When I appeared before the Committee on 18 November 2002, Mr David Rendel MP questioned me about the extent to which the robustness of the financial structure of National Air Traffic Services (NATS) had been tested for possible reductions in air traffic. I promised a note of explanation. This memorandum also covers the supplementary questions sent to me by the Committee. It has been discussed with the National Audit Office.
In the course of responding to Mr Rendel, I questioned figure 21 in the Comptroller and Auditor General's (C&AG's) Report. That led to some confusion, which I hope I can now clear up. The figure shows the impact of a number of tests to which the Airline Group bid was subjected and three further tests run by the National Audit Office (NAO) in the course of their examination of the public private partnership (PPP). The NAO tests were based on the oil crises of 1973 and 1979 and the Gulf War of 1990. The figure is calibrated in chargeable service units (CSUs). I pointed out that CSU data were not available in respect of the period of the two oil crises. I was not aware that in running their three tests the NAO had estimated the number of CSUs for the critical years from the available data for the total number of flights in UK airspace. The NAO point out that they had discussed their approach with the relevant experts at NATS, and officials in the Department did not query it. As I said at the hearing, I accept that.
The C&AG's Report rightly says that the Airline Group bid was subjected to nineteen tests of its financial robustness. Nine mandatory tests were applied to all bids. The final Airline Group bid was subjected to a further 10 tests. Whereas the mandatory tests addressed the impact of various kinds of long-running under-achievement, the additional tests focused on more specific, sharper shocks. The latter are the tests illustrated and briefly described in the first part of Figure 21 in the C&AG's Report. All 19 tests are listed in the Annex to this memorandum.
The tests were designed to examine a range of realistic possibilities to see how robust the financing was against plausible forecasting errors. The Airline Group bid passed all the tests, and this contributed to the conclusion that it offered a sufficiently sound financial structure. The tests were not designed to identify a set of events which would have made the company unable to meet its debt obligations. Given the assumptions used for the testing, that set of events would have had to be so extreme that the issue was not worth pursuing. It is therefore not possible to say from the tests what would have been the least disastrous scenario which would have pushed NATS into default.
The point of figure 21 is to demonstrate that there were historic events-notably the oil crises of the 1970s- which were not replicated in the tests to which the PPP bids were subjected. From the modelling which underlies figure 21, the NAO concluded that "given a reduction in traffic on the scale experienced in both of the oil shocks NATS would have been unable to meet its debt service obligations" (C&AG's Report, paragraph 3.27). The Department did not challenge that statement when commenting on the draft Report. We did question whether events so long ago were a good guide to what might happen in the future, but the NAO were keen to keep the comparison in the report and we did not press the point. At that time we did not ask for, and so were not offered, access to the NAO's modelling work. The C&AG's Report, which we agreed to in the normal way, presented the conclusions arising from this analysis.
We have now seen the NAO's modelling work. In the light of what we have seen, it is clear that in their three supplementary tests the NAO used the same model as was used in testing the Airline Group bid, but applied it differently. There were some differences in the way in which the year-on-year traffic comparisons were calculated. But the key difference is that the original range of scenarios assumed that the economic regulator would allow NATS' prices to increase if there were to be a major and sustained downturn in traffic, whereas the NAO tests did not.
The NAO comment that they consider it inappropriate to include assumptions about possible amelioration in stress tests, which they believe should be performed with all other things remaining equal. In this particular case, the NAO did not make an assumption that the regulator would make adjustments, and regard such an approach as questionable. In the original tests, on the other hand, the assumption was made (by the Department and its advisers, and by the bidders) that in each quinquennial review of NATS' charges the regulator would take all factors into account, including any material adverse events which had occurred during the preceding period. The financial regime was therefore regarded as sufficiently robust if it could withstand foreseeable shocks for the remainder of the control period in which the shocks occurred. Which methodology is correct is a matter of judgement.
One of the tests applied to the Airline Group bid was a reduction in traffic in Year Six of the PPP, that is, the first year of the second control period. This test was chosen because Year Six was the most difficult year for NATS' cash flow. The ability to withstand a shock in that year and to continue to service debt throughout the remainder of the second control period was therefore seen as the key test of the robustness of the bid.
The NAO say that they recognised that Year Six would be a difficult one, but noted that-given their assumption of no intervention by the regulator-the effects in this test would be mitigated by the previous five years of compound growth and by the accumulation of cash reserves. They say that this was a factor in leading them to look at other scenarios.
The C&AG's Report correctly states that the Airline Group bid was not explicitly tested against the possibility of traffic reductions as severe as those which occurred in the oil crises. However, if the methodology used for the original testing is now applied to an oil crisis scenario, the Airline Group bid would withstand such a test. That is to say, the test would have shown that NATS could continue to service debt throughout the first control period, after which the regulator could intervene to compensate NATS' for the loss of traffic in subsequent periods. The NAO methodology, on which the bid failed the test, was more pessimistic: it did not assume any adjustment by the regulator at the end of the first control period, and as a result it showed a default in later control periods.
There is a duty on the regulator under the Transport Act 2000 (among other duties) to ensure that NATS does not find it unduly difficult to finance its activities, and there is also a provision in NATS' operating licence allowing the price cap to be re-opened in exceptional circumstances. Given these facts, the Department considers that it was reasonable to assume that the regulator would adjust at the end of each control period for unusual events-adverse or favourable-in the preceding period. On that basis, the NAO methodology might be considered unnecessarily cautious. But we accept that the regulator had not at the time issued a statement of the regulatory policy it intended to adopt. In the absence at the time of such a statement, the extent to which one could assume intervention by the regulator was a matter of judgement.
The C&AG's Report comments on the fact that this regulatory uncertainty obtained during the bidding process. The report says that it was difficult for bidders to proceed without sure knowledge of the level of prices which NATS would be permitted to charge. I accept and support that comment. Among other things, it would have been helpful to have had a clear statement at an early stage of the regulator's intended approach for the second and subsequent control periods. This would have enabled the testing of the financial structure to have been conducted with greater confidence. Whilst the regulator had its reasons for delaying its regulatory statement in the present case, there is a lesson here for regulated businesses in the future.
December 2002