26 January 2010
Ken Sutherland Unison Networks Ltd 1101 Omahu Road Hastings
RE: IMPLEMENTING THE DEFERRED TAX APPROACH Dear Ken In the process of developing Input Methodologies the Commerce Commission (Commission) has expressed a clear preference for the tax payable approach to the regulatory taxation allowance, whereas Unison has always advocated the deferred tax approach. In its Emerging Views Paper,1 the Commission has asked parties to provide a worked example of how the deferred tax approach could be implemented in practice. You have asked me to provide a report on this matter and this letter serves as my report. I understand that this report will be submitted to the Commerce Commission as part of Unison’s pre-workshop submission on regulatory tax issues. This report is to be regarded as a statement of evidence on my part and I confirm that it is provided in accordance with the Code of Conduct for expert witnesses contained in the High Court Rules. In particular, I acknowledge that I have read the Code of conduct and agree to comply with it. My qualifications as an expert are attached to prior statements and reports Unison has already submitted to the Commission. I confirm that the issues and matters addressed in this report are within my area of expertise. The implementation of the deferred tax approach within the economic regulatory framework under Part 4 of the Commerce Act is partly an issue of tax practice and partly an issue of financial economics. Tax law and practice will determine the actual tax depreciation that arises, and hence the magnitude of any benefits from that source. But the question of ensuring that the net present value of cash flows is the same under the deferred tax approach as it is under the tax payable approach is an issue of financial economics. The Commission claims that “while the NPV-equivalence of the deferred tax and tax payable approaches is straightforward to achieve for new assets, difficulties arise for assets that are already part way through their lifetimes”.2 It is not clear how the Commission has come to this conclusion. As I discuss below, there is a straightforward option for implementing the deferred tax approach that ensures that NPV equivalence is achieved.
1
Commerce Commission, Input Methodologies (Electricity Distribution) Emerging Views, 23 December 2009.
2
Commerce Commission, Emerging Views Paper, paragraph A16, p. 80.
CRA International (Hong Kong) Limited 1902A Tower Two
Lippo Centre
89 Queensway
Hong Kong
+852 2521 5501
+852 2521 5582 fax
Implementing the Deferred Tax Approach 26 January 2010 Page 2
This report provides technical details of how the deferred tax approach can be implemented in practice.
OPENING DEFERRED TAX BALANCE It is desirable that the balance of the deferred tax account reduces to zero at the end of the asset’s life. If this does not happen then there is a residual liability that must be written off. However, the deferred tax account will be continually updated over time, and with a mix of assets of differing remaining lives, it will be impossible to determine how much of the balance should be written off. Calculation of tax expense using regulatory depreciation (derived from the regulatory asset base) and tax payable using tax depreciation (derived from the regulatory tax written down value (WDV)) suggests that the opening balance of the deferred tax account must be set according to the following formula if it is to reduce to zero at the end of the asset’s life:
DT0 = (RAB0 − WDV0 )t Where DT0 is the opening value of the deferred tax balance; RAB0 is the opening value of the regulatory asset base; WDV0 is the opening value of the regulatory tax written down value (WDV); and t is the corporate income tax rate. The opening value of the deferred tax balance for regulatory purposes is therefore derived from the opening value of the regulatory asset base (RAB) and the opening value of the regulatory tax WDV, and cannot be set independently of these two values.
OPTIONS FOR IMPLEMENTING THE DEFERRED TAX APPROACH Subject to the constraint on the opening deferred tax balance, there are several options that could conceivably be used for implementing the deferred tax approach: •
Set the opening regulatory tax WDV equal to the actual tax WDV, adjusted to exclude any premium paid as part of an acquisition. Use the opening RAB and the opening regulatory tax WDV and calculate the opening deferred tax balance from the formula above;
•
If the first option does not yield the same NPV as the tax payable approach then derive a regulatory tax WDV that does produce the correct NPV; and
•
Set the opening regulatory tax WDV equal to the actual tax WDV, adjusted to exclude any premium paid as part of an acquisition. Recognise the difference between the opening RAB and the opening regulatory tax WDV as a source of ‘permanent differences’ over the lifetime of the existing assets, and account for those permanent differences when calculating the regulatory tax allowance.
Options 1 and 3 both set the regulatory tax WDV equal to the actual tax WDV, adjusted to exclude any acquisition premium paid as part of an acquisition. This adjustment reflects the Commission’s preliminary decision on the treatment of acquisitions. Option 2 uses a derived regulatory tax WDV.
Implementing the Deferred Tax Approach 26 January 2010 Page 3
The Commission is concerned to ensure that the net present value (NPV) of cash flows under the deferred tax approach is equal to the NPV of cash flows under the tax payable approach. I refer to this as the “NPV equivalence condition”.
