Read PRiM%20Newsletter%20No%2024_print.pdf text version

PRiM Risk Newsletter

1 1

The PRiM Risk Newsletter

Issue No. 24 January 2011

PRiM

Welcome!

In comparison with the banking sector, risk management in the fund industry has traditionally had a relatively narrow focus on fund performance, centred on portfolio management. The acceptance of higher risk instruments such as derivatives in UCITS III and the far-reaching impact of the financial crisis on fund markets have resulted in a wider focus on risk management in investment funds. The oftendiscussed Article 42 of the Luxembourg Law of 20.12.02 and the subsequent Circular 07/308 are manifestations of the increasing importance of risk management in investment funds. With the approval of the Alternative Investment Fund Managers Directive and UCITS IV, this focus on risk management in investment funds has gained momentum and encouraged PRiM to dedicate this issue of the Newsletter to the topic of risk management in investment funds. Included in this issue are two interviews on risk management in investment funds: one with Alain Hoscheid, Conseiller de Direction Adjoint, Commission de Surveillance du Secteur Financier, the other with three market practitioners and members of the PRiM Board of Directors, Thomas Nummer, Head of Risk at Allianz Global Investors Luxembourg, Luc Neuberg, Head of Risk & Operations, BCEE Asset Management and Michael May, Deputy Global Head of Risk Management, HSBC Securities Services. Also in this issue are articles on: Model risk by Alan Picone, Director, Enterprise Risk Services, Deloitte S.A.; UCITS IV by Ravi Beegun, Partner at KPMG; liquidity risk by François Génaux, Partner, Financial Services Consulting Leader, Thierry López, Risk Management Services Leader and Benjamin Gauthier, Manager, Risk Management Advisory Services, from PwC Luxembourg. As always, your ideas for future issues and your comments are welcome. Please send them to [email protected] Bonne lecture! Paul Kleinbart, Editor

The individual opinions expressed in this newsletter do not necessarily reflect the opinion of PRiM nor of any other contributors to this edition

Editeur responsable:

Marco Zwick PRiM c/o ABBL 59, boulevard Royal L-2449 Luxembourg

Contact: [email protected] Web site: www.prim.lu

IN THIS ISSUE

Page 2 CSSF INTERVIEW: Risk Management for UCITS Funds Page 4 Risk Model & Model Risk: A Key Duality Page 8 New Agreement Between UCITS IV Management Companies and Depositaries Page 11 Liquidity Risk for Investment Funds: Both Sides of The Story Page 15 PRiM Scholarship 2011 Conference report: Philippe Jorion Page 16 KEYNOTE INTERVIEW: RISK MANAGEMENT FOR INVESTMENT FUNDS Page 22 PRiM News

2

PRiM Risk Newsletter

INTERVIEW:

RISK MANAGEMENT FOR UCITS FUNDS

PRiM: Risk management in the fund industry took on a new dimension with the initial implementation of the Law of 20 December 2002 (hereafter "2002 UCI law") and the more recent regulatory developments. What were the driving forces behind these developments? Alain Hoscheid: First of all, it is important to point out that the amendments brought in 2002 to the UCITS Directive 85/611/EEC (known as the UCITS III Directive) led to the introduction of precise requirements in terms of risk management, asking in particular management companies or selfmanaged investment companies to employ a risk management process and to comply with an extensive risk-limitation system (i.e., global exposure, counterparty risk, etc). The widening of the investment spectrum under UCITS III allowing in particular funds to make use of financial derivative instruments as part of their investment policy (and not only for hedging purposes) was the main reason for that development. Indeed the European legislator felt that risk management requirements were necessary for ensuring that investor protection remains preserved. Now as concerns the more recent regulatory developments relating to risk management and that are part of the global UCITS IV package (i.e., Commission Directive 2010/43/ EU, CESR's guidelines on Risk Measurement and the Calculation of Global Exposure and Counterparty Risk for UCITS), the main driving force was the need felt by European Supervisors to enhance harmonisation in the area of risk management across Member States. Of course, the financial crisis that the world went through in 2007/2008 further emphasised the need for having a comprehensive and high-quality set of risk management rules.

Risk Management for UCITS Funds

An interview with Alain Hoscheid, Conseiller de Direction Adjoint, Commission de Surveillance du Secteur Financier (CSSF)

PRiM Risk Newsletter

3

PRiM: When comparing with the current regulatory framework comprised of CSSF circular 07/308, what is the main change that UCITS funds have to expect from the latest European initiatives in the area of risk management? Alain Hoscheid: The focus of the aforementioned circular is mainly on the investment risks directly addressed in the 2002 UCI law, these being global exposure/market risk, counterparty risk and concentration risk. With the latest European initiatives and in particular the Commission Directive 2010/43/EU, the scope of the risk management process will have to extend to other risks, as all risks that may be material for a UCITS fund that a management company or self-managed investment company manages will have to be covered, including liquidity risk and operational risk. As regards in particular liquidity risk, management companies or self-managed investment companies have to implement an appropriate liquidity risk management process, supplemented if applicable by stress testing and they have to ensure for each fund they manage that the liquidity profile is in line with the redemption policy. In this context, it is however very important to highlight that the risk management process of a management company or selfmanaged investment company has to be adequate and proportionate to the nature, scale and complexity of the fund business it manages. As a consequence, a company managing, for instance, more complex funds has to put in place a more elaborate risk management process than this is, in principle, the case for a company managing less complex funds. PRiM: Thank you for sharing your view with us.

Risk Management for UCITS Funds

4

PRiM Risk Newsletter

RISK MODEL & MODEL RISK: A KEY DUALITY

Alan Picone, PhD Director, Entreprise Risk Services, Deloitte S.A.

