Listening and Context for Big Data

There are (occaisionally!) some good questions and conversations going on within some of the LinkedIn groups. One recently was around what use cases there are for unstructured data within banking and finance, and I found this comment from Tom Deutsch of IBM to be quite insightful and elegant (at least better than I could I have written it...) on what the main types of unstructured data analysis there are:

• Listening for the first time 
• Listening better 
• Adding context 

Listening for the first time is really just making use of what you already probably capture to hear what is being said (or navigated) 

Listening better is making sure you are actually both hearing and understanding what is being said. This is sometimes non-trivial as it involves accuracy issues and true (not marketing hype) NLP technologies and integrating multiple sources of information 

Adding context is when you either add structured data to the above or add the above to structured data, usually to round out or more fully inform models (or sometimes just build new ones). 

Big Data Finance at NYU Poly

I went over to NYU Poly in Brooklyn on Friday of last week for their Big Data Finance Conference. To get a slightly negative point out of the way early, I guess I would have to pose the question "When is a big bata conference, not a big data Conference?". Answer: "When it is a time series analysis conference" (sorry if you were expecting a funny answer...but as you can see, then what I occupy my time with professionally doesn't naturally lend itself to too much comedy). As I like time series analysis, then this was ok, but certainly wasn't fully "as advertised" in my view, but I guess other people are experiencing this problem too.

Maybe this slightly skewed agenda was due to the relative newness of the topic, the newness of the event and the temptation for time series database vendors to jump on the "Big Data" marketing bandwagon (what? I hear you say, we vendors jumping on a buzzword marketing bandwagon, never!...). Many of the talks were about statistical time series analysis of market behaviour and less about what I was hoping for, which was new ways in which empirical or data-based approaches to financial problems might be addressed through big data technologies (as an aside, here is a post on a previous PRMIA event on big data in risk management as some additional background). There were some good attempts at getting a cross-discipline fertilization of ideas going at the conference, but given the topic then representatives from the mobile and social media industries were very obviously missing in my view. 

So as a complete counterexample to the two paragraphs above, the first speaker (Kevin Atteson of Morgan Stanley) at the event was on very much on theme with the application of big data technologies to the mortgage market. Apparently Morgan Stanley had started their "big data" analysis of the mortgage market in 2008 as part of a project to assess and understand more about the potential losses than Fannie Mae and Freddie Mac faced due to the financial crisis.  

Echoing some earlier background I had heard on mortgages, one of the biggest problems in trying to understand the market according to Kevin was data, or rather the lack of it. He compared mortgage data analysis to "peeling an onion" and that going back to the time of the crisis, mortgage data at an individual loan level was either not available or of such poor quality as to be virtually useless (e.g. hard to get accurate ZIP code data for each loan). Kevin described the mortgage data set as "wide" (lots of loans with lots of fields for each loan) rather than "deep" (lots of history), with one of the main data problems was trying to match nearest-neighbour loans. He mentioned that only post crisis have Fannie and Freddie been ordered to make individual loan data available, and that there is still no readily available linkage data between individual loans and mortgage pools (some presentations from a recent PRMIA event on mortgage analytics are at the bottom of the page here for interested readers). 

Kevin said that Morgan Stanley had rejected the use of Hadoop, primarily due write through-put capabilities, which Kevin indicated was a limitating factor in many big data technologies. He indicated that for his problem type that he still believed their infrastructure to be superior to even the latest incarnations of Hadoop. He also mentioned the technique of having 2x redundancy or more on the data/jobs being processed, aimed not just at failover but also at using the whichever instance of a job that finished first. Interestingly, he also added that Morgan Stanley's infrastructure engineers have a policy of rebooting servers in the grid even during the day/use, so fault tolerance was needed for both unexpected and entirely deliberate hardware node unavailability. 

Other highlights from the day:

• Dennis Shasha had some interesting ideas on using matrix algebra for reducing down the data analysis workload needed in some problems - basically he was all for "cleverness" over simply throwing compute power at some data problems. On a humourous note (if you are not a trader?), he also suggested that some traders had "the memory of a fruit-fly".
• Robert Almgren of QuantitativeBrokers was an interesting speaker, talking about how his firm had done a lot of analytical work in trying to characterise possible market responses to information anouncements (such as Friday's non-farm payroll announcement). I think Robert was not so much trying to predict the information itself, but rather trying to predict likely market behaviour once the information is announced. 
• Scott O'Malia of the CFTC was an interesting speaker during the morning panel. He again acknowledged some of the recent problems the CFTC had experienced in terms of aggregating/analysing the data they are now receiving from the market. I thought his comment on the twitter crash was both funny and brutally pragmatic with him saying "if you want to rely solely upon a single twitter feed to trade then go ahead, knock yourself out."
• Eric Vanden Eijnden gave an interesting talk on "detecting Black Swans in Big Data". Most of the examples were from current detection/movement in oceanography, but seemed quite analogous to "regime shifts" in the statistical behaviour of markets. Main point seemed to be that these seemingly unpredicatable and infrequent events were predictable to some degree if you looked deep enough in the data, and in particular that you could detect when the system was on a possible likely "path" to a Black Swan event.

One of the most interesting talks was by Johan Walden of the Haas Business School, on the subject of "Investor Networks in the Stock Market". Johan explained how they had used big data to construct a network model of all of the participants in the Turkish stock exchange (both institutional and retail) and in particular how "interconnected" each participant was with other members. His findings seemed to support the hypothesis that the more "interconnected" the investor (at the centre of many information flows rather than add the edges) the more likely that investor would demonstrate superior return levels to the average. I guess this is a kind of classic transferral of some of the research done in social networking, but very interesting to see it applied pragmatically to financial markets, and I would guess an area where a much greater understanding of investor behaviour could be gleaned. Maybe Johan could do with a little geographic location data to add to his analysis of how information flows.

So overall a good day with some interesting talks - the statistical presentations were challenging to listen to at 4pm on a Friday afternoon but the wine afterwards compensated. I would also recommend taking a read through a paper by Charles S. Tapiero on "The Future of Financial Engineering" for one of the best discussions I have so far read about how big data has the potential to change and improve upon some of the assumptions and models that underpin modern financial theory (haven't found an online copy as yet but will update when I do). Coming back to my starting point in this post on the content of the talks, I liked the description that Charles gives of traditional "statistical" versus "data analytics" approaches, and some of the points he makes about data immediately inferring relationships without the traditional "hypothesize, measure, test and confirm-or-not" were interesting, both in favour of data analytics and in cautioning against unquestioning belief in the findings from data (feels like this post from October 2008 is a timely reminder here). With all of the hype and the hope around the benefits of big data, maybe we would all be wise to remember this quote by a certain well-known physicist: "No amount of experimentation can ever prove me right; a single experiment can prove me wrong."

 

Last day of voting in the FTF News Awards!

Today is the last day of voting in the FTF News Technology Innovation Awards 2013. Our TimeScape solution has been nominated under the Best Enterprise Data Management (EDM) category. If you can spare a minute of your time today, then Xenomorph would very much appreciate your vote!

The Anthropology, Sociology, and Epistemology of Risk

Background - I went along to my first PRMIA event in Stamford, CT last night, with the rather grandiose title of "The Anthropology, Sociology, and Epistemology of Risk". Stamford is about 30 miles north of Manhattan and is the home to major offices of a number of financial markets companies such as Thomson Reuters, RBS and UBS (who apparently have the largest column-less trading floor in the world at their Stamford headquarters - particularly useful piece of trivia for you there...). It also happens to be about 5 minutes drive/train journey away from where I now live, so easy for me to get to (thanks for another useful piece of information I hear you say...). Enough background, more on the event which was a good one with five risk managers involved in an interesting and sometimes philosophical discussion on fundamentally what "risk management" is all about.

Introduction - Marc Groz who heads the Stamford Chapter of PRMIA introduced the evening and started by thanking Barry Schwimmer for allowing PRMIA to use the Stamford Innovation Centre (the Old Town Hall) for the meeting. Henrik Neuhaus moderated the panel, and started by outlining the main elements of the event title as a framework for the discussion:

• Anthropology - risk management is to what purpose?
• Sociology - how does risk management work?
• Epistemology - what knowledge is really contained within risk management?

Henrik started by taking a passage about anthropology and replacing human "development" with "risk management" which seemed to fit ok, although the angle I was expecting was much more about human behaviour in risk management than where Henrik started. Henrik asked the panel what results they had seen from risk management and what did that imply about risk management? The panelists seemed a little confused or daunted by the question prompting one of them to ask "Is that the question?".

Business Model and Risk Culture - Elliot Noma dived in by responding that the purpose of risk management obviously depended very much on what are the institutional goals of the organization. He said that it was as much about what you are forced to do and what you try to do in risk management. Elliot said that the sell-side view of risk management was very regulatory and capital focused, whereas mutual funds are looking more at risk relative to benchmarks and performance attribution. He add that in the alternatives (hedge-fund) space then there were no benchmarks and the focus was more about liquidity and event risk.

