Solvency II, IORP and real estate

The general concern for the real estate industry is how insurers and possibly pension funds will view real estate as an investment asset once Solvency II is in force, and in particular, the impact of the capital cost of investing in real estate relative to other asset classes. Key to this is how they will be required to model the capital that they will need to hold in respect of potential future falls in value of their investments.

Insurers can either use a “standard formula” for which the risk factors are set down by the EU regulator, the European Insurance and Occupational Pensions Authority (EIPOA) or seek approval from their national regulator to use their own internally generated risk model. Much of the real estate industry attention has focused on the amount of capital, which insurers need to hold to cover the potential fall in the value of their real estate investment assets under the “standard formula”. Currently the proposal is that the shock to be applied to direct real estate investments is 25%, i.e. insurers should hold capital equivalent to a write down in the book value of their real estate investments by 25%, from a starting point that the assets are carried in the books at market value. In April 2011, the Investment Property Databank (IPD) published a study focusing specifically upon real estate. This study was funded by a consortium of seven key real estate trade bodies. It identified a number of areas of concern, but specifically in respect of the property shock, the report suggested that 15% was a more accurate reflection of pan-European volatility than the 25% suggested by the regulator. Since publication, the IPD has updated its analysis, continuing to support the view that the currently proposed property shock is excessive. Coverage in the real estate press has stated that the EU Commission has not been swayed by the lobbying and that the 25% market shock remains in the proposals. The general view appears to be that there is little opportunity for further change in the basic shock under the standard model and that the industry is better served by concentrating on insurers’ own internally generated models and on areas of detail such as the treatment of real estate funds and real estate lending.

As mentioned above, insurers can either use a “standard formula” for which the risk factors are set down by the EU regulator or seek approval from their national regulator to use their own internally generated risk model. In the case of UK insurers, this regulator is the FSA. There is also the possibility to go for an option that allows a model that uses some aspects of the standard formula within an internally generated model. There are a number of areas of uncertainty, not least of which is that delays to the progress of Omnibus II, with its consequential impact of the level 2 regulations and level 3 guidance, are delaying the point at which the FSA can give approval for models until “some point in 2013”. Given the Commission has announced that the deadline for transposition will be moved to 1 July 2013, this is expected to be in the second half of 2013.

For insurers to use a different volatility adjustment from that enshrined in the standard model, they would need to provide the FSA with compelling evidence that the mix of assets in which they have invested has a volatility that is different from that under the Standard Model. This poses a particular challenge for insurers investing in UK real estate. The 25% market shock set out in the current draft regulation is derived from the IPD UK Monthly Index and is therefore a measure of UK real estate volatility. Showing a demonstrably different historic volatility is likely to be challenging for UK insurers.

The most immediate concern for the real estate industry is the treatment of real estate lending by insurers. With the pressure on banks causing them to retreat from real estate lending, many in the industry had hoped that the attractive risk-adjusted returns for property senior lending would attract insurers into the market. As widely covered in the real estate press, a number of large insurers have established teams to allow them to expand into lending. Fund managers looking to raise debt funds were also targeting insurers, potentially opening up the opportunity to those too small to set up their own teams. Unfortunately, the treatment of real estate lending under Solvency II, which had been looking attractive, is now significantly less appealing and less logical. The draft implementing measures for Solvency II dated 31st October, which have never been formally published but which have been widely circulated introduce a significant change to the proposed treatment of commercial real estate lending within the standard formula. In earlier draft provisions, a specific treatment of property loans was included that took account of the value of collateral, using the property shock to adjust the value the collateral. This provision is now restricted to residential mortgages. Under the 31st October draft, other real estate lending has been removed from the counterparty default risk module and included in the general provisions for corporate bonds under the spread risk module. The starting point under this provision is a credit rating by a nominated credit rating agency. This does not reflect normal commercial practice in property lending as individual commercial real estate mortgage loans are not rated in this way. Furthermore, it is not clear how collateral should be taken into account. In the absence of any clear provision that would allow collateral to be taken into account, the assumption is that it should be ignored. If this interpretation is correct, it means there is no distinction between an unrated real estate loan secured by a mortgage over a commercial property and an unrated unsecured loan, which would seem to be an odd place for the regulation to end up.

