Filling the ratings void

Unlike bond investors, structured products investors lack the benefit of industry-standard risk ratings. But with investors, IFAs and distributors all demanding change, Germany's investment banking industry, analytic firms and ratings agencies across Europe are developing their own risk-mapping tools. Shamillia Sivathambu reports

According to a survey carried out in the July/August issue of Structured Products, an increasing number of independent financial advisers (IFAs) in the UK are keen to see a standard rating system introduced for structured investment products. IFAs effectively want structured products to come with a risk rating akin to that already available for bond investors. "Any standardisation is useful to advisers. The sentiment (endorsement from a third-party) is right, which is to create a level playing field for risk ratings," says Mark Peters, head of research for IFA network Sesame.

Investors are still suffering a hangover from the late 1990s bear market and the precipice bond debacle, so risk exposure is a concern that still lingers in their minds. And though structured products are known for their safety-net features, any product that offers exposure to equity markets comes with a certain degree of risk. As a result, the whole industry is responding to customer demand and moving towards a closer analysis of risk. And some market participants now say that a standard risk-ratings body should be established. "The industry is crying out for an independent ratings board," declares David Stuff, director and head of structured products at London-based Barclays Capital.

Between 15 and 20 products are brought to market each week in the UK alone, and because they are quite similar it can be difficult for the IFA to distinguish, evaluate and ultimately recommend a suitable product to clients, says Jamie Vale, London-based investment sales director of structured products at Legal & General. With a uniform ratings standard, it would be easier for IFAs to recommend one product over another, he adds.

But the question, according to Neil Saunders, manager of structured retail products at Edinburgh-based Royal Bank of Scotland (RBS), is who will provide the rating? And which process would work best when making such assessments?

Germany steps into the breach

While the UK grapples with drawing up a blueprint, Germany is implementing what looks increasingly like the country's first standard ratings system for structured products. The European Business School (EBS), on behalf of Germany's Derivate Forum, which comprises the country's five leading investment banks, has built and launched a risk-calculation tool that rates warrants and certificates according to five risk classes.

Launched in mid-July and aimed at the private investor, the EBS calculates risk on a value-at-risk (VaR) principle using Monte Carlo Simulation, which randomly generates values for uncertain variables over and over to simulate a model. The varying loss probabilities (probability distributions) are then allocated to a product, resulting in the creation of five risk classes. Class one is the safest and class five consists of the more speculative products, explains Siegfried Piel, managing director of trading and derivatives at Frankfurt-based bank Sal Oppenheim.

"The idea is make these products transparent without having a PhD in mathematics," says Stefan Armbruster, director of the investment products group at Frankfurt-based Deutsche Bank. "Investors tend to buy into these products based on gut instinct but we want the market to move towards making decisions that are calculated," he adds. A risk-rating tool will heighten investor confidence and encourage them to further invest in structured products, he believes.

It helps that the ratings standard is being backed by Germany's five largest investment banks (Deutsche Bank, DZ bank, HVB, Sal Oppenheim and West LB), which collectively control the majority of Germany's structured products market. A total of 13,522 derivatives securities are expected to be rated by the end of the year, 90% of which have already been rated and posted on the Derivate Forum website for private investors to access free of charge, Piel says.

Piel expects at least two or three more banks to use the tool by the end of this year. Investment banks already use VaR models for risk management, so the process is familiar to them, he says. "It could also be looked at by the regulator and endorsed as a standard ratings system."

Piel's optimism does not stop there. "Germany is an important market for structured products, but this does not mean the tool can't be developed in other European countries too," he says. He hints at Switzerland's possible interest. "Switzerland is not a transparent market. It cannot cope with more than 4,000 products on the exchange, so they could be interested in a ratings tool like this."

