Edhec Working Paper Solvency II

EDHEC-Risk Institute393-400 promenade des Anglais06202 Nice Cedex 3Tel.: +33 (0)4 93 18 32 53

E-mail: [email protected]: www.edhec-risk.com

Solvency II : A unique opportunity for hedge fund strategies

January 2012

Mathieu Vaissié,Research Associate, EDHEC-Risk Institute, Senior Portfolio Manager, Lyxor AM

2

AbstractThere is growing empirical evidence that the complexity of financial markets makes it increasingly challenging for institutional investors to manage their asset/liability profile efficiently. Changes in the regulatory framework and in accounting rules make it even trickier for insurance companies. Against this backdrop, insurers have no choice but to rethink their overall investment policy. While the benefits of hedge fund strategies in asset liability management have been documented in the academic literature, the integration of these strategies into the global asset allocation of insurance companies may be jeopardised by recent developments on the regulatory front. We argue in this article that a Solvency capital requirement of 49% does not reflect the risks inherent in hedge fund strategies. We find that a capital charge of no more than 25% is deemed appropriate for a diversified hedge fund allocation, and conclude that hedge fund investing is appealing from both a risk-adjusted performance standpoint and a capital efficiency perspective. Contrary to the conventional wisdom, Solvency II could be a unique opportunity for hedge fund strategies to find their way into insurers’ core portfolios.

I am grateful to Noël Amenc, Eric Viet, Ludovic Antony and Anthony Rebreteau for their helpful comments. Any remaining errors are the sole responsibility of the author.

EDHEC is one of the top five business schools in France. Its reputation is built on the high quality of its faculty and the privileged relationship with professionals that the school has cultivated since its establishment in 1906. EDHEC Business School has decided to draw on its extensive knowledge of the professional environment and has therefore focused its research on themes that satisfy the needs of professionals.

EDHEC pursues an active research policy in the field of finance. EDHEC-Risk Institute carries out numerous research programmes in the areas of asset allocation and risk management in both the traditional and alternative investment universes. Copyright © 2012 EDHEC

I. IntroductionThere is growing empirical evidence that the complexity of financial markets makes it increasingly challenging for institutional investors to manage their asset/liability profiles efficiently. Changes in the regulatory framework (i.e. the Solvency II Directive) and in accounting rules (i.e. the International Financial Reporting Standards) make this even trickier for insurance companies. While equities exhibit too high a level of risk, the performance potential of bonds is limited over the long run and they may not be as safe an investment as one could have assumed. Against this backdrop, insurers - especially those with long-term liabilities - have no choice but to fully rethink their overall investment policies.

In an attempt to generate surplus and mitigate their shortfall risk through better diversification, over the last decade some insurers have ventured off the beaten track and gained exposure to “alternative” asset classes (i.e. real estate, commodities, private equity, infrastructure, hedge funds, etc.). While the benefits of hedge fund strategies in asset liability management have been documented in the academic literature (see Martellini and Ziemann [2008] or Darolles and Vaissié [2011a]), the integration of such strategies into the global asset allocation of insurance companies could eventually be jeopardised by recent developments on the regulatory front.

We argue in this article that a Solvency capital requirement of 49% does not reflect the risks inherent in hedge fund strategies. Applying a pragmatic - though robust - internal model approach to a series of investable hedge fund indices over an observation period covering the recent crisis, we find that a stress test of no more than 25% is appropriate for a well-diversified hedge fund allocation.

The remainder of this article is organised as follows. Since the Solvency II framework aims to improve the understanding, and in turn, the control of different types of risk, we start with a discussion of the appropriate way to gain an understanding of the embedded risks of hedge fund strategies. We then put the different hedge fund strategies under the microscope and assess the related stress tests. Lastly, we determine what we consider to be a suitable capital charge for a well-diversified hedge fund allocation. A brief overview of the Solvency II framework, its genesis and general principles, can be found in the appendices.

II. Understanding the risks of hedge fund strategiesThe objective of the Solvency II directive is to “establish Solvency requirements that are better adapted to the risks that are actually taken on by insurance firms and encourage the latter to better evaluate and control their risks”. In this respect, we call into question the way the risks relating to assets falling into the “other equities” category are calibrated in the standard approach. While the heterogeneity of the constituents of this category is clearly stressed in Consultation Paper No. 69, the scarcity and (poor) quality of the available information appears to be the key reason for such a disparate group. We argue in this section that this is no longer necessarily the case, and advocate a proper analysis of the underlying risks of hedge fund strategies.