Bonds and loans for which a credit rating is not available are assigned a risk factor, in an example of presumably unintended humour, termed by the regulator “F^up”.

As mentioned above, the directive itself does not deal with the market shocks which are covered in the level 2 regulation. The detail of the level 2 regulations will need to be brought forward after Omnibus II is finalised. In view of the already very tight schedule between now and insurance companies needing to be fully in compliance by 1st January 2014, there will not be much time for debate and discussion.

If this unfavourable treatment of real estate lending is adopted, what does it mean in practice? Smaller insurers using the standard model will clearly find real estate lending other than residential mortgages less appealing than under the previous iterations of Solvency II. For larger insurers seeking approval for their own models, the position is less clear. What will the regulator accept and approve?

Although the treatment of debt is clearly a major concern for those looking to launch debt funds, it has a broader impact for the real estate industry as a whole if the availability of senior debt is curtailed.

There is more positive news for real estate investment managers regarding the treatment of real estate funds. One of the major concerns had been whether real estate funds would be treated as transparent or opaque. Real estate funds cover a broad spectrum; the nature of the investment vehicles varies considerably, as does the way in which they are financed, the level of gearing and indeed the nature of the underlying investments. For an open-ended vehicle, with low levels of gearing and core real estate as the underlying asset, the transparent treatment would seem most appropriate. At the other end of the spectrum, a closed-ended real estate private equity fund with high levels of gearing and underlying assets with significant operating risk, for example hotels, is difficult to distinguish from any other form of private equity fund. Choosing either approach and applying it to all real estate funds would be highly inappropriate for one end of the spectrum or the other. Defining some form of segmentation through regulation is unlikely to be successful. The treatment now proposed is that the default is that funds are treated as transparent unless this is not possible. Some clarification is still needed as to who decides what is “not possible” and the criteria to be used. Generally, however, the outcome seems to be a pragmatic one.

Somewhat perversely, a reduced attractiveness of real estate lending may increase the attractiveness of direct and indirect equity investment in property. In general, the capital cost is likely to make insurers look more closely at their returns from investments. Although moving to lower risk assets to reduce the market shock is superficially attractive, in the longer term the lower returns will be a deterrent.

Aside from the fundamental question as to whether Solvency II and similar legislation will change the behaviour of insurers and pension funds in the way they perceive real estate as an asset class, fund managers and others will also need to address the reporting implications. This is again an area of uncertainty. EIOPA published its “Consultation Paper on the proposal on Quantitative Reporting Templates” and “Draft Proposal for Guidelines on Narrative Public Disclosure & Supervisory Reporting, Predefined Events and Processes for Reporting & Disclosure” on 8th November last year. This was a consultation that ran until late January, since when EIOPA has been considering the feedback received and expects to finalise the package in the summer. Although not the most snappily titled documents, they are a significant step forward in the process that will determine the public and supervisory reporting obligations of the insurers, which will in turn determine the level of detail of reporting at the fund level. As indicated above, the assumption is that funds will be treated as transparent so reporting will need to be at the level of the underlying investments of the funds, although as also discussed above, some real estate funds may well end up being treated under the equity shock method,which would imply less granularity of reporting at the asset level. Under Solvency II, insurers will have to demonstrate to their supervisors that the data they use is sufficiently complete, accurate and appropriate for their specific needs. External data, including information from asset managers, will need to meet the same standards of quality, detail and verification as internally sourced information. The key word is ‘demonstrate’. Insurers will expect documented assurances and other evidence from their asset managers that the quality, consistency and reliability of the risk information they supply, and the governance and control procedures that underpin this, are up to scratch.

