are as follows:
1. Capital charges for Departments are based on the standard version of the CAPM, which assumes all forms of income are taxed equally at the personal level. By contrast, the SOE methodology uses a version recognising differential personal tax treatment of income types.
2. Capital charges for Departments are based on a weighted average of the costs of both debt and equity capital whereas the SOE methodology uses an unlevered cost of equity.
3. Capital charges for Departments are cash events (the Department must pay the charge to The Treasury) whereas costs of capital for SOEs are only accounting events.
4. Capital charges for Departments are framed in real terms. By contrast the SOE methodology is not prescriptive on this question, although all applications to date use a nominal cost of capital (this leads to countervailing adjustments elsewhere).
5. Capital charges for Departments are expressed on a pre-company tax basis, whereas those for taxed entities subject to the SOE methodology are not.
6. Capital charges for departments are expressed on a pre-company tax basis by using an effective company tax rate of 20%. By contrast the SOE methodology prescribes a rate of 33% in respect of untaxed entities.
These points of distinction are now examined in detail.
3. Choice of CAPM
Capital Charging uses the standard form of the CAPM, which assumes inter alia that all forms of income are taxed equally at the personal level. By contrast, the SOE methodology uses a version based on differential tax treatment of various forms of personal income. Specifically, it is assumed that equity retinas (in the form of dividends and capital gains) are free of personal taxation whilst interest is taxed at a rate equal to the statutory corporate rate. The latter tax scenario is called tax "neutrality", because it implies that debt capital has no tax advantage over equity capital, i.e. the corporate tax advantage of debt, in the form of the tax deductibility of interest, is offset perfectly by the personal tax disadvantage. The former tax scenario, underlying the standard CAPM, is called "classical".
In New Zealand, equity returns are largely free of personal tax, due to
1. the process of dividend imputation, which largely eliminates any personal taxation on dividends (see Smith, 1993, for a comprehensive discussion of this process)
2. the exemption of some investors, such as individuals, from capital gains tax and the deferral opportunity (till sale of shares) which is available to the rest, largely eliminates capital gains tax. In respect of the deferral opportunity, Protopapadakis (1983) estimates that deferral effectively halves the tax.
In addition most investors are taxed on interest at rates equal or close to the statutory corporate rate. Thus neutrality is a better description of the New Zealand tax scenario than classical. Accordingly the CAPM implicit in the SOE methodology is more appropriate than that implicit in Capital Charging.
The reasons why Capital Charging uses what is now a less appropriate CAPM are historical. When Capital Charging was first contemplated, dividend imputation had not yet been introduced in New Zealand. Accordingly it was arguable that the tax regime at that time was closer to classical than neutral. Furthermore, at that time, use of the standard CAPM was accepted market practice and consistency with market practice was (and still is) considered desirable. The subsequent adoption of Dividend Imputation clearly tilted the tax regime towards neutral, and market practice has shifted from the standard to the tax adjusted CAPM. On both grounds the SOE choice of CAPM is superior, and should be adopted for Capital Charging.
4. WACC versus Unlevered Cost Of Equity
Capital Charging for Departments uses a weighted average of the pre-company tax costs of debt and equity, while the SOE methodology uses an unlevered cost of equity. However SOEs are in general partly debt financed, and this implies that the cost of capital is a weighted average of the costs of debt and equity. Thus the disregarding of debt in the SOE methodology implies a belief that debt (i.e. leverage) does not affect this weighted average cost of capital (WACC). As shown in Appendix 1, the necessary and sufficient conditions for this are (a) tax neutrality exists, as defined above, (b) the cost of debt does not embody a default premium, i.e. an allowance through a higher promised return for the probability of default and the expected loss to debtholders arising from such, and (c) since the cost of equity is determined by a theoretical model which does not allow for the inferior liquidity of equity capital relative to benchmark government bonds, then the cost of debt must not embody a liquidity premium, i.e. an allowance through a higher promised return for inferior liquidity relative to government bonds
None of these assumptions is perfectly realised. In respect of (a), the desire for general acceptability of any public sector cost of capital methodology points to the minimisation of contentious parameter estimation problems. This obliges one to choose between one of two competing tax scenarios - neutral and classical as described. As argued in the preceding section, the former is the better description of the New Zealand tax environment.