Two methodologies are traditionally used to analyse the risk/return profile of an investment. The first is the holdings-based analysis. This approach determines the actual exposures of the fund. It involves interviewing managers, collecting data on turnover ratios, reading prospectuses, etc. The main drawback of holdings-based analysis is that information on a portfolio’s detailed positions is not readily available, and it is more often than not disclosed on an infrequent basis and with a significant time lag. Conclusions drawn from the analysis may therefore be misleading in the case of “window dressing practices”, or more generally, for dynamic trading strategies. The second methodology is the returns-based analysis. This approach uses an analysis of a fund’s track record, and is aimed at capturing the behaviour of the fund. In its simplest form, it consists of quantifying the level of realised risk: no attention is paid to the determinants of the risk; only 3

4

the tip of the iceberg is considered. A more advanced form of returns-based analysis involves a constrained regression using a series of risk factors as independent variables (see Sharpe [1988, 1992]). This gives an approximation of the fund’s implicit risk factor exposures; it is therefore less sensitive to window dressing practices than holdings-based analyses, or to the genuine characteristics of dynamic trading strategies (see Ben Dor et al. [2003]). The main drawback of this approach is its sensitivity to the quality of fund returns. Another caveat for the advanced form of returns-based analysis is that the outcome strongly depends on the set of risk factors selected.

The relevance of holdings-based and returns-based analyses (the basic and advanced forms) will thus depend, on one side, on the nature of the fund under scrutiny and the information available on that fund, and on the other, on the investor’s ultimate goal. It has been shown, for example, that the holdings-based approach is well suited to predicting the future holdings of mutual funds, while the returns-based approach tends to give better results in terms of predicting their future behaviour (see De Roon et al. [2004]). Hedge fund strategies have a greater degree of complexity and imply more often than not a higher level of portfolio activity than the typical buy-and-hold strategy followed by mutual funds. The quantity and quality of available information will therefore be a determining factor in the choice between the holdings-based or returns-based approach. In this respect, it should be recognised that the situation has improved dramatically over the last decade.

The more flexibility a manager has in terms of tracking error versus his benchmark (if any), markets traded or portfolio activity, the more leeway he has to leverage his talent (or reveal the lack of it) and boost (or impair) his performance. This insight was formalised in Grinold [2000] with the famous “fundamental law of active management”. So, if alpha exists at all1, it is in the hedge fund arena that one should look for it in the first place2. High net worth individuals, who exhibit a relatively high risk appetite and a clear focus on the return dimension, have thus been eager to pour money into small investment boutiques operating in poorly-regulated environments. In its infancy, the hedge fund universe was dominated by private investors for this reason. Because information was extremely scarce and its quality was questionable, neither holdings-based nor returns-based analysis made it possible at this stage to gain a good understanding of the real risks of hedge fund strategies. As a result, risk analysis was mostly qualitative, based on subjective judgment. Hedge fund investing 1.0 required an “overlay of expert judgment”; hence the rise of funds of hedge funds.

After the internet bubble burst, institutional investors were desperately looking for new solutions to improve the resilience of their portfolios during market corrections. They thus turned to alternative diversification. With the large-scale arrival of this new breed of investors displaying a greater focus on the risk dimension, the hedge fund world went through a Copernican revolution. In an attempt to comply with institutional investors’ demands, and in turn to attract a portion of their fund flows, a large number of hedge funds upgraded their infrastructure, improved their corporate governance and eventually adapted their investment strategy. To some degree, they opened up the “black box”, leading to a material improvement in the quantity, and to a lesser extent, the quality of the information. Access to performance data became easier, and investors started to get a bit more colour on the underlying strategy, and in some instances, on portfolio positioning. The holdings-based approach still failed to provide a good understanding of the risks of hedge fund strategies, but returns-based analysis began to be used to good effect at this stage. The simplest form of returns-based analysis made it possible to get a better - though not perfect, due to the quality of the inputs - representation of the risk/return profile of hedge funds; moreover, as investors were climbing their learning curve, and improving the risk factor selection process as a result, the advanced form of returns-based analysis progressively provided

1 - Since, by construction, alpha is a residual term, it can be argued that it converges to zero when a better understanding has been gained of the key drivers of the performance of the investment vehicle under consideration.2 - The analysis of hedge fund alpha is beyond the scope of this article. Readers interested in further information on hedge fund performance and return persistence can refer, among other examples, to Agarwal and Naik [2000], Amin and Kat [2003], Kat and Menexe [2003], Gupta et al. [2003], Capocci and Hübner [2004], Malkiel and Burton [2005] or Ibbotson et al. [2010].

more insight into the key drivers of their risk/return profile.3 Hedge fund investing 1.5 paved the way for a greater acceptance of alternative investment strategies by the traditional world.

Exhibit 1: Percentage of hedge fund managers’ total capital that comes from institutional investors

Source: Preqin [2011]