Apart from the direct requirements of the regulator, there are two further drivers for greater granularity of reporting:

• Many of the larger insurers will be using their own internally generated models rather than the standard model. The level of detail that they require will be determined not only by the minimum standards set by the regulator, but also by the specific requirements of their modelling;
• One of the consequences of the capital cost of investing in real estate as an asset class can be expected to a greater focus on the returns achieved. Investing in poor performing real estate that only gives a bond level of return is punished under Solvency II. Insurance investors should become more thorough in understanding and evaluating performance and underperformance. To do this, one suspects that theywill become even more demanding of information.

As mentioned above, there is also progress towards Solvency II type regulation for pension schemes. In this case, the relevant directive is IORP. In April 2011 the European Commission asked EIOPA for advice on the EU wide legislative framework for IORPs. The starting point for this was the need to accelerate progress towards the key objective of the regime, which is to increase the number of pan-European pension funds from its current low level. The publication of EIOPA Final Advice in February 2012 followed a period of consultation, the final stage of which ran until 2 January 2012 and was commented upon by 170 stakeholders from 14 Member States and 20 European and international organisations.

Although the recommendation of EIOPA is that any amended IORP should also cover defined contribution schemes, the main impact for the real estate industry would arise if defined benefit schemes changed their investment preferences as a result of the regulatory changes. For pure defined contribution schemes, investment risk is passed through to the member. The immediate challenge for the real estate industry would arise if a market shock approach as adopted for insurance companies was also adopted for pension funds and as a result discouraged defined benefit pension funds from investing in real estate. Pension schemes are typically smaller and less sophisticated than insurance companies and would struggle to produce the internally generated models anticipated for insurance companies.

The EIOPA advice to the Commission is somewhat ambivalent as to the route forward. EIOPA proposed a holistic balance sheet approach which would allow the pension scheme to take account of external factors such as the sponsor covenant. However, it wass unclear as to how market risks would be considered on the asset side of the equation and EIOPA proposed that a quantitative impact assessment as was undertaken for Solvency II should also be undertaken for IORP. EIOPA produced a draft of the proposed quantitative impact assessment on 15th June 2012 and a consultation period that ran until 31 July 2012. The results of this impact study will help determine how much extra capital pensions schemes will be required to hold against the risk of unexpected events. A detailed submission to EIOPA was made by a group of European real estate trade bodies. EIOPA received responses from many other organisations across the pensions and investment industry as well. Its response was published on 2nd October 2012. From the perspective of the real estate industry, some technical changes to bring the IORP Solvency Capital Requirement more in line with the latest developments on Solvency II were accepted, but the basic provisions for real estate remain unchanged.

EIOPA is also the regulator responsible for insurance companies so it should come as little surprise that there are many similarities between the Solvency II rules for insurance companies and the IORP provisions that will apply to pension schemes. Of concern to the real estate industry is that the Solvency Capital Requirement (SCR) in this Quantitative Impact Study (QIS) mirrors that proposed for insurers under Solvency II. In particular, the property shock provisions used in Solvency II that assume a 25% fall in real estate values are replicated in the new proposals. The application of this level of market shock would be as unappealing to defined benefit pension schemes as it is to life insurance companies. Both groups are major investors in real estate as an asset class.

The treatment of real estate lending also mirrors that of Solvency II, by treating real estate lending as a corporate bond. The treatment of corporate bonds under IORP is slightly simplified compared to Solvency II and uses the 3% per year of duration charge without the sliding scale for greater than 5 year duration loans under Solvency II.

Although the consultation period is now over, it is important that the real estate industry continues to lobby on this matter, but also that it continues to make its voice heard over areas of uncertainty in Solvency II such as the treatment of real estate lending. The regulator has made it clear that the same calibration is intended to be used in both. Any deficiencies in Solvency II that remain uncorrected are therefore likely to also apply to pension schemes.