Which future for the variable annuities under Solvency II?
After a first introduction in the US market in the 80s and in Japan in 1999, the concept of 'variable annuities'' (VAs) has been strikingly successful in these markets between 2002 and 2008. The possibility for policyholders to benefit within a single Life insurance product from gains in the financial markets while being protected against falling markets has made these products very popular. However, particularly in the US, the recent financial crisis caused severe losses for some important insurers. As this type of product is starting to arise more often in European countries , the EIOPA has established a task force to draft guidelines for the supervision of variable annuities within the Solvency II framework. At this stage, the guidelines are very conservative and restrictive, especially for companies which will not apply for an internal model on a short-term basis.
A profitable but risky product
Variable annuities are a tax-advantage form of investing in mutual funds providing, different type of protections, for additional fees. The guarantees offered vary considerably but the most common ones are:
The most popular products in the US market are the ones with a lifetime GMWB rider (78% of VAs sales in the US in the first half of 2005 contained such a feature). From the policyholder's perspective, the benefits of a VAs with a lifetime GMWB are:
- Liquidity and benefits from the upside markets (less the withdrawal charges and fees);
- Guaranteed income regardless of the lifetime of the policyholder or the performance of the underlying investments.
Variable annuities can be profitable for both policyholders and insurers, being more flexible, individual and transparent than traditional unit-linked products. However, the downside lies in the multiple and complex risks attached to it. Contrary to unit-linked products, the insurer proposing variable annuities faces multiple risks:
- Insurance risks (mortality, longevity, lapses ...);
- Market risks (interest rate, spread, liquidity ...);
- Counterparty risks;
- Operational risks (model risks, legal risks, delay risks ...).
As proven by the fall of the US market during the crisis and the losses on VAs products encountered by several insurers, the VAs and the underlying hedging programs are very complex. Many companies were overconfident about their understanding of the products design and suffered from severe losses due to the inefficiency of their hedging strategies .
Despite these shortcomings, the VAs are able to fully satisfy the growing customer needs for life insurance products with both strong guarantees and performance. They have the potential to be a key part of the portfolio of every European insurer, after the enforcement of the Solvency Directive in 2013. This is why the EIOPA set up a task force and published in November 2010 a consultation paper (n°83) to establish common European guidelines for the supervision on VAs sold by insurance companies.
Strong Solvency II requirements specific to the VAs
In the report, the task force is clearly concerned by the various risks linked to VAs, especially regarding the modelling approach and the hedging programs. To its opinion, an ad-hoc risk management process (roles & responsibilities, reporting & control, stress tests) has to be clearly defined among the actors (group, local entities, hedging platforms, asset management). The paper also highlights several systemic and macro-prudential issues that might raise if the European VAs market volume were to become significant enough:
- Pro-cyclicality of the hedging programs (same trades made at the same time by every insurer to cover their positions following market movements);
- Amplified by the fact that many models and software are similar between companies (often coming from one consultancy firm)
- And by non deep and illiquid markets such as during the 2008-2009 crisis.
These issues are the reasons why the task force believes that the Solvency II "Standard formula" is not adequate for the calculation of the technical provisions and capital requirements related to VAs. Its approach is too simplistic to take into account or accurately model the complex risks linked to VAs products. The task force stipulates that the calculations should be done with an internal model. To this regard, internal model validation tests, ORSA and Use test will have to integrate specific elements dedicated to the VAs (stress-tests, sensitivity analysis, governance process ...).
This new restriction is quite alarming for the small and medium insurance companies. From 2013, the insurers which will not have a validated internal model, whether full or partial, may not be allowed to sell variable annuities at all, or at least might be forced by their regulators to fund an add-on capital. In both cases, the competitiveness of such companies would be strongly weakened.
Alarming impacts on the small and medium insurance companies
These restrictions appeared shortly after the publication of the last Standard formula specifications in June, for the QIS5 exercise, which proved to be far more conservative than the previous ones and burdensome to implement operationally.
All these recent developments of the Solvency II framework emphasize the idea that it has not been designed and calibrated for small and medium insurers. The directive seems to let each company choose between a Standard formula and an internal model; but on a long-term point of view, the first one is not a viable option in term of strategy and competitiveness: too expensive, not sufficient for proper risk and asset-liability managements and, as shown by the VAs example, it might prohibit the developments of several new exotic products.
The move from a Standard formula to an internal model appears to be inevitable for every insurer. To do so, the Standard formula could be the first step toward a partial or full internal model, since most of the developments done could be re-used. But the gap between the two models is huge. To fill it in on a short-term basis is expensive and time-consuming, particularly for small and medium structures. This may explain the mergers that have recently occurred between several French mutual insurance companies and, more generally, between small or medium insurance companies.
The consolidation of the insurance sector is one answer to the threats posed by Solvency II. But it is not the only one and before 2013 many other axes with strong strategic impacts will have to be reviewed by each insurer due to the change of paradigm: investments allocation, risk appetite, hedging strategies, products portfolio ...
- Consultation paper no. 83: "Draft Report on Variable Annuities"
- US Securities and Exchange Commission: "Variable Annuities: What You Should Know"
- Life&PensionRisk: "Pricing and hedging of variable annuities"
- AXA Equitable: "Variable Annuity Guaranteed Benefits - Dynamic Hedging Considerations"
First product commercialised in the UK in 2005 by the American group The Harford
European Insurance and Occupational Pensions Authority
-18% of Variables annuities sales in 2009
€500-600 millions hedge losses for AXA in 2009