Issue 17

Risk Model & Model Risk: A Key Duality

Risk

Risk is common to virtually every industry success factors are random variables, business decisions hardly ever lead to predictable outcomes and earning drivers depend on unpredictable dimensions. Risk materialises itself in various ways, from business-specific distress to a systemic, contagious disease, passed on through ruin or insolvency.

Alain Picone

We are structurally incapable of preventing risk events from occurring. Is there anything that can be done, besides crossing fingers for favourable outcomes? Perhaps the answer lies in the old saying: "Hope for the best, prepare for the worst".

Risk model

Preparing for adverse situations is actually a programme that is actively pursued by the investment fund industry. Fund vehicles are instruments that concentrate the mosaic of financial risks 1 and beyond, making them very complex structures. Risk management of funds aims at better capturing the risks and their intensities, which means assigning probabilities and magnitudes to potential outcomes. In two words, risk modelling. Since risk events cannot be managed preventatively, emphasis is put on quantifying this uncertainty - risk models are the fundamental tools allowing us to deduce these insights. Risk models are an abstract of a fantastically complex reality. Simplifications are required regarding a model's assumptions to ensure effectiveness in the risk metrics production process. Reality must be faithfully rendered, however. Reconciling these somewhat opposing forces is challenging, yet a necessary condition for proper risk measurement. Even when thoroughly conducted, this programme gives rise to residual risk, the so-called model

1 Non-exhaustive list of risk factors includes market, credit, concentration, currency, interest rate, commodities, operational and liquidity risk.

Bibliography · Committee of European Securities Regulators, CESR's Guidelines on Risk Measurement and the Calculation of Global Exposure and Counterparty Risk for UCITS, 10/788 Commission de Surveillance du Secteur Financier, Circular 07/308 E. Lorentz, The essence of chaos ­ 1991 B. Mandelbrot and R.L. Hudson, The (mis) behaviour of markets : a fractal view of risk, ruin and reward ­ 2005 A.J. McNeil, R. Frey and P. Embrechts, Quantitative risk management ­ 2005 E.J. Elton, M.J. Gruber, S.J. Brown, W.R. Goetzmann, Modern portfolio theory and investment analysis ­ 2000

· · · · ·

Issue 17

PRiM Risk Newsletter

5

risk, i.e., the risk of producing figures and estimates from a flawed model, namely the limited capacity of the risk engine to capture the true economic reality.

Risk model validation

To discriminate as to the adequacy of a risk model, a comprehensive risk validation programme should be deployed. Needless to say, the financial crisis of 2007-2008 has dramatically fuelled this imperious need for mitigating "model risk". In the aftermath of the economic debacle, it indeed became clear that some risk management models had failed to capture the most essential risk drivers of instruments held in the fund portfolios. For the sake of simplicity, complex structures like credit derivatives were artificially mapped onto more simple ones, therefore disregarding their explosive character and the "wilderness of their risks" - to quote the eminent scientist and fractal father Benoît Mandelbrot. Consequently, it is no surprise that risk model validation forms a key pillar within the UCITS IV risk management framework. Illustratively, CESR's guidelines 10/788 tend to strengthen significantly the requirements regarding risk model validation and they will most certainly be transposed in Luxembourg via amendments to CSSF Circular 07/308. Remarkably, risk model validation supersedes the mere local, point-in-time exercise of gaining assurance as to the foundations of the model. It is a process in itself, involving various components whose interaction sets the basis of the validation programme. Ex-ante, out-of-sample back-testing: Before a model is used, it should be thoroughly tested. This generic status is quite convincing when it comes to industries where stakes do not allow errors. Aeronautics or pharmaceutics are illustrative examples of industries which have literally forged a culture of ex-ante validation. Models are "put in the lab" and fuelled with

supposedly realistic future conditions. In scientific jargon borrowed from physics, the rationale is to test the susceptibility of the model, namely its reaction towards some specific stress and factor pushes. Essentially there are two methods to derive such tests: Either taking observed, real conditions and applying them to the model (out-of-sample back-testing) or simulating parametric shocks. Whilst aeronautics extensively uses both approaches, the financial industry does not seem to have reached the same level of maturity. This certainly stems from the nature of losses attached to underpinning risks, namely money versus people's lives. Yet, it is astonishing to observe that the vast majority of financial models are operated within institutions without having been submitted to the testing of previous events. Considering the huge amount of data recorded every second in financial markets, this makes it all the more difficult to understand since all this information should have already paved the way for this out-of-sample validation exercise. Also, the parameterisation and mapping processes of risk models, which are crucial yet complex tasks, would benefit significantly from being calibrated in light of observed events. Qualitative validation - risk drivers and assumptions: Every model is an abstract of the complex reality. The art of modelling consists in rendering the most of the observed reality out of a minimum number of explanatory factors. The performance of models can be gauged in respect to this compromise. Strikingly, meteorological modelling illustrates this tour de force. Whilst more than 80 variables actually influence the climate including oceans currents, air humidity, ocean salinity and seismic activities, it is clear that it is neither realistic nor desirable to model the weather using this set of variables and their related dynamical equations of motion. In fact, the weather forecasting industry