In the interim, Piel believes the new ratings tool will establish itself as an industry standard as private wealth managers begin to adopt it. And evidence of this already exists. Frankfurt-based Lutz Gebser, chairman of the Independent Asset Consultants Association, says: "With the new risk classification from Derivate Forum, we can now give our clients' securities accounts even better tailored structures and adapt them to individual choices and risk profiles. We would recommend this risk classification system to our members."

Other contenders

Germany's Derivate Forum might be the only body to come close to issuing anything like an industry standard when it comes to rating structured products, but with the whole market moving towards a closer analysis of risk, a number of private analytic companies are fast introducing risk ratings for structured products.

UK-based Future Value Consultants (FVC) introduced risk-mapping to its analysis of structured products 18 months ago. Previously, its ratings were based on a product's estimated asset value, transparency and prospective returns. Risk evaluation was only introduced following the effects of the bear market, says Tim Mortimer, FVC's managing director. "We rate products from up to 25 product providers and more than 4,000 IFAs use our ratings," he adds.

Just like the Derivate Forum and EBS, FVC also applies a VaR model when calculating risk. Risk profiles are broken into five categories – 0, 10, 20, 30 and 50 – with 50 representing higher-risk products such as futures funds and single stock-linked products, Mortimer says.

Product providers must pay a fee for the rating but the information is available free of charge to the IFA community. "IFAs can spend only a limited amount of resources on assessing the risk and value of products, so they need the help," he adds.

With the market moving increasingly towards open architecture, in which distributors sell a mix of their own and third-party products, rating standards seem a logical investment for product providers to make. "This means that our proprietary products can be translated into the categories of other banks and gain access to entirely new distribution channels," says Deutsche's Armbruster.

And across the Channel there is France's Testé Pour Vous (TPV). TPV is an independent body acting under the Euro Consumer Group and the Consumer Act, responsible for publishing financial analyses for consumers and financial firms. It has also started to evaluate the risk profile of structured products in the French market.

Like the FVC and the Derivate Forum, TPV also uses the VaR model when calculating the degree of risk attached to products. But unlike others, TPV selects and rates only two or three structured products each month. These tend to be products that are heavily advertised or slightly unusual, according to Anne Morin, TVP's director. "If they are deemed too risky, TVP will recommend alternatives, which might be structured products or something else, such as unit-linked products," she says.

Tradinglab, the investment banking arm of Banca Unicredito, also rates structured products, but this is confined to proprietary products. Tradinglab also employs a VaR model when calculating risk.

Diverging methodologies

While the VaR model may be a popular means of evaluating the risk profile of some structured products, others have found the methodology to be somewhat lacking, namely when evaluating gap risk, which is the risk of sudden, large and unprecedented changes in spreads.

VaR models, which are widely used to measure market risk, seem to work well in normal conditions but not during financial crises, which is arguably when it is most necessary to know how much value is at risk, explains London-based Cheiron Osako, director of European structured finance at Fitch Ratings. "By relying heavily on historical data, the VaR method does not always fully capture the risk of sudden market swings beyond that experienced historically," he adds.

When calculating risk for structures where gap risk is significant, Fitch employs the use of Extreme Value Theory (EVT), a methodology usually used to calculate the probability of sudden and extreme loss of catastrophe bonds. In the context of structured products, namely CPPI products, this involves stressing the data to calculate the probability of extreme loss over a short period of time.

The ratings will then reflect the product's likelihood to default, with an AAA rating being of the "highest credit quality" and an AA-rated structure being of a "very high credit quality", Osako explains.

Measuring market value risk rather than just credit risk is a trend that is common to the three largest ratings agencies. Standard & Poor's (S&P) and Moody's have also begun rating structured products by evaluating their spreads. "Although market value risk relies in part on historical data and historical volatility, we (S&P) go a step further from the traditional VaR model by looking at longer data histories and more extreme events to capture movements not always present in standard historical data," says Perry Inglis, London-based managing director at S&P.

All three ratings agencies have rated an average of two public or private CPPI structures to date, and requests from arrangers and investors to have such products rated are steadily growing in number, notes Paris-based Olivier Toutain, vice-president and senior analyst at Moody's.

