Will Solvency II capital charges on real estate ever be reduced?
Last week, regulators offered hope to those that have been lobbying to lessen the capital constraints imposed on EU insurers when investing in real estate.
The European Insurance and Occupational Pensions Authority (EIOPA) has said it is not ready to amend the 25% capital charge that applies to real estate under Solvency II, but it also appeared to open the door to the idea in a new consultation paper.
Under Solvency II, many insurers are forced to hold cash equal to 25% of the value of their real estate investments, which the European association INREV and others argue overstates the volatility of European real estate, potentially skewing allocations away from the asset class.
But Solvency II is being reviewed and, in a consultation paper issued last week, EIOPA acknowledged that its calibration of the 25% “shock”, or solvency capital requirement (SCR), might be inadequate and said it “will continue its analyses towards a potential change”.
For those that have been working hard to persuade EIOPA to change its approach to real estate, it is not the vindication they were hoping for, but neither does it mean the game is over.
For Jeff Rupp, director of public affairs at INREV, it is a “glass half-full” moment. Although the paper does not mean anything will change, the fact that EIOPA is acknowledging that the real estate capital charge might be inadequate means the issue has to at least be addressed – even if ultimately the status quo is maintained. For this reason, the public consultation, which is open until to 15 January, is a “critical one”.
The central issue is that EIOPA based the 25% SCR exclusively on UK commercial property, because this market offered the deepest set of data. Since 2011, a number of updated studies by IPD (now MSCI, which acquired IPD in 2012) and commissioned by INREV suggest a shock of approximately 15% would more accurately reflect the volatility of real estate on a pan-European basis.
To overcome a paucity of data in certain markets and time periods – real estate downturns tend to prompt a slowdown in transaction activity – MSCI created a transaction-linked index, composed of valuation and transaction data, with the former back tested against the latter. This was designed to correct any potentially understated valuations and so to avoid smoothing of the market’s volatility.
As Rupp says, MSCI’s approach “does not give a perfect reflection of volatility of real estate investment”, but it is a better measure than the one currently employed by EIOPA. And if EIOPA acknowledge that its approach is inadequate, surely they should adopt the best available.
But there are two important things to consider. One, EIOPA has been careful with its words, so as not to actually say it believes is inadequate. “The current calibration was constrained by the availability of real estate annual returns observations where the only source of deep and sufficiently frequent data was available for the UK market,” it says in the consultation paper. The UK is “deemed to be the most volatile one in Europe and thus potentially not representative,” it adds, before saying: “Therefore several real estate investors are claiming this single shock is inappropriate”.
The second is that EIOPA seems to still be holding out for better data. Under the heading “Questions to stakeholders”, EIOPA asks: “Do you know data sources which would help to better calibrate property risk”.
EIOPA is likely to be disappointed. “What we are going to have to say is there isn’t any better data,” says Rupp. “This is an illiquid asset class. There’s infrequent trading. You don’t have very many transactions in smaller markets, and in times of economic stress, real estate buying and selling tends to freeze up.”
The game plan then should be to convince EIOPA that what INREV and MSCI have to date demonstrated is the best methodology available.
However, there is an even more curious question included in the consultation paper and it concerns internal models. Some large insurers with the necessary resources are able to calculate bespoke alternatives to the standard 25% SCR, potentially lowering the cost of holding real estate. These internal models have to be approved by national regulators and many are being used today.
“For internal model users, please indicate the approach chosen to model property risk within your internal model, when applicable,” the consultation paper asks. It seems EIOPA is interested to see if individual insurers can provide some new insight.
Rupp is surprised that EIOPA does not already have this information, since they need to be approved under Solvency II. This might suggest there is “imperfect information flow” between national regulators and EIOPA, he says.
INREV is aware that its insurer members use different methods within their internal models, and Rupp is not sure they will prove useful for EIOPA either – the models are designed specifically for each insurer, including their liabilities.
Rupp says INREV “would like to encourage all insurance members that have internal models” to indicate the approach and data sources they use, without necessarily divulging their ultimate solvency requirements.
In the end, EIOPA says it has three options. The first is to maintain the “status quo”. The second is to calibrate a new SCR “not only with UK data”. The third, it says, is to create two different shocks, which it seems would bifurcate Europe into higher and lower-volatility geographic markets and sectors (potentially separating residential and commercial property).
The latter is not what INREV is arguing for and EIOPA itself admits it would “clearly depart from the Solvency II pan-European approach always chosen until now for other standard formula risk modules”.
INREV will be making a response to the consultation. Rupp says it is not necessarily the last-ever opportunity to effect change, but it is a critical window.
“EIOPA is, at the time of this draft opinion, not in a position to provide the European Commission with a definitive advice implying a change to the current approach,” the consultation paper says.
“The bottom line,” says Rupp, “they are still thinking”.