Section 9 - Sensitivity Analysis of Discount Rates

One of the complications in comparing the various reports is that they use different discount rates for calculating net present values. Sofregaz originally used 7.5 per cent, though in its later calculations it uses 7 per cent, BG/KeySpan use 8 per cent, while the Shannon-McManus report uses 7 per cent. The test rate recommended by the Department of Finance for non-commercial public sector capital projects is 5 per cent, though in the case of commercial projects which may involve higher than average risk, a somewhat higher rate may be appropriate. Evaluations carried out on the first interconnector have used 7 per cent.

That said, the discount rate is meant to represent the cost of capital or the required rate of return for a project. In the low inflation and interest rate environment that exists at the moment, both the cost of capital and the required rate of return on investments in the general economy have been falling. This would suggest that the discount rate used to evaluate this project should also be lower than that used in the past. The real rate of return on Irish State index-linked bonds has recently been around 2.7%.

UK regulators have recently been indicating that permitted real rates of return for the various industries subject to tariff controls may be in the range 5 to 6 per cent. There has as yet been no determination in Ireland of the rate which will be permitted, although it is clear from the Electricity Regulation Bill currently before the Oireachtas that there will certainly be a ceiling. At this stage, our judgement is that it could well be lower than the 7 and 8 per cent real rates used in the various studies listed.

Where the timing of expenditure differs between investment options, the discount rate chosen can have an impact on the determination of the least cost option.

Therefore it is important to investigate the sensitivity of the various options to changes in the discount rates used. We have done this, for the three reports in question, for discount rates of 5, 6, 7, and 8 per cent.

We consider here the revised data produced by Sofregaz, as opposed to that produced for the Gas 2025 report.

As can be seen, UK1 is the least cost option, followed by UK1 S-N, and UK2. Discounting, and the choice of discount rate within the range considered here, make no difference to this ordering, but affect the size of the gap.

This is an interesting case, because for the central demand case and a discount rate of 8 per cent, while the Scotland-Dublin route is cheaper in Capex terms, it is more expensive in NPV terms. It follows, that at some discount rate, the Scotland-Dublin route is cheaper.

This rate cannot be determined with accuracy, because the report only gives 5-yearly capex and Opex figures, and these only for the central demand case. BGE would not provide us with yearly figures, citing confidentiality reasons.

Without yearly figures it is impossible to carry out a proper sensitivity analysis. However, we did carry out an exercise based on the 5-yearly costings in the report, for the central demand scenario.

Subjecting this imputed expenditure schedule to a range of discount rates indicates that the BG/KeySpan and Scotland-Ireland options have an equal NPV at a discount rate just below 6 per cent. Any rate lower than this will favour the latter route.

A different expenditure schedule, for example where expenditure happened in the middle of each five-year period, or was spread evenly throughout each period, would further favour the Scotland-Ireland route.

This reflects two features in the BG/Keyspan report –

the differences in NPVs are modest, so small changes in the make-up of the figures can have an impact.

The Scotland-Ireland route is cheaper in Capex terms under all demand scenarios, despite being more expensive in NPV terms in all but the high demand case.

The lower the discount rate used, the more closely the NPV figures will reflect the Capex figures. As a result, it is not surprising that as the discount rate was reduced the option with the cheapest Capex was favoured.

In the Shannon-McManus report, BG and Sofregaz revisit their calculations, using an agreed databook of assumptions. Both come up with the same ordering of capex for three infrastructure options, equivalent to UK1, UK1 S-N and UK2 in the Gas 2025 report. When a discount rate of 7 per cent is applied, the NPVs are in the same order.

Recalculating the NPVs using discount rates ranging from 5 to 8 per cent does not change the order of the projects.

9.5 Summary

Within the range considered above, only the BG/KeySpan ordering was sensitive to the discount rate used. Lack of detail in this report prevents a proper analysis. However, for their central demand case, and using the timing assumptions we have made, it appears that a discount rate above 5.7 per cent favours the North-South pipeline option, whereas below that rate twinning the Scotland-Ireland interconnector is favoured. Altering the timing assumptions would tend to favour the Scotland-Ireland interconnector.

In all other cases, the ordering of the NPV calculations were robust with respect to the choice of discount rates.

We would note that the gas pipeline industry and the powergen industry are capital intensive, and both will be regulated under Irish legislation. The determination of the allowable real rates of return for these industries will no doubt be an early priority for the regulatory system, and such a determination would help to clarify the uncertainties deriving from this source.

Finally, there is a potential complication, which we merely note here, in that import interconnectors would fall under regulatory review in both Ireland and the UK.

To section 10