Steve Greiner said that it was down to the investment philosophy and how risk is defined and measured. He praised some asset managers where the risk managers sit across from the portfolio managers and are very much involved in the decision making process.

Henrik asked the panel whether any of the panel had ever defined a “mission statement” for risk management. Marc Groz chipped in that he remember that he had once defined one, and that it was very different from what others in the institution were expecting and indeed very different from the risk management that he and his department subsequently undertook.

Mark Szycher (of GM Pension Fund) said that risk management split into two areas for him, the first being the symmetrical risks where you need to work out the range of scenarios for a particular trade or decision being taken. The second was the more asymmetrical risks (i.e. downside only) such as those found in operational risk where you are focused on how best to avoid them happening.

Micro Risk Done Well - Santa Federico said that he had experience of some of the major problems experienced at institutions such as Merrill Lynch, Salomen Brothers and MF Global, and that he thought risk management was much more of a cultural problem than a technical one. Santa said he thought that the industry was actually quite good at the micro (trade, portfolio) risk management level, but obviously less effective at the large systematic/economic level. Mark asked Santa what was the nature of the failures he had experienced. Santa said that the risks were well modeled, but maybe the assumptions around macro variables such as the housing market proved to be extremely poor.

Keep Dancing? - Henrik asked the panel what might be done better? Elliot made the point that some risks are just in the nature of the business. If a risk manager did not like placing a complex illiquid trade and the institution was based around trading in illiquid markets then what is a risk manager to do? He quote the Citi executive who said “ whilst the music is still playing we have to dance”. Again he came back to the point that the business model of the institution drives its cultural and the emphasis of risk management (I guess I see what Elliot was saying but taken one way it implied that regardless of what was going on risk management needs to fit in with it, whereas I am sure that he meant that risk managers must fit in with the business model mandated to shareholders).

Risk Attitudes in the USA - Mark said that risk managers need to recognize that the improbable is maybe not so improbable and should be more prepared for the worst rather than risk management under “normal” market and institutional behavior. Steven thought that a cultural shift was happening, where not losing money was becoming as important to an organization as gaining money. He said that in his view, Europe and Asia had a stronger risk culture than in the United States, with much more consensus, involvement and even control over the trading decisions taken. Put another way, the USA has more of a culture of risk taking than Europe. (I have my own theories on this. Firstly I think that the people are generally much more risk takers in the USA than in UK/Europe, possibly influenced in part by the relative lack of underlying social safety net – whilst this is not for everyone, I think it produces a very dynamic economy as a result. Secondly, I do not think that cultural desire in the USA for the much admired “presidential” leader necessarily is the best environment for sound, consensus based risk management. I would also like to acknowledge that neither of my two points above seem to have protected Europe much from the worst of the financial crisis, so it is obviously a complex issue!).

Slaves to Data? - Henrik asked whether the panel thought that risk managers were slaves to data? He expanded upon this by asking what kinds of firms encourage qualitative risk management and not just risk management based on Excel spreadsheets? Santa said that this kind of qualitative risk management occurred at a business level and less so at a firm wide level. In particular he thought this kind of culture was in place at many hedge funds, and less so at banks. He cited one example from his banking career in the 1980's, where his immediate boss was shouted off the trading floor by the head of desk, saying that he should never enter the trading floor again (oh those were the days...). 

Sociology and Credibility - Henrik took a passage on the historic development of women's rights and replaced the word "women" with "risk management" to illustrate the challenges risk management is facing with trying to get more say and involvement at financial institutions. He asked who should the CRO report to? A CEO? A CIO? Or a board member? Elliot responded by saying this was really a issue around credibility with the business for risk managers and risk management in general. He made the point that often Excel and numbers were used to establish credibility with the business. Elliot added that risk managers with trading experience obviously had more credibility, and to some extent where the CRO reported to was dependent upon the credibility of risk management with the business. 

Trading and Risk Management Mindsets - Elliot expanded on his previous point by saying that the risk management mindset thinks more in terms of unconditional distributions and tries to learn from history. He contrasted this with a the "conditional mindset' of a trader, where the time horizon forwards (and backwards) is rarely longer than a few days and the belief is strong that a trade will work today given it worked yesterday is high. Elliot added that in assisting the trader, the biggest contribution risk managers can make is more to be challenging/helpful on the qualitative side rather than just quantitative.

Compensation and Transactions - Most of the panel seemed to agree that compensation package structure was a huge influencer in the risk culture of an organisation. Mark touched upon a pet topic of mine, which is that it very hard for a risk manager to gain credibility (and compensation) when what risk management is about is what could happen as opposed to what did happen. A risk manager blocking a trade due to some potentially very damaging outcomes will not gain any credibility with the business if the trading outcome for the suggested trade just happened to come out positive. There seemed to be concensus here that some of the traditional compensation models that were based on short-term transactional frequency and size were ill-formed (given the limited downside for the individual), and whilst the panel reserved judgement on the effectiveness of recent regulation moves towards longer-term compensation were to be welcome from a risk perspective.

MF Global and Busines Models - Santa described some of his experiences at MF Global, where Corzine moved what was essentially a broker into taking positions in European Sovereign Bonds. Santa said that the risk management culture and capabilities were not present to be robust against senior management for such a business model move. Elliot mentioned that he had been courted for trades by MF Global and had been concerned that they did not offer electronic execution and told him that doing trades through a human was always best. Mark said that in the area of pension fund management there was much greater fidiciary responsibility (i.e. behave badly and you will go to jail) and maybe that kind of responsibility had more of a place in financial markets too. Coming back to the question of who a CRO should report to, Mark also said that questions should be asked to seek out those who are 1) less likely to suffer from the "agency" problem of conflicts of interest and on a related note those who are 2) less likely to have personal biases towards particular behaviours or decisions.

Santa said that in his opinion hedge funds in general had a better culture where risk management opinions were heard and advice taken. Mark said that risk managers who could get the business to accept moral persuasion were in a much stronger position to add value to the business rather than simply being able to "block" particular trades. Elliot cited one experience he had where the traders under his watch noticed that a particular type of trade (basis trades) did not increase their reported risk levels, and so became more focussed on gaming the risk controls to achieve high returns without (reported) risk. The panel seemed to be in general agreement that risk managers with trading experience were more credible with the business but also more aware of the trader mindset and behaviors. 

Do we know what we know? - Henrik moved to his third and final subsection of the evening, asking the panel whether risk managers really know what they think they know. Elliot said that traders and risk managers speak a different language, with traders living in the now, thinking only of the implications of possible events such as those we have seen with Cyprus or the fiscal cliff, where the risk management view was much less conditioned and more historical. Steven re-emphasised the earlier point that risk management at this micro trading level was fine but this was not what caused events such as the collapse of MF Global.

Rational argument isn't communication - Santa said that most risk managers come from a quant (physics, maths, engineering) background and like structured arguments based upon well understood rational foundations. He said that this way of thinking was alien to many traders and as such it was a communication challenge for risk managers to explain things in a way that traders would actually put some time to considering. On the modelling side of things, Santa said that sometimes traders dismissed models as being "too quant" and sometimes traders followed models all too blindly without questioning or understanding the simplifying assumptions they are based on. Santa summarised by saying that risk management needs to intuitive for traders and not just academically based. Mark added that a quantitative focus can sometimes become too narrow (modeler's manifesto anyone?) and made the very profound point that unfortunately precision often wins over relevance in the creation and use of many models. Steven added that traders often deal with absolutes, so as knowing the spread between two bonds to the nearest basis point, whereas a risk manager approaching them with a VaR number really means that this is the estimated VaR which really should be thought to be within a range of values. This is alien to the way traders think and hence harder to explain.

Unanticipated Risk - An audience member asked whether risk management should focus mainly on unanticipated risks rather than "normal' risks. Elliot said that in his trading he was always thinking and checking whether the markets were changing or continuing with their recent near-term behaviour patterns. Steven said that history was useful to risk management when markets were "normal", but in times of regime shifts this was not the case and cited the example of the change in markets when Mario Dragi announced that the ECB would stand behind the Euro and its member nations. 

Risky Achievements - Henrik closed the panel by asking each member what they thought was there own greatest achievement in risk management. Elliot cited a time when he identified that a particular hedge fund had a relatively inconspicuous position/trade that he identified as potentially extremely dangerous and was proved correct when the fund closed down due to this. Steven said he was proud of some good work he and his team did on stress testing involving Greek bonds and Eurozone. Santa said that some of the work he had done on portfolio "risk overlays" was good. Mark ended the panel by saying that he thought his biggest achievement was when the traders and portfolio managers started to come to the risk management department to ask opinions before placing key trades. Henrik and the audience thanked the panel for their input and time.