With the recent crisis, the hedge fund industry has further gained in maturity. Although this had already been widely discussed in the academic literature, many investors realised that all hedge fund strategies were not created equal (see Amenc et al. [2002], Schneeweis et al. [2003], Fung and Hsieh [2003], Jaeger and Wagner [2005] or Malkiel and Saha [2005]). In the same vein, traditional investors learned - more often than not the hard way - that the beta component could also dominate the alpha benefits in the alternative arena.4 Consequently, institutional investors, who now account for the bulk of flows and assets under management (see Exhibit 1), are adjusting their investment approach in two ways. Firstly, they require even more information on the funds and their underlying risk factor exposures, through more frequent and more granular reports. But as stressed in Goltz and Schröder [2010], these reports still do not always live up to expectations. In an attempt to have greater control over assets and direct access to information, the most demanding investors are turning to separate or managed accounts (see Exhibit 2). Independent oversight of hedge fund operations by the managed account platform provider, together with independent pricing of the underlying positions and independent risk management, do indeed make it possible to meet high standards in terms of both the quantity and quality of information. Secondly, as they climb their learning curve, institutional investors start paying more attention to the genuine risk features of different hedge fund strategies, and they progressively switch from commingled products to bespoke investment solutions that offer a perfect match with their specific needs.5 With managed accounts and their like, investors can have (audited) data points as often as daily, and they can increasingly leverage transparency to get a sense of the aggregate risk factor exposures. The holdings-based approach is now technically feasible, and may under certain circumstances produce good results6, while the simplest form of the Returns-based analysis can now give a true and fair representation of the risk/return profile of hedge fund strategies. Moreover, sophisticated investors can obtain a good understanding of the underlying risks of hedge fund strategies with the advanced form of Returns-based analysis. Hedge fund investing 2.0 is becoming increasingly traditional, making its integration into investors’ global asset allocation easier and more efficient.

We argue, as a conclusion, that it is now possible to perform “a reliable risk/return analysis” on hedge fund strategies, similar to that carried out on traditional asset classes.

5

3 - The improvement was limited, however. Information on positions falling out of hedge fund top holdings remained scarce, and as seen throughout the recent crisis, it is precisely those peripheral positions with a lot of optionality that had driven hedge fund performance.4 - Readers interested in a discussion of the place of beta in the performance of hedge fund strategies can refer to Géhin and Vaissié [2006].5 - Please refer to Martellini and Vaissié [2006] for a discussion on the benefits of tailor-made solutions over off-the-shelf products.6 - Although investors may have access to the details of a hedge fund’s books, this is not sufficient to draw an accurate picture of the actual risks. Processing such a huge amount of data is not straightforward and aggregating risk factor exposures properly requires a specific skill set.

6

Exhibit 2: Organisation model of an advanced managed account platform

Source: Giraud [2005]

III. Hedge fund strategies under the micrscopeSince a great deal of information can now be obtained on hedge fund holdings, it could be argued that the solvency capital requirement (SCR) of hedge fund strategies should be based on their aggregate risk factor exposures. However, the Solvency II directive appears to be very much influenced by traditional investor practices, and certain risk mitigation techniques have proved to be somewhat ill-suited for actively-managed long/short portfolios. The diversification benefits of the short leg of hedge fund portfolios are, as a consequence, more often than not ignored in the calculation of the SCR - leading to an overestimation of the embedded risks, and in turn, a somewhat punitive SCR. Until the structure and dynamics of hedge fund portfolios can be properly taken into account in the Solvency II framework, and provided that the exposure is gained through a secure investment vehicle providing a sufficient level of transparency and liquidity, we argue that the simplest form of returns-based analysis is likely to give a better estimation of risk(s) than the advanced form of returns-based or holdings-based analysis.

There are two practical challenges when running the simplest form of returns-based analysis on hedge funds. First and foremost, as previously mentioned, the quality of the publicly available information is, more often than not, highly questionable (see Liang [2003], Straumann [2009] or Schneeweis [2011]). There is also ample evidence in the academic literature that the information provided by commercial databases is severely impacted by performance measurement biases (i.e. survivorship, selection, instant history, etc.). Some of these biases are inherent in the very nature of the hedge fund industry, and others result from the way information is processed (see Fung and Hsieh [2000 & 2002]). While the estimation of these biases strongly depends on the sample and observation period, most studies conclude that the impact on performance, as well as on the risk dimension, is material. Thus, hedge fund performance data is not always representative of the performance an investor would actually have obtained. This is all the more true today as information available on funds that were shut down or created side pockets in the wake of the Lehman Brothers collapse is scarce. Secondly, hedge funds typically calculate net asset value on a monthly basis. It therefore takes years to collect a meaningful amount of data points. Since most hedge funds have a short history, empirical studies are more often than not carried out on a very limited number of observations. The estimation risk is therefore liable to be exacerbated. In order to tackle these two issues, we will use the hedge fund strategy indices provided by Lyxor.7

The specificity of these indices is that they comprise only managed accounts. Firstly, independent pricing of all the underlying positions and independent risk management ensure that the official net asset values published on a weekly basis on the Irish Stock Exchange offer a true and fair representation of the performance of the constituent funds. The performance of the indices is

7 - Greater detail on the construction methodology of this series of investable hedge fund indices can be found at www.lyxorhedgeindices.com

8 - It should not be concluded that all hedge funds fail to cope with market volatility. But as stressed in Liew [2003], the gap between the best and worst performers in the alternative world is widening over time. A growing proportion of industry players can therefore be expected to be negatively impacted by market gyrations; hence the necessity to apply the symmetric adjustment at the strategy index level.9 - Interested readers can refer to Fung et al. [2008], Kosowski et al. [2007] or Ozik and Sadka [2010] for a discussion of the relationship between fund flows and hedge fund performance.10 - As suggested in Consultation Paper No. 69, the symmetrical adjustment used in the latter case was calibrated using the historical performance of the MSCI World Index.