Risk Model & Model Risk: A Key Duality

6

PRiM Risk Newsletter

Risk Model & Model Risk: A Key Duality

uniformly relies on associating three factors - pressure, temperature and wind velocity by means of equations of motion stemming from the pioneering work of Edward Lorentz2. This masterwork in reducing dimensionality of course comes with its limitations: the time horizon of predictions cannot exceed more than a couple of days. The situation is actually very similar in finance, and the equivalent of Lorentz's representation would typically be factorial models all inspired by the CAPM 3 introduced by Sharpe in 1964. Just like in meteorology, the question as to which risk factors to retain in the model is fundamental. This risk factor coverage, alternatively called mapping, is a key element of the performance of models. For investment portfolios that can easily contain hundreds of securities, it is illusionary to model every security as a one and unique risk factor, be it only from the perspective of handling correlation matrices of tens of thousands of entries. In contrast, common risk factors are sought with the rationale that most of the risks relate to external, market observable conditions, thus minimizing the idiosyncratic, unexplained, risk component. In its quest for reducing the number of degrees of freedom, a qualitative model validation will therefore naturally focus on three essential aspects. Firstly, on the coverage of risk factors and the mapping process after which securities are assigned risk factors and pricing equations. Secondly, on the mathematical assumptions and dynamic equations of motion used to simulate the behaviour of risk factors, part of which gives rise to calibration issues. Thirdly, and most importantly, on the ability of the model to render the observed facts and properties of

2 3 M. Lorentz is generally recognized as the father of the chaos theory, of which so-called "butterfly effect" is very popular CAPM stands for Capital Asset Pricing Model

a financial time series. These are uniformly characterised by a distinctive "statistical texture", long remarked and described, among others, by Mandelbrot. These stylised facts are observed in particular by the presence of fat tails in return distributions, the stochastic nature of volatility and the leverage effect (inverse correlation volatility/return). In addition, these facts bear the essentials of the risk landscape of financial securities, and by extension, of investment funds. Interestingly enough, a number of today's models still do not incorporate these facts, which are essentially "legitimated" for the sake of convenience. These simplification assumptions actually jeopardise the integrity of the risk measurement process, rendering it at best flawed, at worst useless, since they are based on a reality which disregards the specifics of financial markets. Continuous monitoring of a model's performance - ex-post back-testing: Validation of risk models is an ongoing process as the proven performance of the model can only be assessed during its life when effectively operated. For this reason, the monitoring of model quality in light of observations, so-called back-testing, has long been a matter of emphasis from both risk management professionals and financial regulators. Supervisory authorities indeed pay particular attention to the notion of overshooting, which occurs whenever a negative return of the fund is larger in magnitude than the risk metrics, namely the Value at Risk. CESR's 10-788 guidelines reinforce the trend of CSSF Circular 07/308, insisting on the strict monitoring of the number of overshootings observed during a given period. For the first time, some quantification is provided : 4 observations during a 250 day period are deemed to be sufficiently serious to ensure that the UCITS senior management is informed together

PRiM Risk Newsletter

7

with an analysis of the sources of the VaR violations and, most importantly, a statement of the measures, if any taken, to improve the model. In addition, the CSSF expects an annual update, as a minimum, on matters of back-testing results, irrespective of their quality. In simple terms, it is the intention of financial regulators to use back-testing as an influencing tool, whose results could trigger adjustments, or re-foundations of models. In this respect, the back-testing exercise could be considered having a superior impact than other risk assessments such as stress-testing or VaR measurement, since it is supposed to be defined into concrete actions once certain thresholds have been touched. The effectiveness of a back-testing programme stems from its tangibility and its conceptual simplicity. It is an objective, quantitative and non-emotional scoring of the model's quality which cannot be "gamed" and dressed better than it truly is. In addition, the model only needs two columns, namely VaR predictions and return observations, to conduct a basic back-testing programme which still provides insightful results. The hypothesis H0: "My internal risk model performs adequately" can actually be tested within a comprehensive back-testing programme. Binomial testing is one method, as the number of overshootings is distributed across the binomial distribution with parameters linked to the VaR confidence level. Other dimensions, such as the independence of overshootings, their magnitude, or the entire distribution of returns, also complement the back-testing programme. As intellectually relevant as a model might seem, it is, in the end, only as good as the sum of its well-founded predictions during observation periods of various economic cycles. This brings us back to the roots of scientific modelling: describing nature as it is and reproducing its facts as they are.

Independent risk model validation

We have referred several times to CESR's guidelines 10-788 as an amplifier of the financial regulators' focus on risk model validation. Illustratively, these guidelines propose new foundations of the role attribution within the model validation process: "models will have to undergo a validation by a party independent of the building process, ensuring model's soundness and its capacity to adequately capture material risks". This is obviously of a propensity to promote further objectivity in the validation exercise, as long as the external party's judgment is not incentivised in any way. Management companies will therefore have to accommodate this new reality of having their internal models scrutinised. In our opinion, the scope laid down by CESR's guidelines 10-788 defines risk model validation as a stand-alone exercise, superseding "traditional" audit reviews. Risk modelling practices and confidence in risk measurement will come reinforced out of such a programme. Even more ambitious, once validated could models be somewhat certified? Doing so, the financial industry would mimic its peers from aeronautics or pharmaceutics, where the culture of certification is deeply embedded in the mindset of risk professionals. In any case, a simple fact should ultimately be acknowledged: as long as there are risk models, there is model risk. Institutions that will spare no pain in controlling the latter will improve the former and clearly benefit from a distinctive competitive advantage in a world that has hardly ever been so risk-sensitive.

Risk Model & Model Risk: A Key Duality

8

PRiM Risk Newsletter

Ravi Beegun, Partner, KPMG

Issue 17

NEW AGREEMENT BETWEEN UCITS IV MANAGEMENT COMPANIES AND DEPOSITARIES

UCITS IV ManCos and Depositaries

Ravi Beegun

he UCITS IV Directive was transposed into Luxembourg law on 24 December 2010. One of its requirements is to have an agreement in place between the depositary and the management company when the management company is not in the same country as the UCITS fund. This agreement will be important for both the management company and the depositary as a means of clarifying their roles and responsibilities, and thereby also delimiting certain risks for which they will be retaining or sharing responsibility. While these requirements currently do not concern domestic structures (when management companies are in the same country as the UCITS funds), European regulators and the European Commission are currently considering whether to extend it to domestic structures, in order to provide for a level playing field.

