Frustrated Programming (FP)

Single Name CDS Futures

2007-03-21T19:13:00.000-04:00

The latest issue of Futures Industry mentions the impending (March 27) Eurex entry into the CDS index futures market and the more interesting single name CDS futures that CME and CBOT are thinking about. This is definitely something to follow since there's likely to be some growing pains (read: arb?) between the OTC and Futures markets while things get sorted out.

This issue also has an update on the CFTC cross-margining story that we talked about in a previous post.

Andrew Lo in NYC on Systemic Risk

2007-03-13T07:47:00.000-04:00

Andrew Lo was in town last week to talk Systemic Risk but I missed the event. If anyone has soft copy of the presentation please email me. I'm particularly interested in hearing his evidence on increasing systemic risk and thoughts on regulatory need:

Prof. Lo attempts to quantify the potential impact of hedge funds on systemic risk by proposing several new risk measures for hedge funds and applying them to individual and aggregate hedge-fund returns data. His preliminary findings suggest that the hedge-fund industry may be heading into a challenging period of lower expected returns, and that systemic risk is currently on the rise. Professor Lo will conclude by proposing a new regulatory body designed to address these issues directly.

Beta Arb

2007-03-13T07:45:00.000-04:00

Interesting thoughts on a beta arb play.....

Smaller Cap Algo

2007-03-09T06:05:00.000-05:00

A few weeks ago I was talking to an up and coming buy-side incubated prime broker who was offering a number of algorithims to its clients, specifically for micro/small/mid cap equities. I see BofA is talking smaller cap algo now too.

The Paradox of Low-Latency XML for Algo Trading

2007-03-07T08:29:00.001-05:00

Dow Jones announced support for XML-tagged news in response to Reuters support for the same announced in December of 2006.

http://www.finextra.com/fullstory.asp?id=16606

"The ultra-low-latency service delivers economic and corporate news in precise and discrete elements in XML-tagged fields."

Since bandwidth and I/O are big contributors to latency, I am surprised these vendors aren't also offering a more proprietary, compression-based structured news encoding to reduce bandwidth requirements.

I guess the next big announcement in this space will be the availability of structured news services like these being co-located in the exchanges' data centers. That's the real ticket to low-latency.

Index Arb Op Risk

2007-03-01T11:29:00.001-05:00

Where were you at 2:30 PM on Feb 28, 2007? Anywhere near the DJIA?

President's Working Group on Systemic Risk

2007-02-27T20:41:00.000-05:00

Interesting activity in the regulatory space....

FreeSubprimeCreditReport.com

2007-02-27T20:33:00.000-05:00

Lots of press recently about how the mounting losses in the subprime mortgage market may impact the hedge funds who assume this subprime risk from the banks who package it up and sell it off. The first I saw was in The Economist and then Wharton.

Logistics in Algo Trading

2007-02-27T20:29:00.000-05:00

There's a talk at an upcoming Algo Trading conference in London on "Market Impact Models and Trade Scheduling". I wonder how long before the players in this space start taking more advantage of the vast world of operations research in scheduling.

Equity Correlation Swaps

2007-02-27T20:12:00.000-05:00

Columbia had a discussion on Equity Correlation Swaps last night......for those of us who like matrices...

The New World of Portfolio Margining

2007-02-17T08:50:00.000-05:00

Another article in the most recent Futures Industry magazine on new margining rules for the NYSE and CBOE that are intended to come online on April 2 also caught my attention because of its likely impact on the prime brokers and hedge funds, which we've mentioned before in a previous post, as well as on the portfolio management and credit risk management systems we build for our clients.

Specifically, NYSE Rule 431 and CBOE Rule 12.4 which focus on cash equity, listed and OTC equity derivatives should reduce client margin requirements and increase leverage. It does sound like there are some limits imposed, including a $5 MM floor on margin equity before allowing the incorporation of OTC derivatives, and a credit ceiling on 10 times regulatory capital.

It's also very interesting to note that besides the impact on portfolio management and margining, these new rules impose more explicit credit risk management processes and controls.

One thing that Michael Jamroz, the author, does mention though is that true cross-margining for equities and futures is not an option yet because of the Commodity Exchange Act's segregation rules for futures margin, although he does note that the CFTC may at some point be in a position to change that.

The Senate and Systemic Risk

2007-02-17T08:00:00.000-05:00

In the latest issue of Futures Industry magazine, John Damgard mentions two members of the Senate to watch for in 2007.

Senator Dodd from Connecticut, the new head of the Senate Banking Committee who could hold hearings this year on federal regulations to limit systemic risk.

Senator Feinstein from California who favors increasing the jurisdiction of the CFTC over the OTC world.

The Systemic Risk of Cross-Asset Everything

2007-01-18T09:04:00.000-05:00

2006 was a year when cross-asset trading and risk really started to get a lot of press, including this post on cross-asset algo trading from Chris. Particularly interesting to the buy-side was the potential for increased leverage they would get from cross-asset margining and netting.

The thing that sticks out most in my mind though is something that Timothy Geithner, the CEO of the NY Fed said in a September 2006 presentation in Hong Kong. In this presentation he highlights the relationship between leverage and liquidity.