7

subsequently calculated by an independent calculation agent, namely Standard & Poor’s. The quality of the data is, as a result, as good as it can be. Secondly, our sample is made up of the weekly returns of the 14 Lyxor hedge fund strategy indices, from 4 January 2005 to 27 December 2011. We therefore have 365 weekly observations available. Thirty years of track records would have been needed to have the same number of observations with traditional hedge funds. Although necessary, having a significant number of data points is not sufficient. The information content is also essential. In this respect, our sample covers the most eventful period since the Great Depression, with a couple of bull markets, a series of market corrections, a systemic crisis, and lately a “risk on/risk off” environment. The quantity and the quality of information at our disposal is thus reasonably good.

We conduct the stress test for the different hedge fund strategies following the two-step procedure introduced in Consultation Paper No. 69 to calibrate the equity market risk. The first step consists of calculating the standard capital charge. It is determined so as to ensure a 99.5% probability of survival over a one-year period. In other words, the supervisory authority accepts a 0.5% chance that an insurance company will fail to cover its liabilities over a one-year horizon. Put another way, only the probability of a 1 in 200 year market event should have the potential to lead to the collapse of an insurer. The first step therefore boils down to calculating the 1-year Value-at-Risk (99.5%) for the different hedge fund strategies. The second step is to apply a symmetric adjustment mechanism. The main objective of this adjustment is to “avoid unintended pro-cyclical effects”. More specifically, the idea is to avoid an increase in the capital charge, and in turn, a fire sale in the middle of a crisis. We argue that this approach makes sense in the hedge fund world too. Indeed, just as upward/downward trends deriving from directional trades are expected to reverse at some point, market normalisations/disruptions caused partly by convergent/divergent trades are bound to come to an end sooner or later. Furthermore, there is ample empirical evidence that although managed actively, hedge funds are not immune to those reversals.8 This is particularly true when they are leveraged and/or exposed to liquidity risk (see Billio et al. [2010]). The collapse of LTCM in the wake of the Russian crisis in 1998 (see Jorion [2000]), or the quant crisis that took place during the summer of 2007 (see Khandani and Lo [2008]) perfectly illustrate the dramatic impact that such reversals can have on supposedly low-risk approaches such as relative value strategies. This effect is likely to be further compounded by the herding phenomenon, as investors commonly chase recent past performance.9 The adjusted capital stress formula is set out below:

Adjusted capital stress = standard capital stress + adjustment x betaWhere the adjustment is equal to

and It is the value of the strategy index under consideration at time t. The beta is calculated from a regression of the index level on the weighted average index level. As proposed by the CEIOPS we use a 1-year calibration period. The adjusted capital stress is subject to a band of +/- 10% around the standard capital stress.

Exhibit 3 shows the 1-year rolling percentile (0.05%) of the 14 Lyxor strategy indices over the observation period (blue areas). The standard capital charges are set equal to the minimum of these series over the observation period (plain lines). For comparison purposes we also used the capital charge currently advocated in Consultation Paper No. 69 (dotted lines).10 As can be seen from Exhibit 3, all strategies except one show a stress test level that is significantly lower than 49%. The only strategy that is close to this threshold - which was even slightly lower at the height of the crisis - is L/S Credit Arbitrage. Although this is unsurprising given that the credit market was at the epicentre of the crisis, such a result should be interpreted with care. The L/S Credit Arbitrage index is made up of a limited number of constituents, and is therefore highly sensitive to idiosyncratic factors. This intuition tends to be corroborated by the materially lower level of

7

8

stress exhibited by the Convertible Bond Arbitrage index. Over the observation period, realised risk for the other strategies is on average as much as 60% lower than the aforementioned 49% capital charge.

Exhibit 3A: Hedge fund strategy stress tests

Exhibit 3B: Hedge fund strategy stress tests

* Non-UCITS compliant index due to the limited number of constituents

IV. On the suitability of the calibration of the hedge fund capital chargeAlthough the trend is gradually changing, traditional investor exposure to hedge fund strategies remains highly diversified. It is therefore worth assessing the capital charge for a well-diversified hedge fund allocation. For this purpose, we use the Lyxor Composite index as a proxy and apply the two-step procedure described in the previous section. As can be seen from Exhibit 4A, we obtain a stress test of 21.86% over our observation period (i.e. 55% lower than the 49% threshold). Nevertheless, it may be argued that in order to be conservative, it would be more appropriate to take the weighted average of the stress tests of the 14 Lyxor strategy indices rather than that of the Lyxor Composite index. However, this would assume that the different strategies are fully correlated and that no diversification can be expected. We thus took the changes in the allocation of the composite index and the changes in the stress tests of the different hedge fund strategies, and computed the linear combination. As can be seen from Exhibit 4B, because of the “re-correlation” effect that is typically observed during periods of stress11, we obtain similar stress test levels (i.e. 22.20% vs. 21.86%).12 This lends weight to the idea that the calibration of

911 - It should be noted that this expression is somewhat misleading since as seen in Darolles and Vaissié [2011b], the higher co-movements observed during stressed market conditions are largely driven by an increase in the standard deviation as opposed to the correlation terms.

