UCITS III: risk management takes centre stage

Competence at investment risk management is set to be a key determinant of competitive success in a post UCITS III environment. With alternative strategies now available to traditional managers the hedge and long-only worlds will collide and the winners will be those that can combine alpha with sound risk controls. While alpha generation is amenable to human judgement, risk control requires smart risk measurement systems that meet the needs of fund managers, clients and regulators.

UCITS III represents a regulatory revolution. Undertakings for Collective Investment in Transferable Securities are now to be allowed to be wholly invested in non-transferable derivatives. It was hedge funds that caused the about turn, storming the investor market and carrying off the riches with average fee rates of above 2%, around five times that achieved by traditional managers . In response, UCITS III allows regulated funds, free to be sold to the retail market across the EU, to implement a range of strategies previously the preserve of the unregulated, “alternative” world.

In a post UCITS III world regulated firms can design a much wider range of investment products to meet client needs. Derivatives can be used to gear up, to reduce risk and to manufacture returns uncorrelated to security markets. Long-short equity, one of the most popular hedge fund strategies, is now feasible in a regulated environment. Relative Value (long-short bonds) offers traditional fixed income fund managers a means of generating superior returns in a flat or rising yield market. Absolute Returns will soon be available at a fund supermarket near you.

All EU based investment funds must convert to the UCITS III rulebook by mid-February 2007. Up to now it is probable that the focus of the product development teams will have been on the legal and administrative work involved simply in making existing funds compliant with UCITS III. It seems certain that from early next year the product developers will seek to follow those among their peers who have already launched regulated long-short equity and fixed income funds, and that thereafter innovation will follow an exponential growth path.

Many sceptics still disbelieve the case for alternative strategies. However, the case is powerful both from a theoretical and practical perspective. Morgan Stanley has recently demonstrated that portfolios with a short element can significantly improve the potential alpha . From the financial adviser’s point of view, the ability to buy a geared fund for the adventurous client and a limited loss fund for the cautious fits with their obligation to risk profile their clients and is highly compelling.

Consequently, UCITS III is expected to transform the European asset management industry over the next few years, and the “winners” are likely to be those who take the most advantage of the new investment freedoms it offers. This is as likely to be hedge funds moving under the UCITS III umbrella to gain access to the larger pool of assets that will only buy regulated funds, as it is to be traditional houses seeking the higher fee levels promised by alternative strategies. However, regardless of the manager’s background, a key challenge will be the successful and cost-efficient implementation of the regulatory requirement of a formal “Risk Management Process”.

Risk Management Process

Where a firm manages a “sophisticated” fund, one using derivatives extensively for investment, the regulations require that its Risk Management Process (RMP), must, in particular, incorporate a daily calculation of the “Value at Risk” (VaR) of the fund. And that’s where it gets interesting. Having been, pre Unilever-Mercury (2001), almost an irrelevance, integrated risk management is moving from what might be described as “an essential nuisance” to becoming a key a weapon of competitive advantage. If you can’t do VaR you will be cut out of the myriad opportunities to tailor funds to client needs that the use of derivatives offers.

The EU made certain stipulations about the specific method used to calculate VaR, but left a significant degree of freedom to individual jurisdictions, in the hope that, in time, an agreed best practice would emerge. However, in the meantime, there is the potential for different countries to apply different standards, causing complication for managers wishing to base funds in different jurisdictions, and so having to select systems that can meet differing sets of requirements.

For example, Table 1 compares and contrasts the parameters for the VaR calculation as applying in Dublin, where the regulator has provided its own specific guidance, and the UK, where the FSA’s rulebook refers back to the EU directive. Although appearing similar, the requirements in relation to the lookback and lookforward periods are worded to have quite different effects. Dublin, by default, makes a 1 year lookback the minimum, while the UK’s FSA, copying the EU directive, defaults to a 1 year maximum. Dublin provides more flexibility on lookforward, making 1 month simply the maximum, with the manager able to choose a shorter period if liquidity makes it appropriate, while the UK is rigidly stuck to 1 month. Dublin is less flexible in relation to the maximum level of VaR for a fund with no benchmark (an absolute return objective), stipulating that this must not exceed 5%. On the other hand, the UK leaves it to the firm to determine what level of VaR is compatible with the risk profile of the fund.