T

Overview of new requirements

The requirements are specified in Articles 34 to 41 of CSSF regulation No.10-4, transposing directive 2010/43/EC. The agreement should contain at least the following items: a) Procedures to be followed by the depositary and the management company · Procedures by asset class, including safekeeping: a description of the procedures to be adopted for each type of asset of the UCITS fund entrusted to the depositary, including those related to safekeeping · Prospectus or fund rule changes: a description of the procedures to be followed when the management company wishes to change the prospectus or the fund rules, specifying when the depositary should be informed and obtaining the prior approval of the depositary, if applicable

Issue 17

PRiM Risk Newsletter

9

Information from depositary: a description of the means and procedures used by the depositary to transmit all relevant information to the management company to enable the latter to carry out its duties properly, including exercise of rights attached to financial instruments, having timely access and accurate information on the accounts of the UCITS funds · Information from management company: a description of the means by which the depositary will have access to all information it needs to carry out its duties · Management company's conduct of activities: a description of the procedures by which the depositary can obtain information on how the management company is performing its duties, assess the quality of the information received and on-site visits; and · Depositary's performance: a description of the procedures through which the management company can assess the depositary's performance in carrying out its contractual duties. b) Exchange of information, confidentiality and anti-money laundering The agreement should cover the exchange of information in relation to the prevention of money laundering: · Information: list of information which must be exchanged in relation to share/unit transactions (e.g., subscriptions, redemptions, cancellations) · Confidentiality: obligations applicable to both parties · Tasks and responsibilities: of each party regarding anti-money laundering roles and responsibilities;

·

c)

d)

e) f)

g)

and · Access to regulatory authorities: these measures should not prevent regulatory authorities from gaining access to any relevant information. Appointment of third parties If the management company or depositary will use third parties to help perform certain tasks, then the agreement should contain: · Information on the third parties appointed: provide details on any third parties appointed by the management company and the depositary · Appointment criteria and ongoing reviews: commitment by each party to provide information to the other party on such criteria and steps taken to monitor third parties; and · Depositary's declaration on responsibility: stating that the depositary's responsibility is not affected by the use of sub-custodians. Amendments and termination of the agreement The agreement should contain provisions that deal with changes to and termination of the agreement: · Period of validity of the agreement · Conditions for amendments or termination; and · Transition measures when changing depositary. Applicable law Luxembourg law is applicable. Electronic transmission of information There should be adequate measures for the recording of information transmitted electronically between the management company and the depositary. Scope of the agreement The names of the relevant UCITS funds

UCITS IV ManCos and Depositaries

10

PRiM Risk Newsletter

UCITS IV ManCos and Depositaries

falling under the scope of the agreement should be mentioned in the agreement. h) Service level agreement Detailed information may either be included in the agreement itself or in a separate written service level agreement.

Next steps to manage cross-border operational and compliance risks

Given that the new requirements are comprehensive, both the management company and the depositary will need to specify as much as possible their tasks and responsibilities under the agreement. This will be important for ensuring afterwards that procedures and controls exist to achieve proper execution of these tasks. It will also help both organisations to manage their risks in relation to improper execution of tasks under the agreement. The main challenges in drawing up such agreements are likely to be the lengthy discussions concerning responsibility, specifying the detailed information flows and overall legal wording used in the agreements to define respective liabilities. From the implementation side, the challenges will be in terms of structuring existing and new flows within the new framework, verifying that controls are operating to prevent inaccurate information being provided to other parties and putting the right level of technical means and resources in place to provide the relevant information and to review information received by the recipient. It is therefore important to start the process early for Luxembourg management companies of foreign UCITS funds, or for Luxembourg depositaries of UCITS funds managed by a foreign management company. The experience will no doubt be helpful if there is an extension of the requirement for an agreement to include domestic structures.

Issue 17

PRiM Risk Newsletter

11

LIQUIDITY RISK FOR INVESTMENT FUNDS: BOTH SIDES OF THE STORY

T

he recent financial crisis highlighted the fact that the importance of liquidity risk management, as well as the severity and swiftness with which this risk can materialise, had been severely underestimated by the industry, markets and regulators alike.

New regulation

As a result, regulators have issued a torrent of new regulation: - In the context of liquidity risk regulation for the banking sector, the BIS1 document "Basel III: International framework for liquidity risk measurement, standards and monitoring", published in December 2010, is of the utmost importance. In that document, the Basel Committee puts forth a variety of regulatory changes, mainly centred on two new ratios: The short-term Liquidity Coverage Ratio and the longer term Net Stable Funding Ratio, - Regarding the asset management industry, both UCITS2 IV and the AIFMD3 include the concept of liquidity risk in their guidelines. CESR's4 level 2 and level 3 guidelines5 (issued in the context of UCITS IV) clearly emphasise liquidity risk by mentioning it right after market risk in the description of the risks that at least should be covered by the Risk Management Process (RMP). Also, the recently issued "Règlement CSSF 10-4"6 specifies that the management companies must use an appropriate liquidity risk management process (including stress tests) in order to guarantee the right for shareholders to redeem their shares. The AIFMD dedicates a full paragraph7 of its global framework to liquidity risk management.