An Insured View - After the panel closed I spoke with an actuary who said that he had greatly enjoyed the panel discussions but was surprised that when talking of how best to support the risk management function in being independent and giving "bad" news to the business, the role of auditors were not mentioned. He said he felt that auditors were a key support to insurers in ensuring any issues were allowed to come to light. So food for thought there as to whether financial markets can learn from other industry sectors.

Summary - great evening of discussion, only downside being the absence of wine once the panel had closed!

 

Xenomorph short-listed for the FTF News Awards 2013

Xenomorph has been short-listed in the FTF News Technology Innovation Awards 2013, under the Best Enterprise Data Management (EDM) category. The short-list includes:

BEST ENTERPRISE DATA MANAGEMENT SOLUTION

• Eagle Investment Systems – Eagle Data Management
• GoldenSource – GoldenSource EDM
• Markit  - Markit EDM
• SmartStream/Euroclear Partnership
• Xenomorph - TimeScape

If you can spare a few moments of your time, then Xenomorph would very much appreciate your vote!

Feel democracy in action with Xenomorph (and vote for us!)

In these times of disfunctional political parties, external troikas and technocratic governments, why not exercise your democratic rights in a positive way by voting for Xenomorph in the best Enterprise Data Management category of the Inside Reference Data Awards 2013. You know it makes democratic sense.

 

Spreadsheet control and contagion

Just caught saw a reference on LinkedIn to this FT article "Finance groups lack spreadsheet controls". Started to write a quick response and given it is one of my major hobby-horses, I ended up doing a bit of an essay, so I decided to post it here too:

"As many people have pointed out elsewhere, much of the problem with spreadsheet usage is that they are not treated as a corporate and IT asset, and as such things like testing, peer review and general QA are not applied (mind you, maybe more of that should still be applied to many mainstream software systems in financial markets...). 

Ralph and the guys at Cluster Seven do a great job in helping institutions to manage and monitor spreadsheet usage (I like Ralph's "we are CCTV for spreadsheets" analogy), but I think a fundamental (and often overlooked) consideration is to ask yourself why did the business users involved decide that they needed spreadsheets to manage trading and risk in the first place? It is a bit like trying to address the symptoms of a illness without ever considering how we got the illness in the first place. 

Excel is a great tool, but to quote Spider-Man "with great power comes great responsibility" and I guess we can all see the consequences of not taking the usage of spreadsheets seriously and responsibly. So next time the trader or risk manager says "we've just built this really great model in Excel" ask them why they built it in Excel, and why they didn't build upon the existing corporate IT solutions and tools. In these cost- and risk- conscious times, I think the answers would be interesting..."

 

Data Management for Risk at Mediobanca

Very pleased to announce today that Mediobanca, the leading investment bank in Italy, has decided to select TimeScape as its data management system. You can see the press release here.

Analytics Strategy from Numerix

Good post from Jim Jockle over at Numerix - main theme is around having an "analytics" strategy in place in addition to (and probably as part of) a "Big Data" strategy. Fits strongly around Xenomorph's ideas on having both data management and analytics management in place (a few posts on this in the past, try this one from a few years back) - analytics generate the most valuable data of all, yet the data generated by analytics and the input data that supports analytics is largely ignored as being too business focussed for many data management vendors to deal with, and too low level for many of the risk management system vendors to deal with. Into this gap in functionality falls the risk manager (supported by many spreadsheets!), who has to spend too much time organizing and validating data, and too little time on risk management itself.

Within risk management, I think it comes down to having the appropriate technical layers in place of data management, analytics/pricing management and risk model management. Ok it is a greatly simplified representation of the architecture needed (apologies to any techies reading this), but the majority of financial institutions do not have these distinct layers in place, with each of these layers providing easy "business user" access to allow risk managers to get to the "detail" of the data when regulators, auditors and clients demand it. Regulators are finally waking up to the data issue (see Basel on data aggregation for instance) but more work is needed to pull analytics into the technical architecture/strategy conversation, and not just confine regulatory discussions of pricing analytics to model risk. 

Big Data – What is its Value to Risk Management?

A little late on these notes from this PRMIA Event on Big Data in Risk Management that I helped to organize last month at the Harmonie Club in New York. Big thank you to my PRMIA colleagues for taking the notes and for helping me pull this write-up together, plus thanks to Microsoft and all who helped out on the night.

Introduction: Navin Sharma (of Western Asset Management and Co-Regional Director of PRMIA NYC) introduced the event and began by thanking Microsoft for its support in sponsoring the evening. Navin outlined how he thought the advent of “Big Data” technologies was very exciting for risk management, opening up opportunities to address risk and regulatory problems that previously might have been considered out of reach.

Navin defined Big Data as the structured or unstructured in receive at high volumes and requiring very large data storage. Its characteristics include a high velocity of record creation, extreme volumes, a wide variety of data formats, variable latencies, and complexity of data types. Additionally, he noted that relative to other industries, in the past financial services has created perhaps the largest historical sets of data and continually creates enormous amount of data on a daily or moment-by-moment basis. Examples include options data, high frequency trading, and unstructured data such as via social media.  Its usage provides potential competitive advantages in a trading and investment management. Also, by using Big Data it is possible to have faster and more accurate recognition of potential risks via seemingly disparate data - leading to timelier and more complete risk management of investments and firms’ assets. Finally, the use of Big Data technologies is in part being driven by regulatory pressures from Dodd-Frank, Basel III, Solvency II, Markets for Financial Instruments Directives (1 & 2) as well as Markets for Financial Instruments Regulation.

Navin also noted that we will seek to answer questions such as:

• What is the impact of big data on asset management?
• How can Big Data’s impact enhance risk management?
• How is big data used to enhance operational risk?

Presentation 1: Big Data: What Is It and Where Did It Come From?: The first presentation was given by Michael Di Stefano (of Blinksis Technologies), and was titled “Big Data. What is it and where did it come from?”.  You can find a copy of Michael’s presentation here. In summary Michael started with saying that there are many definitions of Big Data, mainly defined as technology that deals with data problems that are either too large, too fast or too complex for conventional database technology. Michael briefly touched upon the many different technologies within Big Data such as Hadoop, MapReduce and databases such as Cassandra and MongoDB etc. He described some of the origins of Big Data technology in internet search, social networks and other fields. Michael described the “4 V’s” of Big Data: Volume, Velocity, Variety and a key point from Michael was “time to Value” in terms of what you are using Big Data for. Michael concluded his talk with some business examples around use of sentiment analysis in financial markets and the application of Big Data to real-time trading surveillance.

Presentation 2: Big Data Strategies for Risk Management: The second presentation “Big Data Strategies for Risk Management” was introduced by Colleen Healy of Microsoft (presentation here). Colleen started by saying expectations of risk management are rising, and that prior to 2008 not many institutions had a good handle on the risks they were taking. Risk analysis needs to be done across multiple asset types, more frequently and at ever greater granularity. Pressure is coming from everywhere including company boards, regulators, shareholders, customers, counterparties and society in general. Colleen used to head investor relations at Microsoft and put forward a number of points:

• A long line of sight of one risk factor does not mean that we have a line of sight on other risks around.
• Good risk management should be based on simple questions.
• Reliance on 3^rd parties for understanding risk should be minimized.
• Understand not just the asset, but also at the correlated asset level.
• The world is full of fast markets driving even more need for risk control
• Intraday and real-time risk now becoming necessary for line of sight and dealing with the regulators
• Now need to look at risk management at a most granular level.

Colleen explained some of the reasons why good risk management remains a work in progress, and that data is a key foundation for better risk management. However data has been hard to access, analyze, visualize and understand, and used this to link to the next part of the presentation by Denny Yu of Numerix.

Denny explained that new regulations involving measures such as Potential Future Exposure (PFE) and Credit Value Adjustment (CVA) were moving the number of calculations needed in risk management to a level well above that required by methodologies such as Value at Risk (VaR). Denny illustrated how the a typical VaR calculation on a reasonable sized portfolio might need 2,500,000 instrument valuations and how PFE might require as many as 2,000,000,000. He then explain more of the architecture he would see as optimal for such a process and illustrated some of the analysis he had done using Excel spreadsheets linked to Microsoft’s high performance computing technology.

Presentation 3: Big Data in Practice: Unintentional Portfolio Risk: Kevin Chen of Opera Solutions gave the third presentation, titled “Unintentional Risk via Large-Scale Risk Clustering”. You can find a copy of the presentation here. In summary, the presentation was quite visual and illustrating how large-scale empirical analysis of portfolio data could produce some interesting insights into portfolio risk and how risks become “clustered”. In many ways the analysis was reminiscent of an empirical form of principal component analysis i.e. where you can see and understand more about your portfolio’s risk without actually being able to relate the main factors directly to any traditional factor analysis. 