In respect of conditions (b) and (c) above, while such premiums exist, their effect on WACC will be small for SOEs. For example, suppose that the interest rate on an SOE's debt ([k.sub.d]) embodies combined default and liquidity premiums of 1% (so that [k.sub.d] exceeds the government bond rate by at least 1%). Also the SOE's leverage is 50%. Both estimates are generous for SOEs. Now WACC is a weighted average of the cost of equity ([k.sub.e]) and the cost of debt (net of its corporate tax deduction at the statutory corporate tax rate To), with weights (here) of 50% each, i.e.
WACC = .5[k.sub.e] + .5[k.sub.d] (1 - [T.sub.c])
With a corporate tax rate of [T.sub.c] = .33, the effect on WACC of [k.sub.d] embodying 1% for default and liquidity premiums is only
(.5).01(1 -.33) = .0033
This is not a substantial sum.
In summary, the use of an unlevered cost of equity for SOEs presumes that WACC is independent of leverage, and this rests on three conditions. Of these, tax neutrality is a better approximation to the actual tax scenario than the only feasible alternative. The other two conditions, while not realised, would exert only trivial effect. Thus WACC is essentially independent of leverage. Simplicity then favours use of an unlevered cost of capital. This conclusion is strengthened by the difficulty in measuring SOE leverage, i.e. leverage ratios in WACC are defined in market value terms, yet market equity values for SOEs are unobservable (and accounting values are in general poor proxies).
In respect of Departments, the use of the standard CAPM, inconsistent as it is with tax neutrality, rules out use of an unlevered cost of capital. If the standard CAPM was replaced here by the tax adjusted version, then use of an unlevered cost of equity would be defensible. However the leverage measurement problem discussed above for SOEs does not arise for Departments. This is because the investment in Departments is financed from a pool, containing both equity (from taxation) and debt, which is also used to finance consumption type expenditures. Thus Departmental leverage is indeterminate (as opposed to merely difficult to measure). Accordingly the leverage level attributed to Departments, under the Capital Charge regime, is that of private sector firms engaged in similar activities or, in their absence, one of a set of default options (such as the average over listed companies). For such firms, leverage is both determinate and measurable.
5. Cash versus Accounting Events
Unlike SOEs, capital charges for Departments are cash events, i.e. Departments make periodic payments to The Treasury to reflect the level of capital used and the rate of charge. Since SOEs are partly financed by debt, on which interest payments are made, and make dividend payments to government in respect of their equity capital, they could not be subject to cash type capital charges (which would essentially duplicate these existing payments).
The question still remains as to why Departments are subject to cash rather than purely accounting charges for cost of capital. The reason for this is essentially motivational. Departments lack the disciplines which motivate private sector firms, and even SOEs to some extent, to rationally use capital. These include:
(a) product market disciplines - inefficient use of capital, when transmitted via higher output prices, leads to loss of business,
(b) capital market disciplines - inefficient use of capital leads to poor financial performance. This leads to shareholder/ debtholder interventions (especially takeovers and receiverships) and/or withholding of new capital, and
(c) labour market disciplines - inefficient use of capital leads to poor financial performance, which damages a manager's reputation and will eventually lead to their replacement.
In general Departments lack strong product market disciplines due to most of their output being purchased by government. Debtholder disciplines are of course absent. Government itself has only weak incentives to intervene since it is a mere agent for the ultimate shareholders, who in turn "appoint" them for a variety of reasons. Finally labour market disciplines are weak because Departmental managers have significant non-financial objectives.
In light of all this, compensating disciplines are desirable. Upgrading the capital charge from an accounting event to a cash event is consistent with this, i.e. increasingly irrational use of capital leads to a cash as well as an accounting deficit, and the incremental effect should be motivational.
6. Real versus Nominal Costs Of Capital
As noted, capital charges for Departments are framed in real rather than nominal terms, while the SOE methodology is not prescriptive on this issue (applications to date though have involved nominal costs). The choice of real rates for Departments springs from their capital charges being cash rather than merely accounting events, and the reason is as follows.