1012 - As highlighted in Consultation Paper No. 69, a similar result (i.e. 23.11%) is obtained with the HFRX Global Hedge Fund Index.

hedge fund risk in the standard approach of the Solvency II framework (i.e. 49%) is not suitable, and that the adjusted SCR of an unlevered and well-diversified hedge fund portfolio should be no more than 25%.

Exhibit 4A: Adjusted stress test of a well-diversified hedge fund allocation

Exhibit 4B: Fully correlated vs. actual stress test of a well-diversified hedge fund allocation

Capital is, and will increasingly be, a scarce resource. It is therefore essential for all investors, including insurers, to factor in the capital charge of the different asset classes when defining their long-term investment policy. As already mentioned, now that a true and fair risk/return profile of hedge funds can be obtained, hedge fund strategies can help investors maximise their surplus while minimising shortfall risk. Having conducted the stress tests, we can then assess the capital efficiency of the different hedge fund strategies and see whether they could fit within insurers’ portfolios. To this end, in Exhibit 5 we present the risk-adjusted performance (i.e. average return from January 2005 to December 2011 divided by the standard deviation of the returns over the same period) relative to the capital charge (i.e. maximum level of stress test calculated above). For comparison purposes, we do the same with equities, the typical performance seeking asset class for most traditional investors. As expected, the different hedge fund strategies exhibit heterogeneous profiles. More importantly, virtually all hedge fund strategies turn out to dominate equities in this framework. Also, a well-diversified allocation to hedge fund strategies clearly appears to be more appealing than a buy-and-hold strategy on equities both from an investment and a regulatory perspective.

Exhibit 5: Hedge fund strategy capital efficiency

As previously mentioned, bespoke solutions are increasingly considered by institutional investors in an attempt to maximise the benefits they derive from hedge fund investing. In this respect it is worth emphasising that it is straightforward to determine the SCR of any specific strategy mix using the basic internal model approach proposed in this paper. Alternatively, the SCR of the different hedge fund strategies can be easily factored into the portfolio construction process, and a solution can be designed that is optimal from both a risk-adjusted performance and a capital efficiency standpoint.

V. Concluding remarksInsurance companies are being compelled to revisit their long-term strategic allocation. The reason for this is twofold. On the one hand, the long-term assumptions typically used for traditional asset classes no longer fit with the “new normal” defined by Bill Gross; expected returns appear to be overstated, and levels of risk somewhat understated. On the other hand, changes in the regulatory framework and in accounting rules add further constraints. Insurers’ capacity to cover their liabilities through their current asset mix is therefore highly questionable. The good news is that there is some evidence in the academic literature that hedge fund strategies could help investors maximise their surplus while mitigating the shortfall risk. The bad news is that the aforementioned changes in the regulatory framework could deter insurance companies from considering the introduction of alternative assets into their overall allocation. There is indeed little chance in the current environment that insurance companies will favour hedge fund strategies over traditional performance-seeking assets knowing that the capital charge is currently materially higher (e.g. 49% vs. 39% for equities). In its current form, the Solvency II framework is thus preventing insurance companies from leveraging alternative diversification and implicitly directing them towards fixed income instruments, which may not be as safe an investment as one would have assumed. Paradoxically, the directive could put insurers’ long-term capacity to control their funding ratios at risk.

New forms of investment vehicles such as separate or managed accounts make it possible for insurance companies to gain exposure to hedge fund strategies with sufficient transparency and liquidity to perform “a reliable risk/return analysis”. As a consequence, we argue that there is no reason why hedge fund strategies should be placed in the “other equities” category, next to “emerging equities”, “private equity” or “commodities”, and suffer such poor treatment as in the standard approach. The Solvency II directive appears to be very much influenced by traditional investor practices, and certain risk mitigation techniques turn out to be somewhat ill-suited for actively-managed long/short portfolios. As a result, though technically possible, there is little

11

12

chance ceteris paribus that holdings-based analysis will give a true and fair representation of the risk profile of hedge fund strategies. In order to obtain a suitable calibration for hedge fund risk, we use a basic - though robust - internal model approach using the two-step procedure detailed in Consultation Paper No. 69. By so doing, it clearly appears that a SCR of 49% is not representative of the risks embedded in hedge fund strategies. A capital charge of no more than 25% is deemed to be appropriate for a well-diversified hedge fund allocation.

In conclusion, hedge fund strategies not only appear to provide insurance companies with an appealing solution from an investment perspective, but they also look to be efficient from a capital efficiency standpoint. Against all expectations, hedge fund strategies could end up playing a greater role in the future investment policy of insurers.