Panel Discussion: Brian Sentance of Xenomorph and the PRMIA NYC Steering Committee then moderated a panel discussion. The first question was directed at Michael “Is the relational database dead?” – Michael replied that in his view relational databases were not dead and indeed for dealing with problems well-suited to relational representation were still and would continue to be very good. Michael said that NoSQL/Big Data technologies were complimentary to relational databases, dealing with new types of data and new sizes of problem that relational databases are not well designed for. Brian asked Michael whether the advent of these new database technologies would drive the relational database vendors to extend the capabilities and performance of their offerings? Michael replied that he thought this was highly likely but only time would tell whether this approach will be successful given the innovation in the market at the moment. Colleen Healy added that the advent of Big Data did not mean the throwing out of established technology, but rather an integration of established technology with the new such as with Microsoft SQL Server working with the Hadoop framework.

Brian asked the panel whether they thought visualization would make a big impact within Big Data? Ken Akoundi said that the front end applications used to make the data/analysis more useful will evolve very quickly. Brian asked whether this would be reminiscent of the days when VaR first appeared, when a single number arguably became a false proxy for risk measurement and management? Ken replied that the size of the data problem had increased massively from when VaR was first used in 1994, and that visualization and other automated techniques were very much needed if the headache of capturing, cleansing and understanding data was to be addressed.

Brian asked whether Big Data would address the data integration issue of siloed trading systems? Colleen replied that Big Data needs to work across all the silos found in many financial organizations, or it isn’t “Big Data”. There was general consensus from the panel that legacy systems and people politics were also behind some of the issues found in addressing the data silo issue.

Brian asked if the panel thought the skills needed in risk management would change due to Big Data? Colleen replied that effective Big Data solutions require all kinds of people, with skills across a broad range of specific disciplines such as visualization. Generally the panel thought that data and data analysis would play an increasingly important part for risk management. Ken put forward his view all Big Data problems should start with a business problem, with not just a technology focus. For example are there any better ways to predict stock market movements based on the consumption of larger and more diverse sources of information. In terms of risk management skills, Denny said that risk management of 15 years ago was based on relatively simply econometrics. Fast forward to today, and risk calculations such as CVA are statistically and computationally very heavy, and trading is increasingly automated across all asset classes. As a result, Denny suggested that even the PRMIA PRM syllabus should change to focus more on data and data technology given the importance of data to risk management.

Asked how best to should Big Data be applied?, then Denny replied that echoed Ken in saying that understanding the business problem first was vital, but that obviously Big Data opened up the capability to aggregate and work with larger datasets than ever before. Brian then asked what advice would the panel give to risk managers faced with an IT department about to embark upon using Big Data technologies? Assuming that the business problem is well understood, then Michael said that the business needed some familiarity with the broad concepts of Big Data, what it can and cannot do and how it fits with more mainstream technologies. Colleen said that there are some problems that only Big Data can solve, so understanding the technical need is a first checkpoint. Obviously IT people like working with new technologies and this needs to be monitored, but so long as the business problem is defined and valid for Big Data, people should be encouraged to learn new technologies and new skills. Kevin also took a very positive view that IT departments should  be encouraged to experiment with these new technologies and understand what is possible, but that projects should have well-defined assessment/cut-off points as with any good project management to decide if the project is progressing well. Ken put forward that many IT staff were new to the scale of the problems being addressed with Big Data, and that his own company Opera Solutions had an advantage in its deep expertise of large-scale data integration to deliver quicker on project timelines.

Audience Questions: There then followed a number of audience questions. The first few related to other ideas/kinds of problems that could be analyzed using the kind of modeling that Opera had demonstrated. Ken said that there were obvious extensions that Opera had not got around to doing just yet. One audience member asked how well could all the Big Data analysis be aggregated/presented to make it understandable and usable to humans? Denny suggested that it was vital that such analysis was made accessible to the user, and there general consensus across the panel that man vs. machine was an interesting issue to develop in considering what is possible with Big Data. The next audience question was around whether all of this data analysis was affordable from a practical point of view. Brian pointed out that there was a lot of waste in current practices in the industry, with wasteful duplication of ticker plants and other data types across many financial institutions, large and small. This duplication is driven primarily by the perceived need to implement each institution’s proprietary analysis techniques, and that this kind of customization was not yet available from the major data vendors, but will become more possible as cloud technology such as Microsoft’s Azure develops further. There was a lot of audience interest in whether Big Data could lead to better understanding of causal relationships in markets rather than simply correlations. The panel responded that causal relationships were harder to understand, particularly in a dynamic market with dynamic relationships, but that insight into correlation was at the very least useful and could lead to better understanding of the drivers as more datasets are analyzed.

 

Rutgers Quantitative Finance Summit

I got my first tour around the NYSE trading floor on Wednesday night, courtesy of an event by Rutgers University on Risk. Good event, mainly around panel discussion moderated by Nicholar Dunbar (Editor of Bloomberg Risk newsletter), and involving David Belmont (Commonfund CRO), Adam Litke (Chief Risk Strategist for Bloomberg), Hilmar Schaumann (Fortress Investment CRO) and Sanjay Sharma (CRO of Global Arbitrage and Trading at RBC).

Nick first asked the panel how do you define and measure risk? Hilmar responded that risk measurement is based around two main activities: 1) understanding how a book/portfolio is positioned (the static view) and 2) understanding sensitivities to risks that impact P&L (the dynamic view). Hilmar mentioned the use of historical data as a guide to current risks that are difficult to measure, but emphasised the need for a qualitative approach when looking at the risks being taken.

David said that he looks at both risk and uncertainty - with risk being defined as those impacts you can measure/estimate. He said that historical analysis was useful but limited given it is based only on what has happened. He thought that scenario analysis was a stronger tool. (I guess with historical analysis you at least get some idea of the impact of things that could not be predicted even it is based on one "simulation" path i.e. reality, whereas you have more flexibility with scenario management to cover all bases, but I guess limited to those bases you can imagine). David said that path-dependent risks such as those in the credit markets in the last crisis were some of the most difficult to deal with.

Adam said that you need to understand why you are measuring risk and understand what risks you are prepared to take. He said that at Wachovia they knew that a 25% house price fall in California would be a near death experience for the bank prior to the 2008 crisis, and in the event the losses were much greater than 25%. His point was really that you must decide what risks you want to survice and at what level. He said that sound common-sense judgement is needed to decide whether a scenario is really-real or not.

Sanjay said that risk managers need to maintain a lot of humility and not to over-trust risk meaurements. He described a little of the risk approach used at RBC where he said they use over 80 different models and employ them as layers/different views on risk to be brought together. He said they start with VaR as a base analysis, but build on this with scenarios, greeks and then on to other more specific reports and analysis. He emphasised that communication is a vital skill for risk managers to get their views and ideas across.

Nicholas then moved on to ask how risk managers should make or reduce risks? - getting away from risk measurement to risk management. Adam said that risks should be delegated out to those that manage them but this needs to be combined with responsibility for the risks too. Keep people and departments within the bounds of what their remit. Be prepared to talk a different business language to different stakeholders dependent upon their understanding and their motivations. David gave some examples of this in his case, where endowment funds what risk premiums over many years and risks are translated/quantified into practical things for example such as a new college building not going ahead etc. 

Hilmar said the hedge funds are supposed to take risks, and that the key was not necessarily to avoid losses (although avoid them if you can) but rather to avoid surprises. Like the other speakers, Hilmar emphasised that communication of risks to key stakeholders was vital. He also added the key point that if you don't like a risk you have identified, then try first to take it off rather than hedging it, since hedging could potentially add basis risk and simple more complication.

Nicholas then Sanjay about how risk managers should deal with bringing difficult news to the business? Sanjay suggested that any bad news should be approach in the form of "actionable transparency" i.e. that not only do you say communicate how bad the risk is to all stakeholders but you come along with actionable approaches to dealing with the risk. In all of his experience and despite the crisis, Sanjay's experience is that traders do not want to loose money and if you come with solutions they will listen. He concluded by saying that qualitative analysis should also be used, citing the hypothetical example that you should take notice of dogs (yes, the animal!) buying mortgages, whether or not the mortgages are AAA rated.

Nicholas asked the panel members in turn what risks are they concerned about currently? David said he believed that many risks were not priced into the market currently. He was concerned about policy impacts of action by the ECB and the Fed, and thought the current and forward levels of volatility are low. In Fixed Income markets he thought that Dodd-Frank may have detrimental effects, particular with the current lack of clarity about what is proprietary trading and what is market-making. He thought that should policies and interests rates change, he thought that risk managers should look carefully at what will happen as funds flow out of fixed income and into equities.

Hilmar talked about the postponement of the US debt ceiling limits and that US Government policy battles continue to be an obvious source of risk. In Europe, many countries had elections this year which would be interesting, and that the problems in the Euro-zone are less than they were, but problems in Cyprus could fan the flames of more problems and anxiety. Hilmar said the Japan's new policy of targetting 2% inflation may have effects on the willingness of domestic investors to buy JGBs. 