Thierry López

François Génaux

The new regulations regarding liquidity risk pose considerable challenges to the fund industry in various aspects. And as far as the regulations are concerned, it must be pointed out

1 2 3 4 5 The Bank for International Settlements. The UCITS (Undertakings for Collective Investment in Transferable Securities) IV Directive seeks to update the regulatory framework applicable to European investment funds. The Alternative Investment Fund Managers Directive aims at regulating a big portion of the world of investment funds that was non-regulated. The Committee of European Securities Regulators that has become the European Securities and Markets Authority (ESMA) as of 1 January 2011. CESR/09-963 CESR's technical advice to the European Commission on the level 2 measures related to the UCITS management company passport, section IV, chap 1, Box 1, and CESR/10-788 CESR's Guidelines on Risk Measurement and the Calculation of Global Exposure and Counterparty Risk for UCITS, Box 1. Regulation CSSF N° 10-4 transposing the EU Directive 2010/43/UE, art. 45 paragraph 3. Proposal for a Directive of the European Parliament and the Council on AIFM and amending directives 2003/41/EC and 2009/65/EC, Article 16 "Liquidity management".

Benjamin Gauthier

6 7

Liquidity Risk for Investment Funds: Both Sides of the Story

François Génaux, Partner, Financial Services Consulting Leader, Thierry López, Risk Management Services Leader and Benjamin Gauthier, Manager, Risk Management Advisory Services, PwC Luxembourg

12

PRiM Risk Newsletter

Liquidity Risk for Investment Funds: Both Sides of the Story

that the regulator is not providing any recommendations on how to meet these new requirements.

the loss due to the low level of liquidity of the market. Liquidity risk, therefore, is a consequence of another event. Now should the asset manager have difficulties liquidating a portion of his portfolio due to the low level of liquidity of some securities, then selling those securities that are more liquid can lead to a deviation in terms of investment policy. Eventually this can also generate breaches and trigger the reimbursement of investors. Here again, the risk does not come directly from the liquidity level itself, but from its consequences.

How to? Existing (asset driven) methods

A few methods are currently used by practitioners, but they have clear limitations. As an example, liquidity risk is commonly assessed by using the weighted average traded volume for equities. If this method presents clear advantages (the data can be easily obtained, the number of days required to liquidate the positions can be derived, etc.), some drawbacks should also be mentioned: For example, having a high level of traded shares on the market can also reflect that the underlying company is in a bad situation and that investors are starting to disinvest. The bid-ask spread is another method mainly used for bonds. Its interpretation is also straightforward: The wider the spread, the lower the liquidity of the related bonds. But here again, some drawbacks have to be mentioned: Complete and reliable information is difficult to find, the interpretation is as volatile as the markets (What is a wide spread?), this metric is not forward-looking, etc.

Both sides of the story

Liquidity risk of investment funds, therefore, has to be monitored by looking at both sides of the balance sheet. The current liquidity of the assets in the portfolio needs to be assessed and monitored, but the subscription/redemption movements need to be taken into account beforehand. Also to obtain a global picture of the liquidity risk, stress tests will need to be designed to define today the buffer that will potentially protect the fund investors tomorrow.

How to? Non widespread (liability driven) methods

From a liability point of view, one may think of analysing the investor's profile in order to determine the potential future trend. But this strategy is very difficult to implement: Defining the investor's geographical positions, styles, etc., is limited by the structure of investments (nominee account, etc.). But relying on statistical techniques allows us to define (imperfectly) future trends based on history. The idea is to capitalise on the different stresses that have been encountered in the past as well as on the investor behaviour to define a reasonable trend for a short future period of time. Then

Liquidity risk: A consequential risk

Due to the above-mentioned limitations, one has to capitalise on everything known and should start to improve the methods by better defining the issues, the environment and the risks. Having a certain percentage of assets in a portfolio that are less liquid or illiquid starts to be an issue and creates risks when investors start to redeem large numbers of shares during a continuous period of time. When the asset manager does not need to hurry to sell a portion of the portfolio, he will probably have opportunities to limit

PRiM Risk Newsletter

13

one can derive potential stress scenarios with confidence intervals (like for a Value at Risk computation) as shown in Figure 1 below:

100 000

In the final step, we will define the liquidity profile of the fund by classifying the securities in the portfolio by liquidity level. This will allow us to define whether, in

FIGURE 1: Expected and stressed redemptions trends

0

Number of subscribed/(redeemed) shares

0 -100 000

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

-200 000

-300 000

-400 000

-500 000

-600 000 Expected trend 67% interval of confidence

Number of days 95% interval of confidence 99% interval of confidence

This first result quantifies the speed at which shares will be redeemed in a specific stressed environment. In an intermediary step and in order to link this information to the portfolio, the level of cash available in the portfolio at the scenario date will be compared with the trend defined above. We can then determine how many days the fund will be able to pay redemptions using only the cash in the portfolio (see Figure 2).

15 000 000 10 000 000 5 000 000 0 0 -5 000 000 -10 000 000 -15 000 000 -20 000 000 1

times of stress, the fund could be at risk or not. In order to perform that analysis, different indicators will need to be used for diversification purposes. Information like the number of pricing providers, stale prices, CDS spreads, etc., will need to be selected and adequately combined to assess the liquidity of each security in the portfolio. Then in order to quantify the liquidity of the portfolio, we will define scales and rank the securities by liquidity levels and buckets (see Figure 3 overleaf).

