13 Oct 2010, Liz Duggan , Ian Blance , Brian Sentance , Interactive Data , SIX Telekurs , Xenomorph , Risk magazine
Interactive Data Liz Duggan, Managing Director, Evaluations
SIX Telekurs Ian Blance, Head of Evaluated Pricing Business Development
Xenomorph Brian Sentance, Chief Executive Officer
Risk: What is considered best practice today in terms of the number of third-party pricing sources that buy-side firms should be using on the transaction?
Ian Blance, SIX Telekurs (IB): I think things have changed quite a bit in the past few years. During the height of the crisis, we saw that reliance on a single source of any kind was proven to be not particularly sound practice. It’s difficult to say how many sources would be acceptable, but certainly more than one. Across the board – and that applies for evaluations and other data items as well – having some kind of plan B is always a good exercise. It probably depends on the kind of instrument you’re talking about as to what people would feel comfortable with. The more complex and illiquid a deal, the more need there is for multiple price sources.
Liz Duggan, Interactive Data (LD): I’d add another dimension to that – in addition to the type of asset class, there is also the market segment or the type of client. In the US, for instance, the mutual fund market segment, as a best practice, typically will have two, and sometimes three, pricing vendors for an asset class as Ian observed. But then you’ll have another camp that says: “I want full transparency and full understanding of one vendor because, if I have two vendors and I have a difference of opinion, then what do I do?” An internal process needs to be wrapped around that.
So what I’ve found in the US – in particular since Topic 820 (what used to be called FAS 157) – is you’ll have two camps: one that says yes, two or more vendors is a preference; and those that will say I want one vendor and I want to know and understand that vendor in as much detail as possible.
Risk: How are the regulatory changes impacting buy-side appetite for complex financial products and evaluated pricing services in the US and the European Union?
LD: The new regulations are definitely having an impact by making the buy side more open to embracing third-party evaluations, especially in the European and the Asia-Pacific regions. In the US, especially in the mutual fund industry, there has been a long-term adoption of third-party evaluations. What I find is that there is growing acceptance over here in Europe and we expect the regulatory influence to continue to accelerate the adoption of third-party evaluations.
IB: I don’t think that complex should be used as a proxy for bad or evil. The fact is that, if a product is complex, provided it’s transparent and it’s clear and it’s understood, then it can usually be valued in a reasonably comfortable fashion. If it’s complex but it’s opaque and not really well understood, then that kind of stuff should come under a lot more scrutiny. The structured products market, for instance, can include quite complex instruments in terms of delivering client requirement for risk/return profile but, if they’re understood and everyone knows exactly how they are created and how they should be valued, if you’ve got pricing sources and you’ve got data, then I think people can justify the continued investment in that type of deal. However, I don’t think anybody is going to be able to justify getting into a product that they don’t understand, that they can’t really explain what the collateral is, that they can’t really understand what the payout scenarios might look like any more.
Risk: Why is transparency so important to evaluating pricing clients, and how is the industry working to deliver on this?
LD: When you talk about transparency, I think it’s very similar to what we were talking about with respect to acceptance of evaluations. There are many different forms of transparency and I think some of the basics that buy-side clients start with include understanding the evaluator’s approach, their process and their methodology. Once they have a good understanding of that plus the processes you have in place, the compliance programmes you have in place, the business continuity you have in place – then they have the basic foundation.
Buy-side firms can then turn their attention to the vendor’s daily interpretation of the fixed-income markets and how are they communicating that out to clients. Clients are also spending a lot of time on the accounting guidance and how that relates to evaluated prices, whether it’s FASB in the US or IASB in Europe. Vendors are making more information transparent to the buy-side firms regarding what inputs were used to create that evaluation. There are many different things that evaluation firms are doing, including sending clients daily updates about their interpretations of the markets, providing transparency to the inputs used in the evaluation process, and creating online tools for clients and their auditors to access that information. That is just a sample of some of the efforts that are going on for transparency.
Brian Sentance, Xenomorph (BS): Another aspect from a data perspective is the way in which the data gets used. I wonder – do you find that clients are requiring the data for, say, risk management purposes?
IB: There are two main reasons why transparency is important. The first is that there’s an external drive from the regulators, from the accounting standards boards, to open that up and the whole notion we talked earlier about of needing to understand the valuation, to be able to categorise it as level one, two or three under International Financial Reporting Standards. And then there is a whole internal movement to be able to understand your risk a lot more. I think a lot of firms that had their fingers burned over the last few years by not adequately understanding their internal risk are using transparency for that purpose; internally, rather than just being forced to do it externally.
Risk: What effects have changes in technology had on evaluated pricing accuracy? How are these improvements helping clients to actively manage their risk?
BS: I think there is a lot more due diligence spent on the pricing methodology from the complex end of things. For the technology aspects, what I see in the medium term is that you’ve got a convergence between the front-office technology, which is currently being used in high-frequency trading. So not so much the low latency but the high-performance computing, enabling you to do faster re-evaluations – which I think is maybe being reflected in terms of the evaluations – people wanting net asset values, people wanting evaluations being done on an intra-day basis. Risk management moving more real time and wanting to have a look more at scenario management, as well as things like value-at-risk. And, basically, the whole question is: let’s look at more data, let’s do more things with it, let’s try and get a better view of how it could possibly react to future changes and, as a result, a lot more of the heavy grunt technology is going to get used eventually in this market.