Sanjay said he was worried. In the "Greenspan Years" prior to 2008 a quasi government guarantee on the banks was effectively put in place and that we continue to live with cheap money. When policy eventually changes and interest rates rise, Sanjay wondered whether the world was ready for the wholesale asset revaluation that would then be required.

Adams concerns where mainly around identifying what will be the cause of the next panic in the market. Whilst he said he is in favour of central clearing for OTC derivatives, he thought that the changing market structure combined with implementing central clearing had not been fully thought through and this was a worry to him. 

Nicholas asked what do the panelists think to the regulation being implemented? David said that regulators face the same difficulty that risk managers face, in that nobody notices when you took sensible action to protect against a risk that didn't occur. He thinks that regulation of the markets is justified and necessary.

Sanjay said that in the airline and pharmacutical industries regulatory approval was on the whole very robust but that they were dealing with approving designs (aeroplanes and drugs) that are reproduced once approved. He said that such levels of regulation in financial services were not yet possible due to the constant innovation found in the markets, and he wanted regulation to be more dynamic and responsive to market developments. Sanjay also joined those in the industry that are critical of the shear size of Dodd-Frank.

Nicholas said that Adam was obviously keen on operational issues and wondered what plumbing in the industry would he change? Adam said that he is a big fan of automation but operational risk are real and large. He thought that there were too many rules and regulations being applied, and the regulators were not paying attention to the type of markets they want in the future, nor on the effects of current regulation and how people were moving from one part of the industry to another. Adam said that in relation to Knight Capital he was still a strong advocate of standing by the wall socket, ready to pull the plug on the computer. Adam suggested that regulators should look at regulating/approving software releases (I assume here he means for key tasks such as automated trading or risk reporting, not all software).

Given the large number of students present, Nicholas closed the panel by asking what career advice the panelists had for future risk managers? Adam emphasised flexibility in role, taking us through his career background as an equity derivatives and then fixed income trader before coming into risk management. Adam said it was highly unlikely over your career that you would stay with one role or area of expertise. 

Hilmar said that having risk managers independent of trading was vitally important for the industry. He thought there were many areas to work with operational risk being potentially the largest, but still with plenty more to do in market risk, compliance and risk modelling. He added that understanding the interdepencies between risks was key and an area for further development.

When asked by Nicholas, David said that risk managers should have a career path right through to CEO of an institution. He wanted to encourage risk management as a necessary level above risk measurement and control. He was excited about the potential of Big Data technologies to help in risk management. David gave some interesting background on his own career initially as an emergining markets debt trader. He said that it is important to know yourself, and that he regarded himself as a sceptic, needing all the information available before making a decision. As such his performance as a trader was consistent but not as high as some, and this became one of the reasons he moved into risk management. 

Sanjay said many of the systems used in finance are 20 years old, in complete contrast with the advancies in mobile and internet technologies. As such he thought this was a great opportunity to be involved in the replacement and upgrading of this older infrastructure. Apparently one analyst had estimated that $65B will be spent on risk management over the next 4-5 years.

Adam thought that there was a need for code of ethics for quants (see old post for some ideas). Sanjay added that the industry needed to move away from being involved primarily in attempting to optimise activity around gaming regulation. When asked by Nicholas about Basel III, Adam thought that improved regulation was necessary but Basel III was not the right way to go about it and was way too complex.

 

 

 

Chartis Research - Data Management for Risk White Paper

New whitepaper on data management for risk from the analysts Chartis Research, including a section on how Xenomorph's TimeScape solution addresses some of the key issues identified.

Not only SQL but too much choice as well?

In relation to the Microsoft/PRMIA event that Brian moderated at last night in New York, I spotted this article recently that tries to map out all the different databases that are now commercially available in some form, from SQL to No SQL and all the various incarnations and flavours in between:

http://gigaom.com/2012/12/20/confused-by-the-glut-of-new-databases-heres-a-map-for-you/

As Brian suggested in his recent post, It's amazing to see how much the landscape has evolved from the domination (mantra?) that there was the relational way, or no way. Obviously times have moved on (er, I guess the Internet happened for one thing...) and people are now far more accepting of the need for different approaches to different types and sizes of business problems. That said, I agree with the article and comments that suggest there do seem to be far too many options available now - there has to be some consolidation coming otherwise it will become increasingly difficult to know where to start. Choice is a wonderful thing, but only in moderation!

 

Alpha, Regulation and Party

Quick thank you to all those who came along to Xenomorph's New York Holiday Party at the Classic Car Club. Below is an extract from talk given by Paul Rowady of the Tabb at the event, followed by my effort and some photographs from the event. 

 ************************************************************

There Is No Such Thing as Alpha Generation

The change in perspective caused by a subtle change in language can galvanize your approach to data, the tools you select, and even the organizational culture. That said, ‘alpha generation’ is a myth; there is only alpha discovery and capture.

By E. Paul Rowady, Jr.

We live in an age of superlatives: unprecedented market complexity and uncertainty caused, in part, by an unprecedented regulatory onslaught and unprecedented economic extremes. As a result, there is an unprecedented focus on risk analysis – and an unprecedented (and anxious) search for new sources of performance from all market demographics.

The big data era is here and will only become the bigger data era. What we need is a new perspective. But fostering such a new perspective may be as subtle as performing a little linguistic jujitsu.

Our business – trading and investment in capital and commodity markets around the globe – has a history of being cavalier or too casual about language; particularly how certain labels, terms or vernacular are used to describe the business and the markets. Some of this language is intentional – the use of certain terminology creates mystique, fosters mythology, manufactures a sense of complexity that only a select group of savants can tame -- particularly when it comes to activities around quantitative methods. And some of it is just plain laziness, stretching the use of labels far beyond their original meaning on the idea that these terms are close enough.

I have become increasingly sensitive to this phenomenon over the years. Call it an insatiable need to simplify complexity, bring order to chaos, to enhance a level of accuracy and precision in how we describe what we do and how we do it. I find that precision of language does impact how complex technical topics are communicated, understood and absorbed. It turns out, language impacts perspective – and perspective impacts strategy and tactics.

So let’s gain a little perspective on alpha generation and alpha creation...(full extract can be found on the TabbFORUM)

Paul in full speech mode at the Classic Car Club

  ***********************************************************

Big thanks to Paul for the above talk. Here's is my follow-up:

  ************************************************************

Thanks Paul for a great talk, certainly I agree that people, process, technology and data are key to the future success of financial markets. In particular, I think attitudes towards data must change if we are to meet the coming challenges over the next few years. For example, in my view data in financial markets is analogous to water:

• Everyone needs it
• Everyone knows where to get it
• Nobody likes to share it
• Nobody is 100% sure where was really sourced from
• Nobody is quite sure where it goes to
• Nobody knows its true cost
• Nobody knows how much is wasted
• Everyone assumes it is of high quality
• And you only ever know it has gone bad after you have drunk it.
• (I should add, that if you own water you are also very wealthy, so wealthy your neighbor might even consider robbing you)

The problem of siloed data and data integration remains, but this is as much a political as opposed to purely technical problem. People need to share data more, and I wonder (I hope) that as the “social network” generation come through that attitudes will improve, but I guess this will also add different pressures to data aggregators as people are less hung up about sharing information. The focus needs to be on the data that business folks need, and should be less about the type of the data or the technical means by which it is captured, stored and distributed – for sure these are important aspects, but we need involve more people in realizing this cult of data. 

And just as Paul has issues with the over-use of “Alpha”, I promise this will be the only time this evening I will mention “Big Data” but today I heard the best description so far of what big data is all about, which is  “Big data is like watching the planet develop a nervous system”. Data is fundamental to all of our lives and we are living through some very interesting times in terms of how much data is becoming available and how we make sense of it.

So, a change of tack. When moving to the New York area a few years back, one of my fellow Brits said that you will find the Americans a lot friendlier than the English, but don’t talk to them about politics or religion. So rules are meant to broken, and religion aside I thought I would briefly have to mention the recent election as one of the big differences between the UK and the USA.

Firstly, wow you guys know how to have long elections. I think the French get theirs done in two weeks but even the Brits do it in a month. A few things struck me from the election: I don’t know whether the Democratic Party is generally supportive of legalizing drugs, but I think we can be certain that President Obama spent some time in the states of Colorado and Washington prior to the first debate. 

 And I hear from the New Yorker that the Republicans are trying a radical new approach to broaden the demographic of the supporter base, apparently to make it inclusive of people who have strong believers in “maths and science”. 

Moving on from a light-hearted look at elections but sticking with the government theme, the regulation is obviously very high profile at the moment. To some degree this is understandable as financial markets have been doing a great job of keeping a low profile with:

• JPMorgan $7B London Whale
• Barclays and the Libor rigging
• Standard Chartered and Iranian money laundering
• Knight Capital with the biggest advertisement in history for automated trading
• ING feeling it was missing out on things with Cuba and Iranian money
• HSBC helping Mexican drug lords to move the money around
• Capital One deceiving its customers
• Peregrine Financial Group deceiving the regulators (generating alpha?)