Table 1: VaR requirements in Dublin and UK

Some market experts have interpreted the combination of a 99% confidence level and a 1 year lookback as endorsing and requiring a Monte-Carlo method for calculating VaR as a simple historical simulation method may not have sufficient data available to make the calculation robust. It would appear that Dublin recognised this difficulty and hence have encouraged a longer lookback which is compatible with other VaR estimation methods.

Another requirement arising out of UCITS III is to carry out periodic “stress tests” on investment products to, as far as possible, ensure that funds will behave as advertised and consumer confidence will not be damaged. This seems to be akin to the “crash testing” that a new car must undergo before being sold to the public, and is a further fundamental requirement of a risk measurement system.

Banks vs Investment managers

No doubt the explicit mention of VaR in the EU Directive was a consequence of the influence of banks, for whom it is now reasonably standard to use VaR to monitor their trading books. VaR no doubt caught on in this space because the primary concern was over the potential short term absolute loss measured against a bank’s capital, recognising also that derivatives may have unusual payoffs that are not readily amenable to modelling.

Asset managers have a different set of concerns. Their products are typically marketed by reference to a market benchmark and it is therefore necessary to measure the “tracking error” – the risk of deviation from the benchmark. This is a fundamental requirement of the clients of the asset managers, be they pension funds, professional fund of funds managers, or professional advisers on behalf of individuals. Since any one fund or mandate is likely only a small part of a larger portfolio, if that fund does not behave with a degree of predictability, it plays havoc with the planner’s asset allocation intentions.

Furthermore, the existence of an external client with an interest in the risk and performance that they are experiencing means asset managers need risk systems that enable them to analyse and characterise the risk bets that they are taking. Finally, asset managers will have a longer term time horizon than a bank. This is likely to be in the region of 3 to 5 years, the typical holding or review period of the underlying client. This longer time horizon means asset managers will want a risk measurement system that puts them in a position to exploit, rather than get whipsawed by, short term changes in market volatility.

The consequence of these different requirements is that asset managers have tended to focus on factor model based risk measurement systems. By mapping the system’s factors onto investment characteristics that are relevant to the particular portfolio, the manager is better placed to understand, control and communicate the bets that are being taken. These bets may relate to the level of market risk, country or sector risk, the investment style or they may be bottom up and stock specific. Factor based systems can also assist in portfolio construction, either simply by indicating which positions should be traded for the best risk:return improvement, or through full blown optimisation.

Hence, the need to produce a VaR on a sophisticated fund does not eliminate the need for reporting on the tracking error and other descriptive information, tasks for which a typical VaR system is not well equipped. So, to compete in the post-UCITS III world, a risk manager needs to source a risk measurement system with the following characteristics:

Risk measurement system requirements

o Report tracking errors, volatility and VaR
o Calculate using different time periods
o Produce historical simulation and Monte-Carlo VaR
o Attribution by country and sector
o User defined investment style analysis
o What if process to test prospective trades
o Able to capture up to date market prices
o Stress testing

MiFID

Following hot on the heels of the February 2007 deadline for conversion to UCITS III, MiFID (Markets in Financial Instruments Directive) replaces the Investment Services Directive (ISD) and is expected to be implemented by November 2007. MiFID widens the range of ‘core’ investment services and activities that can be passported between EU states, and improves the operation of the ‘passport’ for investment firms by more clearly delineating the allocation of responsibility between home state and host state. It should act to further reduce the barriers to cross border competition and so present asset management firms with both significant threats and opportunities.

Conclusion

With UCITS III fully in force from February 2007 and MiFID following shortly thereafter, the EU regulatory authorities are creating an environment that is likely to be characterised by intense product innovation and cross border competition. Experts agree that the confluence of these factors moves risk management to centre stage in the drive for competitive advantage, as without advanced risk systems product innovation opportunities will be limited. To stay ahead, risk managers must now seek out risk measurement systems that can assist in controlling benchmark relative risk, which provide the descriptive statistics that meet the needs of the fund managers and their clients and which also offer the ability to deliver the various forms of the VaR calculation required by different regulators.

Tim Wilding
Director