When talking about creating an industry standard, all three agencies are confident that the market will turn to them for an answer. "In order for a rating to be considered an industry standard it must be recognised by the regulator and the Basel Accord. Ratings agencies, which are Nationally Recognised Statistical Ratings Organisations, are recognised by both," explains a London-based senior analyst from one of the three leading ratings agencies.

Getting on-side with the regulator

The regulators, however, are reluctant to endorse such a standard. "This is not a question we would address," says a source at German regulatory body Bundesanstalt für Finanzdienstleistungsaufsicht (BaFin). "We are interested in understanding how these ratings agencies come to their conclusions, but endorsing such a standard is not a priority for us. It goes beyond BaFin's remit as regulation enforcers."

This sentiment is echoed by the French regulator Autorité des Marchés Financiers (AMF). "It is up to the investor if they want to buy a product and we do not influence that decision," says a spokeswoman from AMF. The product provider has an obligation to provide all the necessary product information and each structured product is stamped with a warning from the regulator, notifying investors that they are buying into "risky" investments. But beyond those requirements, it's up to the provider if a risk rating should be attached as well, the spokeswoman explains.

In the interest of consumer protection, the UK's Financial Services Authority's (FSA) Select Committee initially recommended enforcing a "traffic light" code for retail structured products. Red was to signify very-high-risk products, yellow was for products that offered some capital protection and green would indicate secure investments featuring 100% capital protection, explains a spokesman from the FSA. "But this was a massive over-simplification. What is risky for one person may not be risky for another. The FSA is not a product regulator," he adds.

While the FSA acknowledges its shortcomings in delivering useful ratings standards, it is nonetheless sceptical about empowering one of the ratings agencies to do the job. "Ratings agencies are free agents, they sell a service. If they were to provide an industry standard, the debate arises as to whether or not these agencies should therefore be regulated by us," the FSA spokesman says. There is also a concern that the rating standards will be used when marketing or selling products to retail investors, which would contravene financial advice rules, he adds.

While ratings agencies tend to rate CPPI products as wholesale products, they have no control over how that rating is repackaged, says S&P's Inglis. "It would be a real concern to ratings agencies if the rating, which is intended for the wholesale market, was repackaged and sold to the retail market," he adds.

"This is why you cannot push for an industry standard without looking at financial advice rules and how ratings will be used by the adviser," the FSA spokesman concludes.

Simplicity is key

While investment banks and product manufacturers are keen to see the birth of an independent ratings standard which can be applied across the board and simultaneously assist them when distributing products through the wholesale market, a number of large IFA networks and banks with captive retail distribution remain "cautiously optimistic." With the larger IFAs and banks already rating products in-house, there is concern that an industry standard might be too complex and academic for the retail product pushers.

"While the advice awarded by an independent ratings agency will be useful, the mechanics and process of the ratings tool must be examined," Mark Peters from Sesame explains. The ratings would be used as part of an IFA's larger sales proposition so it's important they understand how the risk is calculated, he adds. Sesame's in-house team measures a product's risk profile based on the credit rating of the underlying provider.

Ian Lowes, managing director of UK-based IFA network Lowes Financial Management, says his internal ratings team relies on common sense rather than quant models when labelling products as "endorsed" or "unendorsed". The team uses simple risk-reward analysis, taking into account where the market is today and where it might be in the future. "There is a degree of subjectivity there, but risk calculations shouldn't be rocket science," Lowe says.

Others take this pragmatic approach further. "The best way to control risk is by controlling the product," says Neil Saunders from RBS. If you don't give the sales force risky products then you limit the potential damage, he says.

Who rates structured products?