IB: The only comment I have about the technology side is a growing need for the ability to crunch those numbers faster. It’s obviously something that’s going to help our clients to get their valuations in a much more timely fashion. And, from the perspective of a valuation shop, some instruments require some pretty serious computations, and to have the technology to do that faster and get it out to the client quicker is clearly a huge benefit for them.
Risk: Are the buy-side firms taking active steps to improve their risk management around valuation risk, and are firms overhauling their approach to data or simply tinkering around the edges?
LD: We have observed that firms are being very active in their management of evaluation or pricing because valuation has been in the headlines a great deal post-Lehman. You’ll find firms being very active in managing their understanding of the methodology, their understanding of your process, engaging in due-diligence visits, etc. There is also increased communication between the pricing vendor and the front office with buy-side portfolio managers and traders increasingly engaging the evaluators to share market colour and market sentiment. I’d say that we have seen an increase in that type of activity post-Lehman as a result of what happened in the financial markets.
BS: I think the number of buy-side firms are picking up on doing more of a product control function – it’s a trend that is coming over from the investment bank and that’s getting formalised. We’re seeing more quantitative people being employed by the buy side directly, which is a change from maybe five years ago. I think, in general, on the data management side there has been a lot of convergence between all the different kinds of data, such as market data and the traditional reference data in counterparty data. Maybe another side of things, which is very particular from a valuations point of view, is what I would term as model data, which are the data input parameters. And maybe the slightly more complex end of things in terms of some of the inputs that aren’t quite quotable data but might need some calibration, is another area in which the convergence of data management to manage that is a single process as opposed to lots of different systems. As an alternative to the more traditional approach of using spreadsheets and managing all of the models that way, it is something that is on the agendas of a number of firms.
Risk: The economic and financial crisis that has engulfed the world over the last three years has created greater sensitivity and volatility in valuation results. How are vendors working with clients to help them manage the impact of this volatility on their processes, portfolios and reporting requirements?
BS: My slant on that is that the increased volatility from the point of view of processes that some of our clients have been running, particularly in the area of data quality, has produced a lot more work for them. There are many more exceptions and we’ve been getting a lot more requests for dynamic data quality rules, so rules that would assess exceptions based on the recent performance of the market or the recent volatility of the market. For instance, if you go from, say, five years ago when maybe things were a lot less volatile, heading outwards, and you’re trying to apply those same data quality metrics to a market of, say, two or three years ago, you’re going to have an awful lot more work to do. So the fact that things change daily by a large amount, maybe that’s not something necessarily to worry about because everything else is moving by a large amount. So that is one area where it has changed some of the requirements from the clients.
IB: I think you also need to differentiate between sensitivity and volatility. Sensitivity can apply even in a quiet market. I mentioned earlier, depending on the structure of a particular instrument, that small changes in assumptions can have a significant impact on the end valuation. You don’t need to have a particularly volatile market, you just need to change one assumption in a very small way and that can, from a sensitivity perspective, have a huge impact. With volatility, on the other hand, what you end up with are a lot more exceptions, wider bid/ask spreads, potentially less in the way of market colour. Volatility tends to have an impact on liquidity as well so, if the market inputs used in the valuation process begin to dry up, that can have an impact too. One tends to get more requests for evaluations the more volatile it gets, as people want to keep on top of the market a lot more than if you had a calmer conditions.
Risk: What advances are being made in building robust pricing tools for illiquid assets and markets? What level of comfort around pricing will be required for buy-side firms to feel comfortable with less liquid assets in the future?
BS: I think it’s similar to a lot of econometrics. I think the quantitative mathematicians have been in for a battering and, in some respects, quite rightly so. There are still a lot of academics that were putting up pricing models for techniques for collateralised debt obligation models that are now re-inventing them, and still don’t really have much of an explanation about why this new model is going to necessarily be a lot better or indeed why the old techniques didn’t work. I think there’s going to be some innovation, there’s going to be a lot more pressure on more robust pricing. I think it’s a long way off, but something that is starting is within all of the pricing models – and particularly with regard to going away from normal market behaviour, where you have something that is pricing and pricing in line with a theoretical model – to maybe moving towards the peculiar state of the market where liquidity is drifting away. Firms are starting to look at what’s happening with that price – I think there’s some innovation going on in trying to involve human behaviour in the behaviour of markets to come up with better theoretical prices behind them. There was a gentleman from SunGard who, in my view, came up with a very good way of looking at it, which was with regard to a lot of the products that you’re trying to evaluate. He was always keen – having come from a loans background – on what is the second means of valuation. If you have a product that is unhedgeable, that is unreplicable – that is, you don’t see any quotes in it – what is it really worth? With the house, you have maybe the value you can sell it in the market, but there are complications. And I think, to have those opaque products where you have a second means of evaluation, what you can hang your hat on and say, “well it’s got to be worth this”, is challenging.