Appendix 1: The genesis of the Solvency II directiveThe foundations of the current prudential framework (i.e. Solvency I) date back to the early 1970s. Needless to say, the world has changed dramatically in the meantime and a set of simple, sometimes arbitrary rules that are accounting-oriented, neither represents the whole range of risks insurance companies are now exposed to, nor does it encourage insurance companies to manage their businesses efficiently. As a result, even if the number of failures among European insurance companies13 turns out to be below that observed elsewhere in the world, all the sector players (i.e. both insurance companies and supervisory authorities) came to the conclusion that the prudential framework had to be upgraded in order to better fit the current reality, hence the discussions surrounding Solvency II. As stated by the EU, the objective is to establish Solvency requirements that are more appropriate to the risks that are actually taken on by insurance firms and to encourage these firms to evaluate and control their risks more effectively. The goal of Solvency II is thus twofold. From a macro standpoint, it is aimed at mitigating systemic risks. From a micro standpoint, it is intended to detect any weakness or threat to an insurance company’s capacity to satisfy its future commitments.

In Solvency I, the capital requirement follows a fixed-rate approach (percentage of technical provisions, turnover or previous claims) and does not explicitly integrate the risks inherent in the activities of an insurance company (i.e. underwriting risks, risks related to the evaluation of technical reserves, etc.) or its day-to-day business (i.e. operational risks, legal risk, reputational risk, etc.). In the same vein, on the assets side, risks associated with the different asset classes (i.e. stocks, corporate bonds, commodities, etc.) are not explicitly included in the calculation of the SCR. Each country draws up a list of eligible assets and the authorised proportions that satisfy the constraint on “safe, liquid, diversified and profitable assets”. At the end of the day, as stressed in Amenc et al. (2006)14, the SCR of an insurance company depends more on the local statutory accounting standards than on the general economic outlook that tends to apply throughout Europe. Another benefit that can be expected from Solvency II is therefore greater harmonisation.

In an attempt to address the aforementioned limits of Solvency I and determine the level of prudential capital required for each insurance company more effectively, a series of more subtle principles - as opposed to hard rules - that are more economic-oriented and forward-looking in nature have been proposed. The new Solvency II provisions have been developed over the last decade in accordance with the EU’s Lamfalussy process: level 1 - framework directive (proposed by the European Commission and validated by both the European Parliament and the European Council); level 2 - implementing measures (proposed by the European Commission and validated by the European Commission with the consent of the European Parliament); level 3 - guidance regarding day-to-day supervision (CEIOPS); and level 4 - enforcement of directive (European Commission). Key milestones can be found in the following illustration.

13 - Relates to EU (re)insurers with annual premiums of more than EUR 5 million (smaller entities can choose to opt in) and EU branches and subsidiaries of non-EU-based groups.14 - Amenc, N., Martellini, L., Foulquier, P., and Sender, S. “The Impact of IFRS and Solvency II on Asset Liability Management and Asset Management in Insurance Companies.” Position paper, Edhec Risk Institute, 2006.

Solvency II timeline

Source Lyxor

Appendix 2: Solvency II general principlesIn a similar way to the Capital Requirement Directive for Banks (i.e. Basel III), the Solvency II framework uses a three-pillar approach (see illustration below).

The three-pillar approach

Source Lyxor

The first pillar contains the quantitative requirements and defines the solvency capital requirement (SCR) and minimum capital requirement (MCR). The SCR defines the target level of capital that an insurance company should hold so that it can “absorb significant unforeseen losses and give assurance to policyholders that payments will be made as they fall due”. As opposed to Solvency I, it takes into account a wide range of risks that insurance companies are exposed to (see illustration below). The MCR, on the other hand, is the level of capital below which the supervisory authority will consider that financial resources are not adequate, when it will automatically intervene. It should be noted that the MCR is to be entirely supported by Tier 1 and Tier 2 capital (with a minimum of 80% of Tier 1). The SCR on the other hand, must be supported by a minimum of 50% of Tier 1 and a maximum of 15% of Tier 3 capital.

13

14

A modular structure of risks

* Adjustments for the risk-absorbing effect of future discretionary benefitsSource: CEIOPS

The second pillar sets out the qualitative requirements for the governance and risk management of insurance companies. The objective is to ensure that an effective risk management system, covering all the risks to which the insurance company is exposed, has been put in place, and is used by senior management to control risk and capital allocation dynamically. An insurer must undertake an “own risk and solvency assessment” (ORSA) to make sure that sufficient capital is held against the risks that have been identified. Some therefore argue that Solvency II could be an opportunity for insurers to improve their overall performance (see Foulquier [2009]15). The supervisory authority will have powers to control the estimation procedures, the quality of the information and the systems used by insurance companies to monitor risks. Should the supervisory authority consider that risks are poorly accounted for, a capital add-on may be applied, or a reduction in risk exposure required.