In applying a capital charge, a fundamental requirement is consistency between the economic substance and the accounting representation. Thus an investment with zero Net Present Value (economically neutral) should be expected to yield zero income net of the capital charge in every year of its life, i.e. an economically neutral project is expected to be neutral in accounting terms. Under inflationary conditions this can be achieved through two possible accounting methods:
1. Apply a real cost of capital to the revalued asset value (at period end) and exclude asset revaluations from income; or
2. Apply a nominal cost of capital to the revalued asset value (at period commencement) and include asset revaluations in income
To illustrate this, consider a Department contemplating purchase of a non-depreciating asset for $10m, in order to avoid operating costs expected to be $1m in the first year and thereafter grow at 3% p.a. due to inflation. The nominal cost of capital is 13% (and hence its real counterpart is 9.71%). The present value of the nominal benefit stream, discounted at 13%, equals the asset cost of $10m. So the project is economically neutral. Since inflation is 3% p.a., we expect the asset value to grow at that rate. So, applying method (1) yields income of
Year 1 Year 2 Cost Saving $1m $1.03m - Capital Charge at 9.71% $1m $1.03m 0 0
and method (2) yields
Year 1 Year 2 Cost Saving $im $1.03m + Asset Revaluation $.3m $.309m - Capital Charge at 13% $1.3m $1.339m 0 0
So both methods are expected to yield zero income in each year. Since the capital charge is a cash event then the first method is also expected to yield a cash outcome of zero in all years. However with method (2), in which the asset revaluation is a non-cash event but the cost savings and capital charges are not, there will be an expected cash deficit in each year equal to the expected asset revaluation. Accordingly, with method (2), even some economically desirable projects (Net Present Value positive) will be expected to generate cash deficits. This may discourage adoption of such projects. So method (1) is preferred because it is free of this drawback.
As noted, the SOE methodology is not prescriptive on this issue. However applications to date have utilised method (2), i.e. a nominal cost of capital coupled with inclusion of asset revaluations in income.
7. Grossing Up For Corporate Taxes
The next point of distinction between the Departmental and SOE processes is that a Department's cost of capital is increased to express it on a pre-company tax basis. By contrast, rates for SOEs (and Crown Entities which are taxed) are not. The reason for this is that Departments, unlike SOEs and taxed Crown Entities, are not subject to corporate tax. Thus, in the absence of grossing up, Departments would, in evaluating investment projects, always find a given project more desirable than an SOE (or any other tax paying entity). This is not a desirable outcome, i.e. all public sector entities should generate the same valuation for a given project. To ensure such consistent valuation, capital charge rates for Departments must be increased (to yield pre-company tax rates) so that project valuations are at least approximately consistent with company tax paying entities. This is done by "grossing up" the cost of equity (dividing by 1 -[T.sub.e], where [T.sub.e] is the effective company tax rate) and disregarding the corporate tax deduction which would otherwise arise for interest.
Lally (1995) demonstrates that this adjustment formula yields identical valuations across tax paying and non-tax paying entities for projects involving perpetual and equal expected cash flows. We illustrate here with a simple example. In order to focus purely on the tax issue, we act as if the taxed entity (an SOE, for example) and the untaxed entity both use an unlevered cost of capital. The project involves purchase of a non-depreciating asset for $10m, in order to avoid operating costs expected to be $1.3m pre-tax p.a. indefinitely. The cost of capital (before grossing up) is 10%, and the effective company tax rate is 33%. The tax paying entity will enjoy an after tax benefit per annum of
$1.3m (1 - .33) = $.87m
Using a discount rate of 10%, the present value is $8.7m. With a cost of $10m, the project is not desirable. By contrast , an entity exempt from tax (a Department) would enjoy a benefit of $1.3m p.a. In the absence of grossing up its discount rate, it would value the project (using the 10% discount rate) at $13m, and hence judge it to be desirable. To avoid this inconsistency, the discount rate for the untaxed entity must be grossed up by the effective tax rate to
.10/1-.33 = .149
Applying .149 to the cash flow of $1.3m p.a. now yields a present value of
$1.3m/.149 = $8.7m
which is identical to that of the taxed entity. The same decision - to reject the project - then results.
This example merely demonstrates equivalence for projects involving perpetual and equal expected cash flows. Any variation from this produces valuation discrepancies between taxed and untaxed entities. An example is a perpetual project involving expected cash flows which are equal over time in real rather than nominal terms. In this event, grossing up by the above method will not be correct, as shown in Appendix 2. Furthermore the difference will be non-trivial for anything beyond very low rates of inflation.
