| Our guidance on applying the model that accompanies the Construction Risk Methodology has not changedsince the issue of our December 2008 Special Comment, but for completeness is summarised again here. |
If CCs or grants7 were not included as an input, the model8 would assume that any shortfall between the senior debt amount and the overall the contract cost was funded entirely by loss-absorbing equity, which would result in a stronger senior debt rating - this is inappropriate as a standard assumption because CCs do not generally have loss-absorbing equity-like characteristics.
Instead, the model prompts users to input amounts for CC as well as senior debt; it then treats the CC as an additional debt-like facility. Under "Neutral" treatment, the senior debt and CC facilities rank pari passu, with the same expected loss for each. Under "Protective" treatment, the CC is assumed to rank junior to the senior debt, and absorbs a proportion of the losses that would have otherwise have accrued to the senior debt; under "Adverse" treatment, the opposite applies, with the senior debt bearing a greater level of losses. In each case, the model shows the aggregate expected loss and rating equivalent - before apportionment between CC and senior debt - and then shows how this total is split between the two. A notional rating for the CC is also shown so that users can observe the zero-sum game as the aggregate overall loss is either shared equally, or borne disproportionately by either the CC or the senior debt.
Where CCs are small (say 5% of the overall funding package), the effect will typically be close to neutral and should be modelled in this way. At such modest levels it is likely that qualitative considerations would be more important than a mechanical quantitative adjustment.
For more material CCs best practice would be to run the model first assuming a neutral effect; and then to make a qualitative assessment of the likely impact on the senior debt's expected loss position and rating, informed by the effect of the protective or adverse model switches as appropriate.
________________________________________________________________________________________________________________________
7 We have not changed the label in the model from "grant" to "Capital Contribution", but may do so in a future release.
8 The methodology provides details of how users can obtain access to the web-based model that accompanies it.