The third pillar sets out disclosure requirements to increase transparency and foster market discipline. In order to ensure consistent reporting across the EU two types of reports are required from all European insurance companies. First, a public report (the Solvency and Financial Condition Report or SFCR), produced on an annual basis and containing qualitative and quantitative information. Second, a private report (the Regular Supervisory Report), produced for the supervisory authority on a quarterly basis, containing information that complements the SFCR, plus quantitative reporting templates developed by the EIOPA. This set of information is obviously expected to give a true and fair representation of the risks insurers are exposed to.

As mentioned above, Solvency II is intended - inter alia - to be more economic-oriented than Solvency I. The valuation of assets and liabilities therefore needs to be market-based as opposed to accounting-based. In this respect, technical provisions will be broken down into hedgeable and non-hedgeable risks. The former will be valued on a marked-to-market basis; the latter with a discounted best estimate method plus a risk margin using a ‘cost-of-capital’ approach (see illustration below).

15 - Foulquier, P. “Solvency II: An Internal Opportunity to Manage the Performance of Insurance Companies.” Position paper, EDHEC-Risk Institute, 2009.

From an accounting-based to a market-based approach

Source: Lyxor

Finally, in the Solvency II directive, the SCR is calibrated so as to ensure a 99.5% probability of survival over a one-year period. In other words, the supervisory authority accepts a 0.5% chance that an insurance company will fail to cover its liabilities over a one-year horizon. Put another way, only the probability of a 1 in 200 year market event should have the potential to lead to the collapse of an insurer. To clarify, the probability of an insurer defaulting over the next twelve months should, at any point in time, remain below the 0.5% threshold. From a technical perspective, estimating the SCR therefore boils down to calculating a 1-year Value at Risk (99.5%)16&17. The final amount of capital an insurance company is required to hold will therefore depend on two components:

1. the level of SCR associated with the six risk sub-modules defined in Consultation Paper No. 72 2. the diversification potential that can be expected if capital is spread across the above-mentioned sources of risk (see Consultation Paper No. 74 for greater detail on the calibration of correlation terms).

Both components are calibrated using historical data.As is the case with Basel III, each insurance company can either implement the standard formula, or adopt its own internal evaluation model with Solvency II. Supervisory approval is obviously required in the latter case.

Should insurance companies opt for the internal model approach, they need to satisfy a series of tests (i.e. use test, statistical quality standards, calibration standards, P&L attribution, validation test, documentation standards, external models and data) in order to validate consistency with the standard formula approach (see Consultation Paper No. 37 for greater detail). The internal model approach is therefore likely to be the preserve of larger insurance companies that have the appropriate level of resources (i.e. administration, legal, compliance, IT, etc.).18

VI. References• Agarwal, V., and Naik, N. “Multi-Period Performance Persistence Analysis of Hedge Funds.” Journal of Financial and Quantitative Analysis, Vol. 35, No. 3 (2000), pp.327-342.

• Amenc, N., Martellini, L., and Vaissié, M. “Benefits and Risks of Alternative Investment Strategies.” Journal of Asset Management, Vol. 4, No. 2 (2003), pp.93-118.

• Amin, G., and Kat, H. “Hedge Fund Performance 1990-2000: Do the Money Machines Really Add Value?” Journal of Financial and Quantitative Analysis, Vol. 38, No. 2 (2003), pp.1-24.

1516 - The shortcomings of this indicator are well known, but a critique of Value-at-Risk is beyond the scope of this article. For a discussion of coherent risk measures we invite interested readers to refer, for example, to Artzner, P.F., Delbaen F., and Eber, J.M. “Coherent Measures of Risk”, Mathematical Finance, 9 (1999).17 - As we have seen in the third section, a symmetric adjustment has been introduced for equity risk to avoid a pro-cyclical effect. But despite EIOPA’s advice, no volatility stress has been applied.18 - Medium-sized companies will be able to opt for a hybrid approach, and treat various activities or risks differently. However, it is not clear as yet how this will be implemented in practice.

16

• Ben Dor, A., Jagannathan, R., and Meier, I. “Understanding Mutual Fund and Hedge Fund Styles Using Return-based Style Analysis.” Journal of Investment Management, Vol. 1, No. 1 (2003), pp.94-134.

• Billio, M., Getmansky, M., and Pelizzon, L. “Crises and Hedge Fund Risk.”, Working paper, University Ca’ Foscari of Venice, 2010.

• Capocci, D., and Hübner, G. “An Analysis of Hedge Fund Performance.” Journal of Empirical Finance, Vol. 11, No. 1 (2004), pp.55-89.

• Darolles, S., and Vaissié, M. “Diversification at a Reasonable Price: Revisiting Alternative Diversification from the Perspective of Institutional Investors.” Working paper, EDHEC Risk Institute, 2011a.

• Darolles, S., and Vaissié, M. “The Benefits of Dynamic Risk Management: Mitigating Downside Risk without Compromising Long-term Growth Prospects.” Working paper, EDHEC Risk Institute, 2011b.

• De Roon, F., Nijman, T., and ter Horst, J. “Evaluating Style Analysis.” Journal of Empirical Finance, Vol. 11, No. 11 (2004), pp.29-53.