8. The Rate For Grossing Up
The final point of distinction between the two processes is that Departmental cost of capital is converted to pre-company tax terms using an effective company tax rate of 20% whereas the SOE methodology prescribes 33% for untaxed entities. The difference reflects The Treasury's view about the effective corporate tax rate at two distinct points in time. The preceding section has shown that grossing up is required to ensure the same valuation of a given project by both taxed and untaxed entities. By extension, for all untaxed entities (whether Departments or Crown Entities), the same rate for grossing up is necessary to achieve the same valuation of a given project.
9. Implications For Policy
The Capital Charging and SOE processes then differ in six respects. Three of these differences - cash versus accounting charges, real versus nominal cost of capital, and whether or not to gross up for company taxes - spring from fundamental differences in the environments in which Departments and SOEs operate. Thus no modification to either process is warranted in these respects. Three further differences lie in the effective company tax rate used in any grossing up, the CAPM variant used and in whether debt financing is allowed for. The first of these differences reflects a change in The Treasury's belief about the effective corporate tax rate. Consistency points to using the same rate. The chosen rate should accord with the current beliefs, and this is the 33% rate prescribed in the SOE methodology.
The last two differences arise from different assumptions about the underlying tax environment. The Capital Charging model assumes a classical tax environment, in which debt has a tax advantage over equity, while the SOE model assumes that debt has no tax advantage. The latter tax assumption is a better description of the current New Zealand situation. That the former assumption was invoked for Capital Charging reflects the New Zealand tax situation at the time of developing this model. Since then the tax environment has changed. Accordingly the Capital Charging model should be revised in two further ways to reflect this.
The first change would be to utilise an unlevered cost of capital. Although this will not materially change the cost of capital (due to debt having no tax advantage), the arguments for doing so are consistency with the SOE methodology and simplicity. The simplification is twofold: a reduction in the number of terms in the formula and avoidance of the need to estimate by how much the cost of debt exceeds the riskfree rate. Accordingly the real cost of capital of a Department, which is currently determined by equation (1), would now be given by
(5) k = [k.sub.u]/1-[T.sub.e] - i
The second change would be to adopt the tax adjusted version of the CAPM for determining [k.sub.u], as shown in equation (3).
In summary then, three changes should be made to the Capital Charging methodology: using an effective company tax rate of 33%, an unlevered cost of capital, and calculation of this from the CAPM in equation (3). Unlike the second of these proposals, the first and third may significantly effect the cost of capital. The effect of all three changes is captured by comparing equations (5) and (3) with (1) and (2). Equation (5) requires an effective company tax rate [T.sub.e]. The SOE methodology recommends the statutory rate of 33%. Using this rate, and the parameters described earlier for each of the two models, the excess of (5) and (3) over (1) and (2) is
(6) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Following Appendix 1, the equity beta [[Beta].sub.e] is related to the unlevered beta [[Beta].sub.u] by
(7) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Since tax neutrality of debt is judged to be a good description of the New Zealand tax situation, then [Delta] = 0. In addition the SOE methodology sets the cost of debt [k.sub.d] at 1% over the riskfree rate. Since this margin allows for the default premium, inferior liquidity and risk, then the allowance for risk will be very small. Accordingly the debt beta [[Beta].sub.d] must also be small, and is treated as zero. Equation (7) then becomes
[[Beta].sub.e] = [[Beta].sub.u] [1 + B/S]
Substituting this into (6), the excess of (5) and (3) over (1) and (2) is then
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
This difference then depends upon the values for [[Beta].sub.u], B/V and [R.sub.f]. The latter varies over time, and the first two over Departments. A typical value for [R.sub.f] is about .07. Using this the last equation becomes
(5)-(1) = .053[[Beta].sub.u] +.0075 B/V - .0175
So the difference is monotonically increasing in both [[Beta].sub.u] and B/V. Values for [[Beta].sub.u] range from about .30 to .70 (see The Treasury, 1997, p58). Also B/V ranges from 0 to .50. Accordingly the difference in the last equation will range from -.002 (when both [[Beta].sub.u] and B/V are at their minima) to .023 (when both are at their maxima). The latter is not trivial. Most of the effect comes from the unlevered beta assuming a value at the top end of its range (.70). This value of .70 represents the average unlevered beta amongst private sector assets. Many activities within Departments have been assessed as average risk, and hence warranting an asset beta of .70 (see Lally, 1995, p16). Consequently the difference in the last equation will be around 2% for a considerable set of Departmental activities.