Figure 2: Cash evolution by redemptions scenarios

Remaing cash in portfolio

2

3

4

Number of days

Expected trend

67% interval of confidence

95% interval of confidence

99% interval of confidence

Liquidity Risk for Investment Funds: Both Sides of the Story

14

PRiM Risk Newsletter

Liquidity Risk for Investment Funds: Both Sides of the Story

Figure 3: Portfolio - Liquidity buckets

60%

50%

40%

% of portfolio

30%

20%

10%

0% 0 to 1 1 to 2 2 to 3 Liquidity buckets 3 to 4 4 to 5

This will enable the risk manager to assess his final liquidity risk and, should it be required, define the adequate buffers. This practical method allows the (risk) manager and the board members to generate respectively and review signals (that can easily be summarised in scorecards) aiming at identifying potential liquidity risk of the investment funds they manage. That will help them to identify quickly potential issues based on quantitative information that will/could be complemented by more qualitative judgments.

Conclusion

Existing methods to assess liquidity risk are mainly asset driven and thus are limited. Liquidity is not a risk as such; liquidity risk is always a consequence of another "liability driven" event. "Though we might hate to admit it, there are always two sides to every story..."8

8

Phil Collins, Both Sides of The Story, 1993

PRiM Risk Newsletter

PRIM SCHOLARSHIP 2011

Invitation to apply for this year's scholarship

15

A Conference with Professor Philippe Jorion

The PRiM Risk Management Scholarship was awarded for the first time in 2010. It actually comprises two scholarships of 3,000 each ­ one for students enrolled in the Luxembourg School of the Finance, the other for participants in the PRiM/IFBL training programme Financial Risk Management, which is taught by Professors Hübner and Jorion. The scholarships are awarded on a competitive basis. To compete for the scholarship, applicants must submit an essay in English and in MS-Word format with a maximum length of 2,000 words. The scholarship is paid in the form of a reimbursement of fees for the Luxembourg School of the Finance or the PRiM/IFBL training programme Financial Risk Management. The money awarded may not be used for any other purpose. Winners are chosen by the PRiM Board of Directors and all decisions made by the Board are final. The essay topic for the 2011 PRiM Risk Management Scholarship is:

Although in the past financial risk management has developed primarily in the banking sector, it is now becoming an essential component in the asset management industry. How do the key requirements for financial risk management in the banking sector differ from the equivalent needs of the investment fund industry? How do you see this trend developing in the future?

On October 14, 2010, Philippe Jorion, the renowned expert on risk management who is the Chancellor's Professor of Finance at the University of California, Irvine's School of Business, delivered a lecture on "The Adequacy of VaR Calculations in the Fund Industry". Professor Jorion has published numerous books and articles on risk management and is perhaps best known for his famous book Value at Risk: The New Benchmark for Managing Financial Risk, which has become an industry standard. The event was organised and sponsored by PRiM together with the IFBL. The primary objective of Professor Jorion's lecture was to present and evaluate alternative methods for measuring financial market risk in the asset management industry. Professor Jorion began his presentation by analysing the means of measuring leverage across a fund portfolio by using average leverage. He then provided examples of VaR (Value at Risk) measures and accounting for the effect of moving windows and model risk with hedged strategies. Professor Jorion also explained how to design effective stress tests tailored to investment portfolios, how to backtest VaR measures and how to evaluate tradeoffs between fast and slow moving risk forecasts. In the conclusion of his presentation, Professor Jorion underlined the necessity of carefully measuring and monitoring financial risk. Risk managers should not rely on just one method of evaluating risk. They need to use a variety of tools, including exposures, VaR and stress tests. VaR by itself is not a panacea. It has its limitations, but together with other tools it can be a valuable means of measuring risk and thus provide the foundation for managing risk effectively.

The deadline for submitting an essay to compete for the 2011 scholarship is March 31st, 2011. If you have any questions or would like to submit an essay for the scholarship competition, please send an E-mail to: [email protected] prim.lu. Please mention "PRiM Scholarship" in the subject line and provide your complete contact details in your E-mail. Good luck!

PRiM Information

T H E A D E Q U A C Y O F VA R CALCULATIONS IN THE FUND INDUSTRY

16

PRiM Risk Newsletter

KEYNOTE INTERVIEW:

RISK MANAGEMENT FOR INVESTMENT FUNDS

To gain more insight into the issues surrounding risk management for investment funds, PRiM spoke with three key players from different areas in the fund industry: Thomas Nummer, Head of Risk at Allianz Global Investors Luxembourg; Luc Neuberg, Head of Risk & Operations, BCEE Asset Management; Michael May, Deputy Global Head of Risk Management, HSBC Securities Services. All of them are members of the PRiM Board of Directors. Thomas Nummer is also Co-Chairman of the ALFI Risk Management Committee and Luc Neuberg is in charge of the ALFI Sub-committee on Market Risk.

PRiM: Risk management in the fund industry took on a new dimension with the implementation of the Law of 20.12.02. What do you think were the driving forces behind that development?

Luc Neuberg

Keynote Interview

Thomas Nummer

Michael May

Thomas Nummer: The fact that the Law of 20.12.02 allowed UCITS funds to invest in a wider range of financial instruments, such as derivatives and different money market instruments, meant that better controls and risk management were required by European legislators in order to maintain a high level of investor protection. So an improved independent risk management can be seen as the logical consequence of accepting higher levels of risk among a fund's investments. The focus on risk management in the Law of 20.12.02 had a very strong influence on management companies, which already then received a European passport, although the requirements concerning risk management were less detailed than foreseen in UCITS IV. Chapter 13 in the Law of 20.12.02, the Commission recommendation (2004/383/EC) from that time and the CSSF Circular 05/176 provided more detailed guidance and thus were a further impetus for improving risk

PRiM Risk Newsletter

17

management for the so-called UCITS III funds improving risk management for the so called UCITS III funds in 2004 and 2005. The current CSSF circular 07/308 issued in 2007 has given more detailed guidance on the organisational and also technical requirements concerning UCITS risk management.

risk categories affecting a UCITS fund.