• Fung, W., and Hsieh, D.A. “Performance Characteristics of Hedge Funds and Commodity Funds: Natural Versus Spurious Biases.” Journal of Financial and Quantitative Analysis, Vol. 35, No. 3 (2000), pp.291-307.

• Fung, W., and Hsieh, D. “Benchmark of Hedge Fund Performance, Information Content and Measurement Biases.” Financial Analysts Journal, Vol. 58, No. 1 (2002), pp.22-34.

• Fung, W., and Hsieh, D. “The Risks in Hedge Fund Strategies: Alternative Alphas and Alternative Betas.” in L. Jaeger, ed., The New Generation of Risk Management for Hedge Funds and Private Equity Funds, London: Euromoney Institutional Investors PLC, 2003, pp.72-87.

• Fung, W., Hsieh, D., Naik, N., and Ramadorai, T. “Hedge Funds: Performance, Risk, and Capital Formation.” Journal of Finance, Vol. 63, No. 4 (2008), pp.1777-1803.

• Grinold, R. “The Fundamental Law of Active Management.” Journal of Portfolio Management, Vol. 15, No. 3 (2000), pp.30-37.• Géhin, W., and Vaissié, M. “The Right Place for Alternative Betas in Hedge Fund Performance: an Answer to the Capacity Effect Fantasy.” Journal of Alternative Investments, Vol. 9, No. 1 (2006), pp.9-18.

• Giraud, J.R. “Mitigating Hedge Funds’ Operational Risks: Benefits and Limitations of Managed Account Platforms.” Working paper, Edhec Risk Institute, 2005.

• Goltz, F., and Schröder, D. “Hedge Fund Transparency: Where Do We stand?” Journal of Alternative Investments, Vol. 12, No. 4 (2010), pp.20-35.

• Ibbotson R., Chen, P., and Zhu, K. “Sources of Hedge Fund Returns: Alphas, Betas and Costs.” Working paper, Yale School of Management, 2010.

• Jorion, P. “Risk Management Lessons from Long-Term Capital Management.” European Financial Management, Vol. 3, No. 3 (2000), pp.277-300.

• Kat, H., and Menexe, F. “Persistence in Hedge Fund Performance: The True Value of a Track Record.” Journal of Alternative Investments, Vol. 5, No. 4 (2003), pp.66-72.

• Khandani, A., and Lo, A. “What Happened to the Quants in August 2007? Evidence from Factors and Transactions Data.” Journal of Financial Markets, Vol. 14, No. 1 (2008), pp.1-46.

• Kosowski, ,R., Narayan Y., and Melvyn, T. “Do Hedge Funds Deliver Alpha? A Bayesian and Bootstrap Analysis.” Journal of Financial Economics, Vol. 84, No. 1 (2007), pp.229-264.

• Liang, B. “The Accuracy of Hedge Fund Returns.” Journal of Portfolio Management, Vol. 29, No. 3 (2003), pp.111-122.

• Liew, J. “Hedge Fund Index Investing Examined.” Journal of Portfolio Management, Vol. 29, No. 2 (2003), pp.113-123.

• Malkiel, B., and Saha, A. “Hedge Funds: Risk and Return.” Financial Analysts Journal, Vol. 61, No. 6 (2005), pp.80-88.

• Martellini, L., and Vaissié, M. “Optimal Allocation to Hedge Funds.” RISK, Vol.19, No. 3 (2006), p.76-80.

• Martellini, L., and Ziemann, V. “The Benefits of Hedge Funds in Asset Liability Management.” Bankers, Markets & Investors, No. 97 (2008), pp.16-30.

• Jaeger, L. and Wagner, C. “Factor Modelling and Benchmarking of Hedge Funds: Can Passive Investments in Hedge Fund Strategies Deliver?” Journal of Alternative Investments, Vol. 8, No. 3 (2005), pp.9-36.

• Ozik, G., and Sadka, R. “Smart Money or Smart about Money? Evidence from Hedge Funds.” Working paper, Boston College, 2010.

• Preqin. “Preqin Global Investor Report: Hedge Funds.” 2011.

• Sharpe, W. “Determining a Fund’s Effective Asset Mix.” Investment Management Review, Vol. 2, No. 6 (1988), pp.59-69.

• Sharpe, W. “Asset Allocation: Management Style and Performance Measurement.” Journal of Portfolio Management, Vol. 18, No. 2, pp.7-19.

• Schneeweis, T., Kazemi, H., and Martin, G.A.“Understanding Hedge Fund Performance Research Issues Revisited - Part II.” Journal of Alternative Investments, Vol. 5, No. 4 (2003), pp.8-30.

• Schneeweis, T., Kazemi, H., and Szado, E. “Hedge Fund Database ‘Deconstruction’: Are Hedge Fund Databases Half Full or Half Empty?” Journal of Alternative Investments, Vol. 14, No. 2 (2011), pp.65-88.

• Straumann, D. “Measuring the Quality of Hedge Fund Data.” Journal of Alternative Investments, Vol. 12, No. 2 (2009), pp.26-40.

17