This understatement of cost of capital under the Capital Charge regime is significant for the following reason. The model for determining a Department's cost of capital will also serve to determine the appropriate discount rate for valuing prospective projects. Errors in specifying the cost of capital of a Department then leads to errors in valuing prospective projects. In particular, prospective projects of the same risk as a Department's existing portfolio (and thereby warranting the same discount rate) will be valued using a discount rate which is too low by up to 2%. Overvaluation then results, leading to overinvestment.
In summary, it is recommended that the Capital Charging methodology be altered in three respects, to reflect changes in the tax environment since its adoption. The first change is to increase the effective company tax rate used to 33%, consistent with the SOE methodology. The second is to use an unlevered cost of capital, for simplicity and consistency with the SOE methodology. The third change is to employ the tax adjusted version of the CAPM. Failure to adopt the second has no significant effect on the resulting cost of capital. By contrast, failure to adopt the first and third will understate the nominal and real costs of capital by up to 2.3%, with the maximum operating if the activities in question have average private sector risk. This condition holds for many Departmental activities, and will flow through to errors in valuing prospective projects. A bias towards investment then results.
This paper describes and compares two processes for assessing and assigning cost of capital in the public sector: that for Departments under the Capital Charge process and that recommended for SOEs. The processes are similar in the sense of using "private sector" type technology, and in particular the CAPM. However they differ, or may differ, in six respects. Three of these spring from fundamental differences between the environments in which departments and SOEs operate, and comprise cash versus accounting charges, real versus nominal cost of capital, and the issue of whether to convert cost of capital to pre-company tax terms. The remaining points of difference comprise the effective company tax rate used for grossing up the cost of capital for untaxed entities, the CAPM variant used and whether there is allowance for debt financing. Each of these points of difference are attributable to different assumptions about the tax environment. This in turn reflects changes since Capital Charging was initiated. At the very least, the difference in the effective company tax rate and in the CAPM version used should be eliminated by revising the Capital Charging process. Failure to do so leads to underestimating a Department's cost of capital by up to two percentage points. The consequence is a bias towards investment in new projects.
Airways Corporation of New Zealand Ltd, Annual Report 1995-96.
Australian Department of Finance. (1987), The Choice of Discount Rate For Evaluating Public Sector Investment Projects.
Conine, T. (1980), "Corporate Debt and Corporate Taxes: An Extension", Journal of Finance, 35, 1033-1036.
Hartman, R. (1990), "One Thousand Points of Light Seeking a Number: A Case Study of Congressional Budget Office's Search for a Discount Rate Policy", Journal of Environmental Economics and Management, 18, 3-7.
Kemp, P. (1990), "Capital Charging in the NHS: The Next Steps", Public Finance and Accounting, February, 17-20.
Lally, M. (1992), "The CAPM Under Dividend Imputation", Pacific Accounting Review, 4, 31-44.
-- (1995), "The Cost of Capital for Government Entities", Accounting Research Journal, 8, No. 1, 14-25.
Lintner, J. (1965), "The Valuation of Risky Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets", Review of Economics and Statistics, 1-37.
Protopapadakis, A. (1983), "Some Indirect Evidence on Effective Capital Gains Tax Rates", Journal of Business, 56, 127-138.
Ross, S. (1976), "The Arbitrage Theory of Capital Asset Pricing", Journal of Economic Theory, 13, 341-360.
Sharpe, W. (1964), "Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of Risk", Journal of Finance, 19, 425-442.
Smith, A. (1993), "Comprehensive Dividend Imputation, Neutrality and Double Taxation of Corporate Profits", The Bulletin For International Fiscal Documentation, 47, No. 10, 568-580.
The Treasury. (1997), Estimating the Cost of Capital for Crown Entities and State-Owned-Enterprises.
Transpower New Zealand Ltd, Annual Report 1995-96.
Weston. J and Copeland, T. (1991), Managerial Finance, 9th edition, Dryden Press.
Martin Lally, School of Economics and Finance, Victoria University of Wellington
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||New Zealand|
|Publication:||New Zealand Economic Papers|
|Date:||Dec 1, 1998|
|Previous Article:||The Economic Benefits of Schooling in New Zealand: Comment and Update.|
|Next Article:||The Asian Financial Crisis: How Long Will The Downturn Last?|