It exists.
And it’s detailed in the first ever annual report from the UK’s Asset Protection Agency (APA).
It covers the period from December 2009 to the end of March 2010, and features some great bits of detail on the UK government’s Asset Protection Scheme (APS), which saw HM Treasury promise to cover some of the potential losses on £282bn worth of Royal Bank of Scotland assets.
But why not let APA describe the thing? From the report:
The APA is of the opinion that the APS meets all of these requirements for the following reasons:
- • the value of the APS is driven by the underlying credit risk of the assets in the Scheme, very much like a synthetic collateralised debt obligation (CDO), the credit risk of those underlying assets is therefore the ‘underlying’ in the APS;
• in order for HMT to get exposure to that underlying risk without using a derivative it would have had to purchase the underlying assets / risk in a funded transaction, i.e. an asset purchase scheme, clearly when compared to that alternative the APS requires no such initial net investment; and
• the APS is settled at a future date, potential payouts will occur at future dates should the First Loss Amount be exceeded and has a contractual maturity of 2099.
That’s from the rather interesting discussion (on page 35) of how the APS is valued for HMT’s accounts. Interestingly, the whole thing seems to be counted as a derivative (as opposed to say, an insurance contract — something which was discussed by the APA), for the reasons outlined above.
And as a derivative, the APS is valued at fair value. The Agency is even using a Gaussian Copula to determine that value — albeit one that has been slightly tweaked to take into account recent (crisis) events:
The assets covered by the APS are mapped to credit default swap curves derived from observable market information, which are used to derive default probabilities. The Gaussian copula models the joint default behaviour of the underlying assets. Stochastic recovery modelling, a relatively recent innovation developed in response to the recent market turmoil and now a broadly accepted industry standard, allows the model to be calibrated to the market prices of traded CDO tranches.
A Monte Carlo engine simulates default times for each asset and recoveries in different scenarios. Cash flows due to each party in each scenario are discounted and the results averaged to determine the net present value of the position.
The valuation model reflects the delay incorporated into the APS from the time that a loss is incurred on an asset to any protection payment by discounting protection cash flows. It also reflects the exit option, which is not a standard feature of CDOs. It assumes that RBS will not exit the APS before four years have elapsed, which corresponds to total fee payments of £2.6bn, which is £100m higher than the minimum required payment of £2.5bn. At that point, it will remain in the APS only if the expected value of future protection is greater than the value of future payments. In effect, it is assumed that RBS will have perfect foresight at that point.
There’s a (snarky) joke in there but we shall resist, for once.
Moving swiftly on, there’s some good news for UK taxpayers. The APA expects total losses from the scheme to come in at £57bn — just below the £60bn threshold at which HMT has to start sharing the bank’s losses. Which means, once you include fees RBS is paying, the taxpayer could make a profit.
And here are a few other interesting tidbits from the report.
Some criticism of RBS’ methodologies, as picked up by today’s FT:
The APA has expressed concern over the quality of credit memos and financial analysis received from RBS and the consistency with which RBS manages the Covered Assets in accordance with the AMO. The APA continues to discuss this with RBS and to monitor the position.
And some wrangling over £2bn of structured finance transactions:
The covered amount at 31 March 2010 includes approximately £2.0 billion of assets in the derivatives and structured finance asset classes which, for technical reasons, do not currently satisfy, or are anticipated at some stage not to satisfy, the eligibility requirements of the APS. The APA and RBS continue to negotiate in good faith whether (and if so, to what extent) coverage should extend to these assets. Also, the APA and RBS are in discussion over the classifications of some structured credit assets and this may result in adjustments to amounts for some asset classes; however underlying risks will be unchanged. Whilst good progress is being made, the final outcome is dependent on RBS and the APA reaching an agreement by the due date on various areas of interpretation. Should this not be achieved and the APA does not grant an extension to RBS, cover on these assets may be restricted.
And finally a personal update on the welfare of the Agency’s staff :
There were no recorded working days lost to sickness absence in the reporting period.
A healthy bunch at the APA.
Shame the RBS assets have been so sick.
Related links:
Agency hits at RBS for handling of risky assets - FT
Taking out the trash at RBS – FT Alphaville
Revision and APS ambiguity at RBS – FT Alphaville