 All these occurred in 2012, when it seems that the dust had barely settled over MF Global and UBS. So it is possible to understand the reaction of people and politicians to what has gone on and the need for more stable capital markets, but my biggest concern is that there is simply too much regulation, and complex systems with complex rules is a great breeding ground for the law of unintended consequences. To illustrate how over time we humans, and in particular governments, seem to be regressing in terms of using more words to describe ever more complex behaviours I found the following list online:

• Pythagoras 24 words
• Lords Prayer 66 words
• Archidmedies Priciple 66 words
• 10 commandments 179 words
• Gettysburg Address 286 words
• Declaration of independence 1300 words
• US Govt sale of cabbage 26,991 words

Dodd-Frank is about 2,300 pages, which apparently is going to spawn some 30,000 pages of rules – that is enormous. Listening to a regulator speak last week, he said the regulators had about 10,000 pages done, 10,000 in progress and 10,000 not even started yet. Worse than this, he added that regulators were not trying to shape the financial markets of the future but rather dealing only with the current issues. Regulators should take their lead from quantum physics in my view, as soon as you observe something it is changed. Financial markets are complex, and making them even more complex through overlaying complex rules is not going to result in the stability that we all desire. 

Anyway, thanks for coming along this evening and I hope you have a great time. Quick thank you to our clients and partners without whom we would not exist. Thanks to the hard work our staff put in over the year, but in particular thanks to Naj and Xenomorph's NYC team for organizing this evenings event.

   ************************************************************

 Some photographs from the event below. Big thanks to NandoVision for some of the images:

Clients, partners and staff catch up over a drink or three

 

Ted Pendleton of Numerix and Paul Rowady of Tabb Group earlier on in the evening

 

This waiter had a pleasant interuption in service prior to the fashion show by Hiliary Flowers

 

Jim Beck talks with PRMIA NYC members: Qi Fu, Sol Steinberg and Don Wesnofske

 

Cass Almendral, Hillary Flowers and Brian later at the bar

 

Not sure how this ballet-themed dress works in a convertible?

 

Russ Glisker and Mark O'Donnell talk cars with Paul

 

A far more practical outfit for this Porsche

 

Some of the fashion models rush to discuss the finer points of Alpha Harvesting with Paul...

  ***********************************************************

Thanks again to all involved in putting the party together and for everyone who came along on the night. If I don't get round to another post over the Holiday Season, then best wishes for a fantastic break and a great start to 2013.

Numerix and Xenomorph partner in risk management

Just a quick post to highlight Xenomorph's Numerix partnership announcement that went out earlier this week. In summary we have done some great work with Numerix on combining their ability to price and risk manage very complex trades with TimeScape's ability to manage all the data such types of instruments need.

The integration is a great demonstration of the flexibility of TimeScape's data model (see recent post and LinkedIn discussion) and addresses some of the issues discussed and illustrated in an earlier post on data management for risk. Quick thank you to the clients involved in testing and using the integration, to the Numerix team for their assistance on this and to my New York colleagues who led the TimeScape integration work.

Big Data - Integration still a key (external) issue

Good breakfast event from SAP and A-Team last Thursday morning. SAP have been getting (and I guess paying for) a lot of good air-time for their SAP Hana in-memory database technology of late. Domenic Iannaccone of SAP started the briefing with an introduction to big data in finance and how their SAP/Sybase offerings knitted together. He started his presentation with a few quotes, one being "Intellectual property is the oil of the 21st century" by Mark Getty (he of Getty images, but also of the Getty oil family) and "Data is the new oil" by both Clive Humby and Gerd Leonhard (not sure why two people quoted saying the same thing but anyway).

For those of you with some familiarity with the Sybase IQ architecture of a year or two back, then in this architecture SAP Hana seems to have replaced the in-memory ASE database that worked in tandem with Sybase IQ for historical storage (I am yet to confirm this, but hope to find out more in the new year). When challenged on how Hana differs from other in-memory database products, Domenic seemed keen to emphasise its analytical capabilities and not just the database aspects. I guess it was the big data angle of bring the "data closer to the calculations" was his main differentiator on this, but with more time I think a little bit more explanation would have been good.

Pete Harris of the A-Team walked us through some of the key findings of what I think is the best survey I have read so far on the usage of big data in financial markets (free sign-up needed I think, but you can get a copy of the report here). Some key findings from a survey of staff at ten major financial institutions included:

• Searching for meaning in instructured data was a leading use-case thought of when thinking of big data (Twitter trading etc)
• Risk management was seen as a key beneficiary of what the technologies can offer
• Aggregation of data for risk was seen as a key application area concerning structured data.
• Both news feed but also (surprisingly?) text documents were key unstructured data sources being processed using big data.
• In trading news sentiment and time series analysis were key areas for big data.
• Creation of a system wide trade database for surveillance and compliance was seen as a key area for enhancement by big data.
• Data security remains a big concern with technologists over the use of big data.

There were a few audience questions - Pete clarified that there was a more varied application of big data amongst sell-side firms, and that on the buy-side it was being applied more KYC and related areas. One of the audience made that point that he thought a real challenge beyond the insight gained from big data analysis was how to translate it into value from an operational point of view. There seemed to be a fair amount of recognition that regulators and auditors are wanting a full audit trail of what has gone on across the whole firm, so audit was seen as a key area for big data. Another audience member suggested that the lack of a rigid data model in some big data technologies enabled greater flexibility in the scope of questions/analysis that could be undertaken. 

Coming back to the key findings of the survey, then one question I asked Pete was whether or not big data is a silver bullet for data integration. My motivation was that the survey and much of the press you read talks about how big data can pull all the systems, data and calculations together for better risk management, but while I can understand how massively scaleable data and calculation capabilities was extremely useful, I wondered how exactly all the data was pulled together from the current range of siloed systems and databases where it currently resides. Pete suggested that this was stil a problematic area where Enterprise Application Integration (EAI) tools were needed. Another audience member added that politics within different departments was not making data integration any easier, regardless of the technologies used. 

Overall a good event, with audience interaction unsurprisingly being the most interesting and useful part.

Apex for Interactive (Reference) Data

Launch event for Interactive Data's new reference data service Apex on Wednesday night, hosted at Nasdaq Time Square and introduced by Mark Hepsworth. Apex looks like a good offering, combining multi-asset data access, batch file and on-demand API requests from the same data store, plus hosted data management services, and a flexible licensing/distribution/re-distribution model.

Some good speakers at the event. Larry Tabb ran through his opinions on the current market, starting with regulation. He painted a mixed picture of the market, starting with the continuing exit by investors from the equity mutual funds market, offset to some degree by rapid growth in ETF assets (54% growth over past 3 years to $1,200billion). Obviously events such as the Flash Crash, Libor, the London Whale and Knight Capital have not increased investors confidence in markets either.

On regulation he first cited the sheer amount of regulation being attempted at the moment going through systemic risk/too big to fail, Dodd-Frank, Volcker, derivatives regulation, Basel III etc. Of particular note he mentioned some concerns over whether there is simply enough collateral around in the market given increased capital requirements and derivative regulation (a thought currently shared by the FT apparently in this article).

Given the focus of the event, Larry unsurprisingly mentioned the foundational role of data in meeting the new regulatory requirements, which for the next few years he believes will be focussed on audit and the ability to explain and justify past decisions to regulators. Also given the focus of the event, Larry did not mention his recent article on the Tabb Forum on federated data management strategies which I would have been interested to hear Interactive's comments on, particularly given their new hosted data management offerings. (You can find some of our past thoughts here on the option of using federated data.)

Mike Atkin of the EDM Council was next up and described a framework for what he thought was going on in the market. In summary, he split the drivers for change into business and regulatory, and categorised the changes into:

• Transparency
• Systemic Risk
• Capital and Liquidity
• Clearing and Settlement
• Control and Enforcement

He then that the fundamental challenge with data was to go through the chain of identifying things, descibing them, classifying/aggregating them and then finally establishing linkages. He then ended this part of his presentation with the three aspects he thought necessary to sort this out from industry data standards, to methods of best practice and on to having infrastructure in place to enable these changes. 

Mike then went on to recount a conversation he had had with a hedge fund manager, who had defined the interesting concept of a "Data Risk Equation":

N x CC x S / (Q x V)

where:

N: is the number of variables

CC: is a measure of calculation complexity

S: is the number of data sources needed

Q: is a measure of quality

V: is a measure of verifiability 

I think the angle was the Hedge Fund guy was simply using a form of the above to categorise and compary the complexity of some of the data issues his firm was dealing with.

Aram Flores of Deutsche Bank then talked briefly. Of note was his point that the new regulation was forcing DB to use more external rather than internal data, since regulation now restricted the use of internal data within regulatory reporting. Sounds like good news for Interactive and some of its competitors. Eric Reichenberg of SS&C GlobeOp then gave a quick talk on the importance of accurate data to his derivative valuation services. The talks ended with a well-prepped conversation between Marty Williams and one of their new Apex clients, who jokingly refered to one of the other well-known data vendors as the Evil Empire which raised a few smiles - fortunately the speaker didn't start to choke at this point so obviously Darth Vader wasn't spying on the proceedings...