For example, I believe that the so-called global exposure limit/market risk limit is a risk category that needs further improvement. The different standard approaches used for the calculation of global exposure, i.e., the commitment approach and the VaR-based approach, remain an area, in which a better understanding is needed. Most funds were Luc Neuberg: The changes regarding risk accustomed to applying a commitment management in the Law of 20.12.02 are approach to determine their global indicative of a better recognition of the exposure in relation to derivatives only. importance of risk management in the fund The introduction of a VaR-based approach industry. Fund managers and fund boards applied to all assets of a fund, have come to realise that they including derivatives, which need to monitor not just the "Fund managers and caused people to think more performance of their funds, fund boards have about risk management but also the level of risk to come to realise and its importance for which a fund is exposed. funds. In addition to the that they need to Michael May: A significant expertise, it is monitor not just the technical to understand the aspect of this development important performance of their limitations/assumptions of was the fact that the Basel II Accord was being formulated both approaches, which is funds, but also the at that time, causing people level of risk to which a important not only for a risk to think more about risk, manager, but also for the fund is exposed." particularly operational people who are responsible risk, and risk management. for a fund or a management - Luc Neuberg company. Although most people in the fund industry at that time were aware of the importance of risk An increased focus on liquidity and pricing management, there was not a lot of detail risk has also been a recent change in the on where and how risk management fits fund industry, primarily attributable to into the world of investment funds. the financial crisis. In most areas of risk management, the fund industry needs to PRiM: In which areas do you think improve. Banks are more experienced at improvements are still needed most? risk management than funds. They know that good risk management is a necessity, Thomas Nummer: Beside the various not just a routine for complying with technical aspects, one should clearly regulations. Funds need to learn from highlight as an area of improvement the banks. Although each person may view need to understand risk management risk management differently in accordance through a holistic approach. Each with his individual situation, there must management company/risk manager needs be standards and common principles for to set a risk framework and strategy in order applying risk management in both banks to build the `house' first and in a second layer and funds. to think about improvements in relation to

Keynote Interview

18

PRiM Risk Newsletter

Keynote Interview

Luc Neuberg: Certainly the area of of this process are the value and influence liquidity risk needs to be developed. given to risk managers, which are usually There are very few standards on how we determined by the corporate culture of a measure and understand liquidity risk. In bank. One significant area that is in need the fund industry, risk management is a of development is fiduciary risk. Although young but fast growing discipline largely the AIFM Directive is likely to provoke dependent on tools and methods from numerous discussions around fiduciary risk, other areas of risk management, such as we need a clear definition of the concept, VaR calculation and general methods of which can serve as the basis for improving portfolio management. We need to develop our methods of managing fiduciary risk. new methods and tools that are specific to Far too often risk management resorts to funds. Just like the "Quants" who helped "ticking boxes" or simply following agreed explain portfolio risk, we need to define processes and routines without being clearly how risk management applies to the creative and thinking about what is really everyday operation of a fund. Particularly needed. in Luxembourg, we need PRiM: The need for risk to adapt a very practical "Far too often risk management in funds approach to risk management can differ significantly in funds rather than focusing management resorts in accordance with the just on portfolio risk. to "ticking boxes" perspective from which or simply following Michael May: Liquidity it is viewed (e.g., from the risk definitely needs agreed processes and perspective of an investor improvement, which clearly manager routines without being or a fund Do you or a illustrates the difference custodian). think creative and thinking that the Luxembourg fund between risk management in banks and funds. Among industry is well prepared to about what is really fund managers and boards, meet the requirements of the needed." the focus has traditionally diverse players? What are the been on portfolio risk more strengths and weaknesses - Michael May of Luxembourg's current than on the risks that funds face on an everyday basis. position? In banks, on the other hand, the primary Thomas Nummer: focus has been on credit risk. Only in the Luxembourg is an excellent place to financial crisis did liquidity risk suddenly launch and operate a fund. The industry become a major concern of both banks has built up a unique pool of knowledge and funds. Funds had a real problem on managing UCITS funds, including responding to the challenges of liquidity sophisticated products. There are only a risk, because it was something new for few places in Europe ­ if any others at all ­ them. Banks are pushing hard to develop where such a pool of experience, including better methods of risk management, risk management for funds, can be found. particularly for fighting systemic risk (cf., That is surely one of Luxembourg's principal Basel III). Because banks have major capital strengths. requirements, they need to manage risk very effectively. The challenge for the fund Although many fund managers are industry is to adapt the risk management represented in Luxembourg, key parts of banks for use in investment funds. Part