European Business School and Derivate Forum

www.derivate-forum.de

Products rated: more than 13,000 certificates and warrants from Deutsche Bank, DZ Bank, Sal Oppenheim, HVB and West LB

In which markets: Germany

Methodology: VaR

Similar to TPV's methodology (see page 13), the VaR is calculated by using historical data and Monte Carlo Simulation. The product is broken down so that EBS can calculate the risk of each component (underlying asset). Interest rate risks, volatility risks, currency risks and default risks are calculated by looking at historical data. As soon as the probabilities show a negative against the benchmark (eg. DJ Eurostoxx 50, S&P 500), a risk rating is put in place. If a product produces a VaR that ranges between 0 and 2.5%, it is categorised under risk class 1, which is the safest. A VaR ranging between 17.5% and 100% is considered highly speculative and is rated under risk class 5.

Fitch Ratings

www.fitchratings.com

Products rated: Two private CPPI deals

In which markets: Europe

Methodology: Extreme Value Theory

Going one step further than S&P or Moody's (see page 13), Fitch sometimes employs the use of Extreme Value Theory (EVT) when calculating risk for structures with significant gap risk. It is usually used to calculate the probability of sudden, extreme loss of catastrophe bonds. When dealing with extreme loss over a short period of time, using a normal distribution calibrated on historical data may produce optimistic simulations, so Fitch stresses the data to calculate the probability of extreme loss over a short period of time. The ratings will then reflect the product's likelihood to default, with an AAA rating being of the "highest credit quality" and an AA-rated structure being of a "very high credit quality", Osako says.

Future Value Consultants

www.futurevc.co.uk

Products rated: Over 500 CPPI, zero coupon and fund-linked structures from Barclays Capital, Keydata, Abbey etc.

In which markets: UK

Methodology: VaR (same as EBS and Derivate Forum) with five risk classes that range from 0 to 50, with 0 being the most secure. In addition to rating a product's risk, FVC rates its value, transparency and prospective returns. To estimate value, FVC uses its own models and market date, while returns are determined by assessing future market volatility and equity forecasts. A product's construction determines its transparency.

Moody's Investors Service

www.moodys.com

Products rated: One public deal for BNP Paribas' CPPI Dynamo Series 1 and an undisclosed number of private "CPPI-like" deals.

In which markets: Europe

Methodology: Spread evaluation

Similar to the methodology applied by S&P, Moody's calculates market value risk for structures without capital guarantee by looking at long-term and short-term spreads. It calculates the spread evolution of each asset to determine the price of the underlying at any point in time. This allows them to calculate the loss to the notes issued by the vehicle.

Unlike S&P however, if a structure has a capital guarantee feature, ratings are based on the bank's credit exposure and not market value risk.

Standard & Poor's

www.standardandpoors.com

Products rated: Two private CPPI deals

In which markets: Europe

Methodology: Spread evaluation

Traditionally applied to calculate the market value risk of assets in Sivs, some collateralised debt obligations (CDOs) and, more recently, leveraged super seniors, S&P rates the risk profile of a CPPI structure by simulating defaults, spreads and spread volatilities to determine high confidence levels that are unbreached by the simulations and are reflective of ratings probabilities. Once this has been determined, trigger levels are identified to change the investment mix to a more risk-averse portfolio. This reduces asset default probabilities and simulated portfolio returns. The ratings will reflect the product's ability to withstand severe spread movements before it defaults.

Testé Pour Vous (TPV)

www.testepourvous.com

Products rated: Rates two to three structured products a months including CPPI, zero coupon and fund-linked structures from BNP Paribas, Sinopia Asset Management Axa Investment Managers etc.

In which markets: France

Methodology: Value-At-Risk (VaR)

Unlike the three main ratings agencies, the simulated data is not stressed and for this TPV uses the standard VaR model. VaR is expressed as a multiple of the standard deviation of the portfolio's return. It is widely used for risk management by banks and other financial institutions. By using historical market trends, volatilities and Monte Carlo Simulation, VaR uses complex algorithms to calculate the maximum loss a derivative instrument could incur in a day's trading. Using historical data, Monte Carlo expresses returns as a histogram of hypothetical returns.