So overall a good event, new product offering looks interesting, speakers were entertaining and the drinks/food/location were great. 

The Missing Data Gap

Getting to the heart of "Data Management for Risk", PRMIA held an event entitled "Missing Data for Risk Management Stress Testing" at Bloomberg's New York HQ last night. For those of you who are unfamiliar with the topic of "Data Management for Risk", then the following diagram may help to further explain how the topic is to do with all the data sets feeding the VaR and scenario engines.

I have a vested interest in saying this (and please forgive the product placement in the diagram above, but hey this is what we do...), but the topic of data management for risk seems to fall into a functionality gap between: i) the risk system vendors who typically seem to assume that the world of data is perfect and that the topic is too low level to concern them and ii) the traditional data management vendors who seem to regard things like correlations, curves, spreads, implied volatilities and model parameters as too business domain focussed (see previous post on this topic) As a result, the risk manager is typically left with ad-hoc tools like spreadsheets and other analytical packages to perform data validation and filling of any missing data found. These ad-hoc tools are fine until the data universe grows larger, leading to the regulators becoming concerned about just how much data is being managed "out of system" (see past post for some previous thoughts on spreadsheets).

The Crisis and Data Issues. Anyway enough background above and on to some of the issues raised at the event. Navin Sharma of Western Asset Management started the evening by saying that pre-crisis people had a false sense of security around Value at Risk, and that crisis showed that data is not reliably smooth in nature. Post-crisis, then questions obviously arise around how much data to use, how far back and whether you include or exclude extreme periods like the crisis. Navin also suggested that the boards of many financial institutions were now much more open to reviewing scenarios put forward by the risk management function, whereas pre-crisis their attention span was much more limited.

Presentation. Don Wesnofske did a great presentation on the main issues around data and data governance in risk (which I am hoping to link to here shortly...)

Issues with Sourcing Data for Risk and Regulation. Adam Litke of Bloomberg asked the panel what new data sourcing challenges were resulting from the current raft of regulation being implemented. Barry Schachter cited a number of Basel-related examples. He said that the costs of rolling up loss data across all operations was prohibitative, and hence there were data truncation issues to be faced when assessing operational risk. Barry mentioned that liquidity calculations were new and presenting data challenges. Non centrally cleared OTC derivatives also presented data challenges, with initial margin calculations based on stressed VaR. Whilst on the subject of stressed VaR, Barry said that there were a number of missing data challenges including the challenge of obtaining past histories and of modelling current instruments that did not exist in past stress periods. He said that it was telling on this subject that the Fed had decided to exclude tier 2 banks from stressed VaR calculations on the basis that they did not think these institutions were in a position to be able to calculate these numbers given the data and systems that they had in place.

Barry also mentioned the challenges of Solvency II for insurers (and their asset managers) and said that this was a huge exercise in data collection. He said that there were obvious difficulties in modelling hedge fund and private equity investments, and that the regulation penalised the use of proxy instruments where there was limited "see-through" to the underlying investments. Moving on to UCITS IV, Barry said that the regulation required VaR calculations to be regularly reviewed on an ongoing basis, and he pointed out one issue with much of the current regulation in that it uses ambiguous terms such as models of "high accuracy" (I guess the point being that accuracy is always arguable/subjective for an illiquid security).

Sandhya Persad of Bloomberg said that there were many practical issues to consider such as exchanges that close at different times and the resultant misalignment of closing data, problems dealing with holiday data across different exchanges and countries, and sourcing of factor data for risk models from analysts. Navin expanded more on his theme of which periods of data to use. Don took a different tack, and emphasised the importance of getting the fundamental data of client-contract-product in place, and suggested that this was a big challenge still at many institutions. Adam closed the question by pointing out the data issues in everyday mortgage insurance as an example of how prevalant data problems are.

What Missing Data Techniques Are There? Sandhya explained a few of the issues her and her team face working at Bloomberg in making decisions about what data to fill. She mentioned the obvious issue of distance between missing data points and the preceding data used to fill it. Sandhya mentioned that one approach to missing data is to reduce factor weights down to zero for factors without data, but this gave rise to a data truncation issue. She said that there were a variety of statistical techniques that could be used, she mentioned adaptive learning techniques and then described some of the work that one of her colleagues had been doing on maximum-likehood estimation, whereby in addition to achieving consistency with the covariance matrix of "near" neighbours, that the estimation also had greater consistency with the historical behaviour of the factor or instrument over time.

Navin commented that fixed income markets were not as easy to deal with as equity markets in terms of data, and that at sub-investment grade there is very little data available. He said that heuristic models where often needed, and suggested that there was a need for "best practice" to be established for fixed income, particularly in light of guidelines from regulators that are at best ambiguous.

I think Barry then made some great comments about data and data quality in saying that risk managers need to understand more about the effects (or lack of) that input data has on the headline reports produced. The reason I say great is that I think there is often a disconnect or lack of knowledge around the effects that input data quality can have on the output numbers produced. Whilst regulators increasingly want data "drill-down" and justfication on any data used to calculate risk, it is still worth understanding more about whether output results are greatly sensitive to the input numbers, or whether maybe related aspects such as data consistency ought to have more emphasis than say absolute price accuracy. For example, data quality was being discussed at a recent market data conference I attended and only about 25% of the audience said that they had ever investigated the quality of the data they use. Barry also suggested that you need to understand to what purpose the numbers are being used and what effect the numbers had on the decisions you take. I think here the distinction was around usage in risk where changes/deltas might be of more important, whereas in calculating valuations or returns then price accuracy might receieve more emphasis. 

How Extensive is the Problem? General consensus from the panel was that the issues importance needed to be understood more (I guess my experience is that the regulators can make data quality important for a bank if they say that input data issues are the main reason for blocking approval of an internal model for regulatory capital calculations). Don said that any risk manager needed to be able to justify why particular data points were used and there was further criticism from the panel around regulators asking for high quality without specifying what this means or what needs to be done.

Summary - My main conclusions:

• Risk managers should know more of how and in what ways input data quality affects output reports
• Be aware of how your approach to data can affect the decisions you take
• Be aware of the context of how the data is used
• Regulators set the "high quality" agenda for data but don't specify what "high quality" actually is
• Risk managers should not simply accept regulatory definitions of data quality and should join in the debate

Great drinks and food afterwards (thanks Bloomberg!) and a good evening was had by all, with a topic that needs further discussion and development.

 

 

Bankenes Sikringsfond Selects Xenomorph's TimeScape for Faster Data Analysis and High-Quality Decision Support

Just a quick note to say that we have signed a new client, Bankenes Sikringsfond, the Norwegian Banks’ Guarantee Fund. They will be using TimeScape to fulfill requirements for a centralised analytics and data management platform. The press release is available here for those of you who are interested.

We Can’t Upgrade, the Data Model’s Changed!

New article with some of my thoughts on data models, interfaces and software upgrades has just gone up on the Waters Inside Reference Data site.

 

Charting, Heatmaps and Reports for TimeScape, plus new Query Explorer

We have a great new software release out today for TimeScape, Xenomorph's analytics and data management solution, more details of which you can find here. For some additional background to this release then please take a read below.

For many users of Xenomorph's TimeScape, our Excel interface to TimeScape has been a great way of extending and expanding the data analysis capabilities of Excel through moving the burden of both the data and the calculation out of each spreadsheet and into TimeScape. As I have mentioned before, spreadsheets are fantastic end-user tools for ad-hoc reporting and analysis, but problems arise when their very usefulness and ease of use cause people to use them as standalone desktop-based databases. The four-hundred or so functions available in TimeScape for Excel, plus Excel access to our TimeScape QL+ Query Language have enabled much simpler and more powerful spreadsheets to be built, simply because Excel is used as a presentation layer with the hard work being done centrally in TimeScape.

Many people like using spreadsheets, however many users equally do not and prefer more application based functionality. Taking this feedback on board has previously driven us to look at innovative ways of extending data management, such as embedding spreadsheet-like calculations inside TimeScape and taking them out of spreadsheets with our SpreadSheet Inside technology. With this latest release of TimeScape, we are providing much of the ease of use, analysis and reporting power of spreadsheets but doing so in a more consistent and centralised manner. Charts can now be set up as default views on data so that you can quickly eyeball different properties and data sources for issues. New Heatmaps allow users to view large colour-coded datasets and zoom in quickly on areas of interest for more analysis. Plus our enhanced Reporting functionality allows greater ease of use and customisation when wanting to share data analysis with other users and departments.

