Fortis holds as part of its investment portfolio so called Structured Credit Instruments (SCI). Structured Credit Instruments are securities, created by repackaging cash flows from financial contracts and encompass asset-backed securities (ABS), mortgage-backed securities (MBS) and collaterised debt obligations (CDO’s). The exposure on Structured Credit Instruments is in the Consolidated Financial Statements reported by category of financial instrument and as such included in Investments available-for-sale, Investments held at fair value through profit or loss, Assets held for trading, Due from Customers and Other assets. The exposure includes the US subprime residential mortgage-related assets.

At 31 December 2007, the net exposure on the global Structured Credit Instruments can be detailed as follows:

Other includes mainly the asset pools of Scaldis, as reported under Other assets in Banking. Scaldis is fully consolidated within Fortis and is a conduit that purchases eligible assets from investment grade, non-investment grade and unrated sellers. The asset pools contain continuous financing of third party clients’ assets such as consumer and auto loans, trade receivables, mortgages and lease receivables.

Special Purpose Entities (SPE) assets relate to investments by entities of Fortis in debt securities of Special Purpose Entities set up by Fortis for the purpose of asset securitisation or structured debt issuance and are included in the Fortis consolidation scope. The exposure of Special Purpose Entities is mainly reported under the balance sheet captions: Due from Customers (Residential Mortgages) and Due from Banks (Reverse repurchase agreements).

On 11 January 2006, the European Commission endorsed IFRS 7, Financial Instruments: Disclosures. IFRS 7 is applied by Fortis as from 1 January 2007 and includes requirements to disclose whether fair values of financial assets are determined, in whole or in part, directly by reference to published price quotations in an active market or are estimated using a valuation technique based on assumptions that are or are not based on available observable market data.

Structured Credit Instruments are mainly measured at fair value. The Structured Credit Instruments not carried at fair value are mainly related to the asset pools of Scaldis and are measured at amortised cost. The fair value measurement of financial assets, part of the Structured Credit Instruments, can be categorised based upon the valuation methods applied:

• category 1: fair values determined in whole or in part, directly by reference to published price quotations in an active market
• category 2: fair values determined, in whole or in part, using a valuation technique based on assumptions that are supported by available observable market data
• category 3: fair values determined, in whole or in part, using a valuation technique based on assumptions that are not supported by available observable market data.

The categorisation within the fair value hierarchy is based upon the lowest level of input that is significant for the fair value measurement. Note 36 ‘Fair values of financial assets and financial liabilities’ contains a description of the valuation methodologies applied for the measurement of the fair value.

The following table presents the financial instruments measured at fair value included in the Structured Credit Instruments by category of fair value measurement, indicating the transparency of the inputs to measure the fair value as at 31 December 2007:

The percentages reported are based on the net exposure including the assets held by Special Purposes Entities of Fortis.

During the second half of 2007 the market circumstances for Structured Credit Instruments changed dramatically leading to a significant decrease in the observability of the data used for market pricing of assets primarily related to US subprime residential mortgage-related assets, including Asset-Backed Securities (ABS) and Collaterised Debt Obligations (CDO’s). Under these market conditions, it was no longer possible for Fortis to value the CDO origination portfolio based on published price quotations as those were no longer available. Consequently, a valuation technique, based on a discounted cash flow model, was applied.

As for other financial assets, Fortis applies a two step approach in the impairment testing process of financial instruments. Firstly, an assessment is made whether objective evidence exists that a financial asset is impaired, followed by the recognition and measurement of an impairment loss. The assessment of objective evidence is based on observable data (‘triggers’) about loss events.

The Fortis established mandatory triggers concerning CDO products that can lead to an impairment, are the following:

• for all tranches Fortis owns, any Event of Default under the indenture for the issuing CDO
• the downgrade of the tranche to a non-investment grade.

If a mandatory trigger is met, the instrument is considered impaired. Besides the mandatory triggers four judgemental triggers are used:

• overcollateralisation falls below 100%
• interest coverage falls below 100%
• downgrade of any tranche with 1 or more notches below the original rating
• fair value drops below 80% of the acquisition price.

If a judgement trigger is met, further credit quality analysis is undertaken. Impairment losses are measured as the difference between the asset’s carrying amount and the present value of the estimated future cash flows.

During 2007, the process to estimate impairment for the so called Super Senior exposure originated by Fortis, was based on a cash flow analysis considering observable and non-observable data for relevant benchmark instruments, implied cumulative losses in mortgage pools and the likelihood of events of default in the underlying ABS-CDO collateral. The inputs to the valuation model principally comprise remittance data from mortgage service companies. These are received towards the end of each month and relate to the preceding month’s cash flows on the mortgages underlying the relevant mortgage-backed securities. The model assesses the level of risk in the underlying mortgage portfolio and estimate the fair value of the positions.

The model is based on estimates of the cumulative losses for the different buckets (RMBS and CDO’s) included in the different CDO structures. The model considers a value for CDO buckets and gives a valuation of the RMBS buckets based on the following steps:

• computation of the default pipeline for each portfolio
• extrapolation of Cumulative Default Rate of the mortgage pool over its lifetime using a default curve based on market data
• computation of the Cumulative Loss with loss given default based on recent market information
• calculation of the net present value of the cash flows.

Fortis, however, estimated that the outcome of the model based on historical market data did not reflect the rapid deterioration of the US subprime market. Consequently it has been decided to consider a stress scenario using various stress testing assumptions, including a higher cumulative loss than the historical market data to determine the level of impairment.

Although the Fortis stressed cash flow model is calibrated to current market transactions as measured by the most readily available proxy information that would be used by other market participants, such models have inherent limitations, and different assumptions and inputs generate different views on the level of impairment. The impairment losses on Structured Credit Instruments are reported in the income statement under the caption Change in Impairments (note 46).

Fair value changes of Structured Credit Instruments reported as Assets held for trading or Investments held at fair value through profit or loss are reflected in the caption Other realised and unrealised gains and losses of the income statement (note 41).