The spreadsheet supplied by NZDF calculates the savings of the current A/B option over the option of retaining the A4s and then later purchasing C/Ds as having a present value (PV) of $351million²⁰. The spreadsheet divides the savings of the F16 deal into two parts [all savings are expressed as PVs in "reals" (ie in today's dollars - after accounting for inflation); but see below for a comment on the methodology used to achieve this in the spreadsheet]:

• The first part of the saving is the benefit of purchasing F16 A/Bs now and selling the A4s (as scheduled), compared with retaining the A4s until their effective end-of-life and purchasing F-16 C/Ds beginning in 2007/08. This saving is calculated to be $160m.

• The second part is the benefit of the lease option rather than immediate purchase. This option is assumed to exist only for the current A/B deal, and is assumed not to be available for the subsequent C/D option. This saving is calculated to be $191m.

The robustness test has taken two forms. Firstly, I have examined the format of the spreadsheet to ascertain whether, on the assumed parameters, the methodology appears reasonable. Secondly, I have altered some of the key parameters to test how robust the results are to changes in key assumptions.

METHODOLOGY

The one slight inconsistency in the spreadsheet is that values are all expressed as reals but a nominal discount rate ("today's" government stock rate, assumed to be 7.29%) is used as the discount rate. To ensure consistency, a real discount rate should be used if prices are expressed as reals. Assuming inflation at the mid-point (1.5% pa) of the RBNZ's target band, this would give a real government stock rate of 5.79%. Using this discount rate gives a PV of savings of $321m (ie $30m less than in the spreadsheet).

There is an additional issue as to whether the government stock rate or the WACC (weighted average cost of capital - assumed to be 10% real in the spreadsheet) should be used to discount future cash flows - i.e. whether we should use 5.79% real or 10% real. Because the nominal government stock rate is between these two extremes, and the results are not too sensitive to choice of discount rate, I have henceforth conducted all calculations with the discount rate used in the original spreadsheet (i.e. 7.29%).

PARAMETER ALTERATIONS

Scenario & Parameter	Initial Value	Changed Value	PV Savings NZ$mill
0. Original PV Savings			351
1a. Real infl rate A/B MLU p.a.	0%	6%	253
1b. Real infl rate A/B MLU p.a.	0%	10%	145
2. A/B disposal value	15%	0%	346
3a. C/D aircraft market value	US$30m	US$25m	273
3b. C/D aircraft market value	US$30m	US$20m	195
4. A4 disposal value in 2007	US$0m	US$30m	328
5. C/D initial costs	US$202m	US$150m	311
6. A/B MLU	US$7.6m	US$15m	257
7. PV of C/D lease option*	NZ$0	NZ$101	250
8. Scenarios 1a+7			152
9. Scenarios 1b+4+7			21

*PV of a lease option offered on similar terms to the A/B lease option to begin in 2007, discounted back to today's value.

MLU Costs

Scenario 1a builds in a real inflation rate for the A/B mid-life upgrade (MLU) at 6% p.a. (i.e. a nominal inflation rate of 7.5% p.a. if CPI inflation is 1.5% p.a.). This is consistent with an approximate doubling of the MLU cost from US$7.6m to US$15m as per scenario 6. These two scenarios calculate the PV of savings for the A/B option at $253-257m.

However a 6% real inflation rate is below the historical real inflation rate for military equipment, which Pugh²¹ has estimated as being 10% p.a. Scenario 1b uses this value for the MLU and calculates the PV Savings at $145m. (This is below the value gained from the lease, demonstrating the importance of the lease arrangements for the robustness of the project savings.)

Disposal Values

Scenario 2 tests for the significance of assumptions regarding the disposal value of A/Bs. Reducing the disposal value to zero has almost no effect on the savings (because the discount rate has already discounted the present value of disposed A/Bs to close to zero in the base case).

Scenario 4 values the disposal value of the A4s at US$30m, which is mid-way between their current estimated value now (US$60m) and their currently estimated value on disposal (US$0m). The effect on the savings is small (a difference of $23m).

Cost of C/Ds

The cost of C/Ds is an important parameter. The base case assumes each C/D will cost US$30m. This is based on best available information.²² However, it does not factor in the possibility that there might again be a "good offer" on F16s (C/Ds) at some stage in the future. Keeping open the decision of when to buy (ie not buying A/Bs now) is the equivalent of having an option to purchase aircraft in future if a good or better offer comes along - at the risk that there will be no such "good offer" in the relevant timeframe.

Scenarios 3a and 3b explore the sensitivity of the results to an ability to purchase C/Ds at prices of US$25m and US$20m each in 2007 (the scheduled date if A/Bs are not purchased now). The PV of savings for the A/B option falls to $273m and $195m respectively.

Alternatively, the price might remain at US$30m but a similar lease arrangement to that now being offered for the A/Bs might be provided. The base case assumes that there will be no beneficial lease option available when a C/D purchase takes place, which could be unrealistic. Scenario 7 factors in such a lease arrangement. It assumes that a similar lease arrangement to that now being offered will become available in 2007 with the same PV of benefits as now being obtained from the lease option. This benefit ($191m in 2007) is then discounted to today's dollars at 7.29% per annum to yield a benefit of $101m. This reduces the PV of savings for the A/B option to $250m.

The spreadsheet also assumes the set-up costs for the C/Ds will be the same as for the A/Bs. However, there may be reduced set-up costs if storage etc has been different. The effect of this is illustrated in scenario 5.

Multiple Scenarios

Scenario 8 combines the possibility of modest (6%) real inflation for the MLU and a lease deal for the C/Ds. Scenario 9 combines stronger real inflation for the MLU plus a C/D lease plus a modest disposal value for the A4s. In both cases the PV of savings remains positive, although in the latter case, the PV of savings is in the order of the potential uncertainty surrounding adoption of different discount rates (i.e. is virtually zero).

CONCLUSIONS TO PART 1

Scenario 9, is possibly a "very-best-case" for the C/D option. Significantly, none of the scenarios indicate that the A/B purchase now is inferior to keeping the A4s and purchasing C/Ds in 2007. This suggests that the A/B option is appropriate unless other factors indicate otherwise. These factors could include the following:

• It is decided that an air strike capacity is not warranted given pressures on other defence elements, with decisions to use the defence budget instead to enhance other force elements (see Part 3).

• It is decided that a smaller air strike capacity could be maintained than currently envisaged in which case options such as rotating the use of A4s may become an option, so delaying the need to purchase an upgraded aircraft.

• It is decided that total defence expenditure should be cut (although government has already indicated that this is not to be the case).

• Other ways of accounting for the purchases (in particular an accrual accounting approach) indicate answers that are not compatible with the spreadsheets that have been provided. Part 2 of this paper examines the purchase in this light.

• Parameter assumptions change significantly with new information, significantly diminishing or removing the currently estimated savings (see Addendum 1).

20 All dollar savings here are rounded to the nearest million. return

21 Pugh, Philip (1993) "The Procurement Nexus", Defence Economics, 4(2), 179-194. return