OPTION 1: UTILISE OPENING RAB AND ACTUAL TAX WDV Under this option the opening value of the tax WDV for regulatory purposes is set to the opening value of actual WDV. Using this option, the NPV of cash flows under the deferred tax approach does not equal the NPV of cash flows under the tax payable approach, except for the case of a new asset or a Hypothetical New Entrant (HNE) where both the RAB and the tax WDV are initially equal. An example spreadsheet is provided with this letter. The example in the spreadsheet assumes that: •
the opening regulatory asset base is 1,000;
•
the opening actual tax WDV is 400;
•
the assets have 20 years of life remaining;
•
the annual rate of revaluations is 2.0%; and
•
the weighted average cost of capital is 8.0%.
Given these assumptions the tax payable approach provides an NPV of 1,000, but the deferred tax approach provides an NPV of only 820. This option does not meet the NPV equivalence test and is not a suitable basis for implementing the deferred tax approach.
OPTION 2: DERIVED REGULATORY TAX WDV It is possible to derive a regulatory tax WDV that (a) is entirely independent of the actual tax WDV, and (b) ensures that the NPV equivalence condition is met. It is possible to derive a formula that will calculate the derived regulatory tax WDV from relevant parameters. However, the resulting formula is long, complex, and not at all intuitive. Alternatively, the regulatory tax WDV that yields NPV equivalence can be solved by a spreadsheet model. While this provides demonstrable proof that NPV equivalence is met, the resulting regulatory tax WDV again has no intuitive explanation. The example spreadsheet shows that, given the assumptions listed earlier, the opening regulatory tax WDV is 1,717.86, and the corresponding opening deferred tax balance is -215.36. Neither the regulatory tax WDV nor the deferred tax balance has an intuitive explanation, and thus may lack credibility even though NPV equivalence is achieved.
OPTION 3: ACCOUNTING FOR PERMANENT DIFFERENCES I have noted under Option 1 that if the RAB and the tax WDV are initially equal then NPV equivalence is achieved. The Commission is concerned, however, with situations where assets are part way through their lives and, as a result, the opening RAB and regulatory tax WDV are not equal.
Implementing the Deferred Tax Approach 26 January 2010 Page 4
The simplest solution to this problem is to treat the difference between the opening RAB and the regulatory tax WDV as a source of “permanent differences”. Permanent differences are any items that are treated as income or expenditure in a firm’s accounting income, but not in its taxable income. In this instance the difference in asset value between the RAB and the regulatory tax WDV gives rise to depreciation that is counted as a deduction from accounting income but not from taxable income. Appropriate treatment of the permanent differences allows NPV equivalence to be achieved. The total value of the permanent differences arising from the difference between the RAB and the regulatory tax WDV is:
PDTOTAL = (RAB0 − WDV0 )t where PDTOTAL is the total value of permanent differences arising from the differences in asset values. The total value of permanent differences can then be amortised over the life of the asset, with the amortised amount added to tax in the calculation of the regulatory tax expense. The right hand side of the formula for calculating the total value of permanent differences is identical to the right hand side of the earlier formula for calculating the opening deferred tax balance. With the difference in asset values accounted for by way of permanent differences, the opening deferred tax balance can then be set to zero. In summary, the implementation of this option requires that: •
Permanent differences are calculated according to the formula immediately above, and then amortised over a reasonable period;
•
The amortised permanent differences are added to the regulatory tax expense; and
•
The opening balance of the deferred tax account is set to zero.
This approach always achieves NPV equivalence, it has the benefit of being transparent, and the calculations are no more complex than those required under the tax payable approach. The attached spreadsheet provides an example of the implementation of this option. Utilising the same assumptions as before the spreadsheet demonstrates that the NPV of cash flows under the deferred tax approach is the same as the NPV of cash flows under the tax payable approach. For a single asset it may be reasonable to amortise the permanent differences over the remaining life of the asset. In practice, however, the firm’s assets will have a mix of ages with considerable variability in their remaining life. The permanent differences could be amortised on a straight line basis over the average remaining life of assets (or any other period), or an accelerated method of amortisation could be used. The exact method chosen will not alter the NPV equivalence result.
OTHER COMMENTS Finally, I note that the Commission’s charts in Appendix A of the Emerging Views Paper appear not to have implemented the deferred tax method using the approach described above. If that is the case, then the chart in the Commission’s Figure A.1 is not correct.
Implementing the Deferred Tax Approach 26 January 2010 Page 5
The chart in the Commission’s Figure A.2 shows the difference between tax depreciation and regulatory depreciation expressed as a percentage of the “regulatory investment value”. If the values in that chart were multiplied by the corporate income tax rate (i.e. reduced to 30% of their current level), then they would approximate the amortised permanent differences to be added to the regulatory tax allowance. It is not clear, however, that the chart in its current form provides any particularly helpful information. Yours sincerely
Andrew Shelley Consultant
Enclosure