PRiM Risk Newsletter

19

of the value chain such as designing a evolve to a continuous risk monitoring and fund, managing the fund's portfolio and not just an answer to some legal restrictions. managing the related portfolio risks are The size of the Luxembourg fund market often performed by the parent company and the diversity of the fund business make or other outsourcing partners in a foreign the harmonisation of controls extremely country. In such an environment with difficult. considerable cross-border delegation, Michael May: As the managing risk adequately biggest cross-border and capturing all relevant risk "Although many fund market in the world, categories is indeed a major Luxembourg should have challenge. fund managers are a focus that is different represented in Fund managers launching from asset management. Luxembourg, key parts Fund managers can choose Luxembourg funds expect management companies of the value chain [...] different domiciles, but to perform adequate, in are often performed for those funds set up to independent oversight of Luxembourg we need by the parent all risk categories, even if focus on fund operations. The some of them are managed flexibility of the Luxembourg company or other by external parties. In encourages outsourcing partners market addition, it is also expected outsourcing, which we must in a foreign country. consider a major area of focus. that we demonstrate a good understanding of operational [...] With considerable In this context, Luxembourg risks, particularly the risks is a "safe" location for creating cross-border associated with a `typical' a fund, because of the delegation, managing effective supervision of the Luxembourg outsourcing risk adequately and business model. Luxembourg fund market. capturing all relevant Further advantages of In the ALFI Risk Committee Luxembourg that should be (which includes four risk categories is [...] a developed are its flexibility, major challenge." dedicated working groups the expertise available in the on market risk, counterparty/ market and the wide variety issuer risk, liquidity risk and - Thomas Nummer of languages. operational risk), we look at risk management through PRiM: Recent improvements a holistic approach across the entire in risk management for funds have resulted fund industry. One primary focus in 2010 in increased costs. What do you think will be were the operational issues that are most the commercial impact of this trend in the important for Luxembourg: outsourcing long term? risk, operational processes concerning Thomas Nummer: The increased regulatory counterparty risk (particularly collateral requirements will generally trigger an management) and operational risk in additional workload in risk management, general. which means that all of us will have to Luc Neuberg: This holistic approach is deal with increased costs. But since the certainly a central point of the new legal regulatory landscape across Europe and the requirements. Risk management should world is changing, it is not a trend that will

Keynote Interview

20

PRiM Risk Newsletter

Keynote Interview

affect funds in Luxembourg more than in other jurisdictions.

Investors, particularly institutional investors, increasingly expect not only performance, but also inter alia effective risk management from asset managers. So a strong risk management culture will be seen as a competitive advantage for an PRiM: UCITS asset manager and for an entire market in process that general. It is also important to understand that the "Portfolio optimisation regulations do not require is also a reduction all asset managers to have the same standards of risk of risk. From the management, since it is also a investor's point of question of proportionality.

business models will change. Consolidation (particularly after the implementation of UCITS IV) will become increasingly important, but scalability will become a problem. Managing all risks will be prohibitively expensive, but some degree of risk management will remain a necessity. IV is likely to reinforce the started with the Law of 20.12.02. What changes in risk management for the fund industry do you expect to see in the future?

Thomas Nummer: Through UCITS IV, an independent view, alpha was used risk management function Luc Neuberg: Companies will become an integral part as a commercial traditionally viewed risk of a management company argument during management as a cost (not even more so than in the the last decade, profit) centre. Now it is seen past. In this context, I think even if it is only a as a necessity and even as that in addition to portfolio a source of revenue. Risk theoretical measure. risk the management management improves the of outsourcing risk and The monitoring of quality and safety of a fund, operational risk in general beta should also be which is in the interest of the will become increasingly fund's investors. Portfolio used commercially and important in Luxembourg. optimisation is also a reduction Risk management should presented as a real of risk. From the investor's not be viewed only as a added value." point of view, alpha was used control function; it should as a commercial argument also be understood as an - Luc Neuberg independent function that during the last decade, even if it is only a theoretical supports business, just like measure. The monitoring of beta should product development and operations. also be used commercially and presented We should also expect that regulators as a real added value. Risk management and investors will require an adequate, actually produces a reduction of costs in independent risk management for nonportfolio management, implying that the UCITS, which will surely be triggered by risk manager can become instrumental in the Alternative Investment Fund Managers ensuring the good performance of a fund. Directive. Michael May: Although the requirement It will be a major challenge for the fund for a fund to perform will not disappear, industry to attract risk managers, directors investors are becoming more conservative and/or conducting officers who understand and they are willing to accept lower returns risk management from a holistic point of in favour of greater security. It is likely that

PRiM Risk Newsletter

21

view, i.e., who have knowledge of both portfolio risk and operational risk. Luc Neuberg: I think the trend is towards decreasing complexity (cf., the CESR guideline from July 28, 2010). Funds need to become more transparent and less complex. This trend also applies to risk management. We need to make risk management less esoteric and more open, so that anyone can understand it. Risk management must be a continuous process that runs from the writing of a prospectus to the calculation of a NAV.

"It will be a major challenge for the fund industry to attract risk managers, directors and/or conducting officers [...]who have knowledge of both portfolio risk and operational risk."

- Thomas Nummer

Michael May: To a large degree, risk management is common sense, if you have operational experience. We do not need more "Quants", without wishing to diminish the importance of their contribution, but it is essential we have risk managers with operational experience. Oversight requirements for custodians are becoming "To a large degree, more demanding, following risk management is the Madoff scandal and more recently the approval of the common sense, if AIFM Directive. In this context, you have operational fiduciary requirements are experience. [...] likely to change, so that there will be no distinction between It is essential we fiduciary responsibilities and have risk managers operational service provision. PRiM: Thank you for sharing your views with us.

with operational experience."

- Michael May

Keynote Interview

22

PRiM Risk Newsletter

PRiM NEWS

PRiM News

PRiM wishes all its readers and members a very happy and prosperous 2011!

Membership renewals go out to all members in January each year and should now have been received by all. Please contact the PRiM secretariat if you have not received a renewal advice (or if you no longer wish to be a member) by e-mail to [email protected] We thank all members for their continuing support and we look forward to working together with you during 2011. Any ideas for events or other initiatives that you have are always welcome. Please send them to Marco Zwick, the President of PRiM, or to any member of the PRiM Board of Directors. Visit our website at www.prim.lu.

Renewal of membership

NEXT EDITION: March 2011 THEME: E-risk in the Financial Services Industry CONTRIBUTORS WELCOME

PRiM NEWS

Information

22 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

602342


You might also be interested in

BETA
Microsoft Word - C1140LUX-22-11-2011 AIFM LEVEL2 - - December 1 Luxembourg.doc
Microsoft Word - HF Industry Trends Q4 2009 _3_.doc