Additionally, the new Query Explorer front really shows off what is possible with TimeScape QL+, in allowing users to build and test queries in the context of easily configurable data rules for things such as data source preferences, missing data and proxy instruments. The new auto-complete feature is also very useful when building queries, and automatically displays all properties and methods available at each point in the query, even including user-defined analytics and calculations. It also displays complex and folded data in an easy manner, enabling faster understanding and analysis of more complex data sets such as historical volatility surfaces. 

Paris Financial Information Summit 2012

I attended the Financial Information Summit event on Tuesday, organized in Paris by Inside Market Data and Inside Reference Data.

Unsurprisingly, most of the topics discussed during the panels focused on reducing data costs, managing the vendor relationship strategically, LEI and building sound data management strategies.

Here is a (very) brief summary of the key points touched which generated a good debate from both panellists and audience:

Lowering data costs and cost containment panels

• Make end-users aware of how much they pay for that data so that they will have a different perspective when deciding if the data is really needed or a "nice to have"
• Build a strong relationship with the data vendor: you work for the same aim and share the same industry issues
• Evaluate niche data providers who are often more flexible and willing to assist while still providing high quality data
• Strategic vendor management is needed within financial institutions: this should be an on-going process aimed to improve contract mgmt for data licenses
• A centralized data management strategy and consolidation of processes and data feeds allow cost containment (something that Xenomorph have long been advocating)
• Accuracy and timeliness of data is essential: make sure your vendor understands your needs
• Negotiate redistribution costs to downstream systems

One good point was made by David Berry, IPUG-Cossiom, on the acquisition of data management software vendors by the same data providers (referring to the Markit-Cadis and PolarLake-Bloomberg deals) and stating that it will be tricky to see how the two business units will be managed "separately" (if kept separated...I know what you are thinking!).

There were also interesting case studies and examples supporting the points above. Many panellists pointed out how difficult can be to obtain high quality data from vendors and that only regulation can actually improve the standards. Despite the concerns, I must recognize that many firms are now pro-actively approaching the issue and trying to deal with the problem in a strategic manner. For example, Hand Henrik Hovmand, Market Data Manager, Danske Bank, explained how Danske Bank are in the process of adopting a strategic vendor system made of 4 steps: assessing vendor, classifying vendor, deciding what to do with the vendor and creating a business plan. Vendors are classified as strategic, tactical, legacy or emerging. Based on this classification, then the "bad" vendors are evaluated to verify if they are enhancing data quality. This vendor landscape is used both internally and externally during negotiation and Hovmand was confident it will help Danske Bank to contain costs and get more for the same price.

I also enjoyed the panel on Building a sound management strategy where Alain Robert- Dauton, Sycomore Asset Management, was speaking. He highlighted how asset managers, in particular smaller firms, are now feeling the pressure of regulators but at the same time are less prepared to deal with compliance than larger investment banks. He recognized that asset managers need to invest in a sound risk data management strategy and supporting technology, with regulators demanding more details, reports and high quality data.

For a summary on what was said on LEI, then seems like most financial institutions are still unprepared on how it should be implemented, due to uncertainty around it but I refer you to an article from Nicholas Hamilton in Inside Reference Data for a clear picture of what was discussed during the panel.

Looking forward, the panellists agreed that the main challenge is and will be managing the increasing volume of data. Though, as Tom Dalglish affirmed, the market is still not ready for the cloud, given than not much has been done in terms of legislation. Watch out!

The full agenda of the event is available here.

OIS Discounting and Curve Management - Wilmott Event - Thursday 31st May

Quick plug for Xenomorph's Wilmott Forum Event on OIS curves tomorrow in downtown Manhattan. The event is done in partnership with Numerix, and will be looking at the issue of OIS vs. Libor discounting from the point of view of a practioner, financial engineer and systems developer. You can register for the event here, and so we hope to see you at 6pm for some great talks and some drinks/socialising afterwards.

TabbFORUM Video: Data Is Revenue

Video interview with Paul Rowady of the Tabb Group, primarily about how data management can break out from being just a back office function and become a source of competitive advantage in both the front office and in risk management.

For those of you with a curious mind, the perseverence to watch the video until the end and possibly not such advanced years as me and Paul, then the lead singer of Midnight Oil that he refers to at the close of the video is Peter Garrett, who looks like this:

Whereas I look like this:

See, completely different. Obviously Peter has a great choice in hairstyle though...

CVA - a business driver for breaking down asset silos

Xenomorph's analytics partner Numerix sponsored a PRMIA event at New York's Harvard Club this week on Credit Valuation Adjustment (CVA). The event also involved Microsoft, with a surprisingly relevant contribution to the evening on CVA and "Big Data" (I still don't feel comfortable losing the quotes yet, maybe soon...). Credit Valuation Adjustment seems to be the hot topic in risk management and pricing at the moment, with Numerix's competitor Quantifi having held another PRMIA event on CVA only a few months back. 

The event started with an introduction to CVA from Aletta Ely of JP Morgan Chase. Aletta started by defining CVA as the market value of counterparty credit risk. I am new to CVA as a topic, and my own experience on any kind of adjustment in valuation for instrument was back at JP Morgan in the mid-90s (those of you under 30 are allowed to start yawning at this point...). We used to maintain separate risk-free curves (what are they now?) and counterparty spread curves, which would be combined to discount the cashflows in the model.

Whilst such an adjustment could be calibrated to come up with an adjusted valuation which would be better than having no counterparty risk modelled at all, it seems one of the key aspects of how CVA differs is that a credit valuation adjustement needs to be done in the context of the whole portfolio of exposures to the counterparty, and not in isolation instrument by instrument. The fact that a trader in equity derivatives was long exposure to a counterparty cannot be looked at in isolation from a short exposure to a portfolio of swaps with the same counterparty on the fixed income desk.

Put another way, CVA only has context if we stand to lose money if our counterparty defaults, and so an aggregated approach is needed to calculate the size of the positive exposures to the counterparty over the lifetime of the portfolio. Also, given this one sided payoff aspect of the CVA calculation, then instrument types such as vanilla interest rate swaps suddenly move from being relatively simple instrument that can be priced off a single curve to instruments that needed optionality to be modelled for the purposes of CVA.

So why has CVA become such a hot topic at the banks? Prior to the 2008/2009 crisis CVA was already around (credit risk has existed for a long time I guess, regardless of whether you regulate or report to it), but given that bank credit spreads were at that time consistently low and stable then CVA had minimal effects on valuations and P&L. Obviously with the advent of Lehmans then this changed, and CVA has been pushed into prominence since it has directly affected P&L in a significant manner for many institutions (for example see these FT articles on Citi and JPMorgan)

A key and I think positive point for the whole industry is the CVA requires a completely multi-asset view, and given regulatory focus on CVA and capital adequacy then as a result it will drive banks away from a siloed approach to data and valuation management. If capital is scarcer and more costly, then banks will invest in understanding both their aggregate CVA and the incremental contribution to CVA of a new trade in the context of all exposures to the counterparty. Looking at incremental CVA, then you can also see that this also drives investment into real or near-realtime CVA calculation, which brings me on to the next talks of the evening by Numerix on CVA calculation methods and a surprisingly good presentation on CVA and "Big Data" from David Cox of Microsoft.

Denny Yu of Numerix did a good job of explaining some of the methods of calculating CVA, and in addition to being cross asset and all the implications that requires for having the ability to price anything, CVA is both data and computationally expensive. It requires both simulation of the scenarios for the default of counterparties through time, but also the valuation of cross-asset portfolios at different points in time. Denny mentioned techniques such as American Monte-Carlo to reduce the computation needed through using the same simulation paths for both default scenarios and valuation.

So on to Microsoft. I have seen some appalling presentations on "Big Data" recently, mainly from the larger software and hardware companies try to jump on the marketing band wagon (main marketing premise: the data problems you have are "Big"...enough said I hope). Surprisingly, David Cox of Microsoft gave a very good presentation around the computation challenges of CVA, and how technologies such as Hadoop take the computational power closer to the data that needs acting on, bringing the analytics and data together. (As an aside, his presentation was notably "Metro" GUI in style, something that seems to work well for PowerPoint where the slide is very visual and it puts more emphasis on the speak to overlay the information). David was obviously keen to talk up some of the cloud technology that Microsoft is currently pushing, but he knew the CVA business topic well and did a good job of telling a good story around CVA, "Big Data" and Cloud technologies. Fundamentally, his pitch was for banks and other institutions to become "Analytic Enterprises" with a common, scaleable and flexible infrastructure for data management and analysis. 

In summary it was a great event - the Harvard Club is always worth a visit (bars and grandiose portraits as expected but also barber shop in the basement and squash courts in the loft!), the wine afterwards was tolerably good and the speakers were informative without over-selling their products or company. Quick thank you to Henry Hu of IBM for transportation on the night, and thanks also to Henry for sending through this link to a great introductory paper on CVA and credit risk from King's College London. Whilst the title of the King's paper is a bit long and scary, it takes the form of dialogue between a new employee and a CVA expert, and as such is very readable with lots of background links.