OREANDA-NEWS. Fitch Ratings has published an Exposure Draft where it states its proposals to amend its Covered Bonds Rating Criteria with the aim of making some rating steps more focused on the most relevant credit aspects. Fitch invites feedback from market participants until 29 July 2016. Click the link at the end of this release for full report.

The agency is proposing to modify the different uplift factors above an issuing bank's Long-Term Issuer Default Rating (IDR) leading to the potential maximum covered bond rating. These comprise the IDR uplift, the payment continuity uplift (PCU) and the recovery uplift. Fitch is also proposing new refinancing spread level (RSL) assumptions for mortgage and public-sector cover pools, making them more comparable between asset types and jurisdictions.

Under the proposal, an IDR uplift of up to two notches can be assigned to programmes in jurisdictions with advanced resolution frameworks and where fully collateralised covered bonds and secured debt are exempt from bail-in. This also applies where Fitch believes payments will continue to be made without recourse to the cover pool even if the issuer has defaulted on its senior debt. However, for this to happen, the risk of undercollateralisation must be sufficiently low at the point of resolution, in Fitch's view, based on legislative, contractual and programme-specific safeguards. See Special Report "Fitch's Jurisdictional Analysis of the Risk of Undercollateralisation of Covered Bonds - Excel file", also published today.

Fitch is proposing to replace its Discontinuity Cap (D-Cap) assessment by the PCU concept. The latter will focus on assessing whether, once recourse against the cover pool is enforced, the liquidity mechanisms are sufficient to protect against payment interruption risk. This risk derives from maturity mismatches between the cover assets and the covered bonds.

The PCU is expressed in notches above the IDR adjusted by the IDR uplift, ranging from zero to eight. As a benchmark, programmes with strong liquidity provisions, such as a 12-month maturity extension, will be eligible for a six-notch uplift in developed markets. Programmes with a conditional pass-through amortisation will continue to have a potential uplift of eight notches. Other threats to payment continuity, stemming from asset segregation or alternative management, could reduce the assigned PCU if they are identified as a high risk.

Fitch is proposing to move towards a loss-driven assessment of recoveries given default. This is a change from current approach, where calculations may suggest a higher level of precision than that implied by the recovery ranges used by Fitch. We expect that fully collateralised programmes secured by standard assets, such as mortgage loans and public-sector debt, should be able to generate at least a good level of recoveries and will be eligible to a one-notch recovery uplift in all rating scenarios. Programmes where overcollateralisation (OC) given credit to by Fitch in its analysis offsets at least the stressed credit loss levels implied by the agency's static model output are expected to experience outstanding recoveries. These will be eligible to a two-notch recovery uplift, or three notches if the tested rating on a probability of default (PD) basis is in non-investment grade.

RSLs are used by Fitch to derive a sale price for assets to cover any shortfall of funds for covered bond payments once recourse against the cover pool has been enforced. Fitch is proposing to base sovereign RSLs on a through-the-cycle analysis of observed five-year sovereign bond spreads and a qualitative assessment of the bonds' liquidity such as the sovereign's reserve currency flexibility (RCF) and the size of the government bond market. For other assets, Fitch is proposing to derive RSLs through a benchmarking exercise, as the market generally prices other assets based on the spread of the sovereign bonds.

The proposed RSL assumptions for public-sector assets located in countries rated 'A' or below are substantially lower, following the removal of certain credit risk premium embedded in the peak observed spreads. On the other hand, sovereign RSLs proposed for some countries in the high-investment-grade category will be higher as a result of the qualitative overlay assessment, particularly in those countries without RCF or with a small-sized sovereign bond market on a relative basis. Existing and proposed RSLs can be found in "Exposure Draft: Fitch's Cover Assets Refinancing Spread Level (RSL) Assumptions - Excel file".

Fitch publicly rates 130 covered bond programmes, of which 111 are mortgage programmes, 16 are public-sector programmes and three are secured by other collateral. A sample of 68 covered bond programmes (61 mortgage, five public sector and two others) was tested for the proposed changes. Out of the sample, 23 are likely to be upgraded and 10 are likely to be downgraded.

The possible upgrades are mainly from low-investment-grade (Italy, Spain and Ireland) and sub-investment-grade (Portugal and Greece) countries. Upgrades are also likely among UK programmes not already rated 'AAA'. The magnitude of upgrades is limited to one rating category, with the exception of one Italian programme, which could be upgraded by five notches. The main drivers of the upgrades are either due to a higher IDR uplift or a higher PCU, or a combination of both.

Of the tested programmes, 48 would benefit from a two-notch IDR uplift as opposed to 16 now as Fitch would take a broader view on European issuers eligible for a covered bond IDR uplift. Soft-bullet programmes in Ireland, Italy and the UK would become eligible for a six-notch PCU, recognising the proposed emphasis by Fitch on liquidity provisions to determine the maximum notches of PCU.

Ten programmes could be downgraded in the high-investment-grade countries. Among them, two would be affected by the removal or lowering of the IDR uplift as their issuers have support-driven IDRs; two are programmes that Fitch deems more at risk of undercollateralisation in the event of resolution; and one would be downgraded due to the removal of the recovery uplift due to a combination of currency mismatches and OC constraint. A further five programmes could be downgraded due to the new recovery approach as the breakeven OC for their rating, which would be based on a higher tested rating on a PD basis, would be higher than the OC currently relied upon by Fitch in its analysis. These downgrades are likely to be limited to one or two notches.

Significant decreases in breakeven OC are likely for some programmes rated by Fitch on a recovery basis only, notably where the PCU is zero, such as for certain Portuguese programmes under their current documentation and all Spanish programmes. Programmes of banks with an IDR of at least 'A+' and which benefit from an IDR uplift increased to two notches could also achieve a 'AAA' rating on a recovery basis. Increases in breakeven OC are likely if RSLs rise (for instance in some highly rated countries) or if credit for recoveries decreases to one from two notches (for instance for programmes largely exposed to assets in a different currency than the covered bonds' denomination).

Among the remaining 62 programmes not included in the tested sample, Fitch does not expect further upgrades. This is because the tested sample was selected to include all non-AAA-rated programmes that could be upgraded due to the combined proposals relative to the IDR uplift, the PCU and the recovery uplift. The changes could lead to an increase of the breakeven OC for the current rating of programmes that have not been formally tested. However, the agency believes further OC-related downgrades are unlikely, because many programmes maintain a higher OC than the breakeven OC for their current rating and issuers can generally increase collateral to secure their covered bonds.

If the exposure draft proposals are fully implemented in the current form, about half of the 24 existing Multi-Issuer Cedulas Hipotecarias ratings could be upgraded by one notch.

Fitch invites feedback from market participants on the proposed amendments to Fitch's Covered Bond Rating Criteria. Comments should be sent to Criteria. Feedback@fitchratings. com by 29 July 2016. Fitch will publish on its website any written responses it receives, in full, including the names and addresses of such respondents, unless the response is clearly marked as confidential by the respondent. Following the review period and consideration of responses received, Fitch expects to finalise and publish the criteria in 3Q16.