LD: For the most part, there are pricing applications available that clients can use for a particular instrument. For some instruments, there is going to be a time gap between when the more creative instrument comes out versus when a third-party pricing vendor is going to be evaluating it. You need three things to produce consistent evaluations: a model; regular and consistent market data; and evaluators who are subject-matter experts that can interpret the market data. So, when the more creative instruments are coming out of Wall Street, there will always be a gap until the pricing vendors are able to develop those three things – the models, consistent market data and evaluators – who can interpret that information.
Risk: Is counterparty risk being factored into valuations effectively? Certainly, it has been a very hot topic in the media lately. What improvements have recently been made in this area and what challenges do firms have yet to tackle?
LD: I think the entire industry would like to tackle counterparty risk. I don’t know if there’s a clean answer for it. I think Ian mentioned earlier needing to know what an evaluation is and then describing a valuation as a firm’s interpretation of the market data the vendor has collected at that point in time. When we are looking at a security and we are giving our opinion on what we see in the marketplace, we’re not necessarily taking into account which two firms are part of that transaction. The evaluation is based on the market information that is being understood at a given point in time. So counterparty risk is something that I can speak about only from the third-party pricing vendor’s perspective. We should continue to examine how can we apply something to the evaluation process to factor in counterparty risk. Today evaluations are really reflecting back to the industry, what market data vendors have collected and processed into the evaluation process, and not taking into consideration the counterparties involved in the transaction.
IB: It’s often difficult for a vendor to know who’s on both sides of the trade. You mention valuations as opposed to evaluated pricing and I think that’s probably the right way of positioning it. An evaluated price is typically agnostic and independent as to the counterparties in the transaction. However, if a specific firm is going to correctly value their trade they certainly need to take into account the credit risk of their counterparty. So, I think, in that respect, you could use the independent evaluated price for the trade as a baseline and then make some kind of adjustment as to the counterparty risk that you perceive. I think that an area where it is very, very relevant is in swap valuations. Where previously the swap counterparty has always been assumed to be irrelevant, I think that now there’s probably a case for adjusting valuations for counterparty risk. From a vendor perspective, that’s almost customised to the individual firm, and therefore begins to cross the line of objectivity and independence, so it is most appropriate as an internal risk management exercise as opposed to an external data or vendor exercise.
Risk: How should firms go about organising or consolidating the data and models that they use for pricing to reduce operational risk?
IB: The most obvious thing for me is, where you can eliminate a manual process, it’s clearly going to help you out on your operational risk side. So, having an automated source for valuations coming in a feed that you can build some processes around, such as exception reporting, rather than having some guy picking up the phone and asking their tame broker what the value of that deal is and then typing it into a spreadsheet, that then gets shoved into an accounting, trading or risk management application. To me that is fraught with potential error, so automation, clearly – as far as you can build a robust process around it – works.
BS: Yes, I have a few things to say on this. It’s my particular area. I think the obvious tool that is used in this area to bring data and models together is spreadsheets, and I think a lot of organisations need to decide whether or not they are going to embrace using the spreadsheet and treat it as a proper tool or whether it’s something that is to be used as a prototyping tool to try to move things into systems. It may or may not surprise you but the [UK] Financial Services Association does not advocate that people shouldn’t use spreadsheets. There are very good presentations that they give about treating spreadsheets as a corporate asset. If you’re using them in the valuation process, you can set up good automated processes. But it’s just that spreadsheets fall into this kind of no-man’s land between end-user computing by the people who understand the products and the data and the business, and the technologists who say “that’s a spreadsheet, we don’t want it here, we don’t want to support it or have anything to do with it”. So I think there’s an operational risk there from the point of view of entering data and everything else.
The other aspect in this area, and it’s good you’ve put data and models together, is a lot of the time we see data management projects concerning reference data or maybe market data, and they might be feeding in ultimately to valuations. It’s the valuation that really and truly is the most valuable data of all. So why that data isn’t within the data management process, why models and data aren’t part of the automated process, is something that I think is slowly changing as things get a little more integrated.
Risk: Are boards of directors and senior managers putting in place the necessary governance structure to sustain a culture where valuation risk is managed at all levels in an organisation?
LD: I’m seeing a good level of activity in that area where clients are engaging – not just those that were responsible for evaluations – but they are now including their compliance officers, portfolio managers, general counsel and others. So what we’re seeing is it’s becoming multi-functional. There’s more involvement than there had been traditionally and there is a deeper desire to understand more detail about the evaluation process.
BS: You’re getting valuation committees on the buy side and sell side. I think the change with valuation management, as with data management, is: what is the business motivator? And maybe previously, if everything was going swimmingly and funds were heading up and there was very little risk, the business motivation wasn’t there. On the sell side, the traders are now interested in the valuations from the point of view of seeing what capital they’re using up because capital is limited. On the buy side, the clients are getting much more demanding in terms of “how much is this worth – explain please”. And so, I think it’s the business motivator that will drive these.
Click here to view the article in PDF format