Let me give two examples of evolving market practices that may help alleviate one concern only to exacerbate another. Collateral plays an increasingly important role in counterparty credit risk management, particularly for highly leveraged counterparties. The increased importance of variation margining plays a critical role in counterparty credit risk management. These changes help limit the exposure of the core financial institution to losses among their leveraged counterparties, but they also act to exacerbate volatility, with asset price declines forcing further margin calls, adding for further market declines. Where initial margin is thin in relation to potential exposure, counterparties are more exposed to adverse movements in asset prices, and in a situation of stress the actions they take to reduce their exposure to further losses are likely to have a greater negative impact on market dynamics. In market conditions where initial margin may be low relative to potential future exposure, the self-preserving behavior of leveraged funds and their counterparties may be more likely to exacerbate rather than mitigate an unexpected deterioration in asset prices and market liquidity. As financial firms demand more collateral, funds are forced to liquidate positions, adding to volatility and pushing down asset prices, leading to more margin calls and efforts by the major firms to reduce their exposure to future losses. In the context of the previous discussion of externalities, firms’ incentives to minimize their own exposure can amplify the initial shock and impose on others the negative externality of a broader disruption to market liquidity.

S&P Rating Hedge Fund Op Risk

2006-12-21T09:13:00.000-05:00

Given that the SEC hedge fund regulation story is waning fast....

Hedge funds may have a new incentive to voluntarily improve their operational risk story after getting wind of this effort by S&P ..... besides the obvious benefits of additional capacity and business scalability that investment in reducing op risk can yield.

How will prospective investors use these ratings to guide their investment decisions?

How will this impact the current composition of hedge fund performance indices?

When combined/weighted with AUM and general strategy info, I wonder if these ratings can be aggregated across funds to better estimate systemic risk?

Synthetic Hedge Funds

2006-12-07T08:29:00.000-05:00

Suren's post reminded me of a recent Economist article that talked about synthesizing hedge fund portfolios and provided some useful links to additional reading material, including something from the Andrew Lo camp. I'm a big fan of Andrew's work, particularly his 2000-2001 views on hedge fund risk that will likely be taking root in 2007-2008.

Hull on Hedge Fund Risk

2006-10-01T21:33:00.000-04:00

A recent Economist article hints at what John Hull spells out pretty clearly in his lessons for "users of derivatives":

1) Define Risk Limits
2) Take the Risk Limits Seriously
3) Do Not Underestimate the Benefits of Diversification
4) Monitor Traders Carefully
5) Do Not Blindly Trust Models
6) Do Not Ignore Liquidity Risk
7) This one is my favorite, having had a previous life experience on a system designed to improve the valuation of a certain balance sheet asset so that it could be used for greater leverage.....Be Cautious about Making the Treasury Department a Profit Center

CEP ... It's Easy As ... ESP

2006-09-04T21:10:00.000-04:00

First Aaron joined StreamBase, then the great ACM Queue magazine featured an interview with Steven Ross-Talbot, one of the current voices in the pi-Calculus event space ... but lately I've been reading Marco and Chris Donnan's latest. It's also probably not too long until Marc takes his latest one step further and gets deeper into the utility of ESP.

The Multi-Dimensional Data Grid

2006-09-04T20:59:00.000-04:00

Thanks to Chris Webb for pointing us to Google's Big Table. Since we're big believers in the dimensionality of things, when you read through Google's paper it becomes evident that they're on to something with broader applicability here. There's also potential synergy with the Data Warehousing and OLAP worlds, or maybe even the ESP space.

Also take a look at the Pentaho folks doing Open Source Business Intelligence....

More on the Application of FPGAs, ASICs, GPUs, etc... to Finance

2006-08-28T06:16:00.000-04:00

After last week's post on FGPAs, I've talked with a number of people over the past week about high performance quant operations in hardware and have happened on a few more variations on the theme.

Accelogic was the first vendor that got me turned on to FPGAs.

Mark's post about pushing C code down into both FPGAs and ASICs points out Celoxica.

FPGA Journal and The Structured ASIC Association

Mike Roberts gave me a heads up on Altera who he knows from a previous life.

Chris Donnan alerted me to an interesting application of GPUs to the Folding at Home problem.

And last but not least, some more reading on the application of the FFT in Jump Diffusion.

I'm also guessing that this silicon should be on the agenda at the World Congress on Computational Finance.

Black Scholes Merton Lives On !!!

2006-08-26T16:24:00.000-04:00

Interesting historical commentary and thoughts on the longevity of BSM especially when you move beyond constant volatility into the stochastic world....

Preparing for the Days of Manycore CPUs

2006-08-26T15:43:00.000-04:00

Eric aggregates some interesting thoughts on raising the level of programming consciousness to support the inherent parallelization in the manycore CPUs and related hardware of tomorrow ... and it's nice to see Erlang getting some press again.

GPGPU for Faster Options Math in Silicon

2006-08-20T15:58:00.000-04:00

Ever since the FPGA discussion piqued my interest a few months ago I've been paying more attention than usual to other high speed opportunities, the most compelling of which has to be the GPGPU movement.

What first got my attention were two articles in IEEE Computer magazine on Game Physics in Silicon & the power of GPUs.

Check out NVIDIA who also happens to have a good story around SLI (see chapter 8), which allows multiple GPUs to work together in parallel. Rounding out the vendor mix is Ageia whose PhysX API seems to be the closest to Microsoft and the XBox.

Can you imagine having GPU access from .NET as well? The answer is Accelerator.

For general GPGPU info check out http://www.gpgpu.org/

So what goes faster when you throw it on the GPU?

For starters, how about the Fast Fourier Transform.

Peter Carr talks about the FFT in the context of jumpier Black Scholes.

Binomial options pricing can also benefit from the FFT, with better sensitivity to stoch vol, etc....

Exciting stuff indeed !!!

Insurance & Reinsurance Regulations in the US

2006-08-13T18:49:00.000-04:00

The Economist turned me on to some interesting info on pending changes to the Insurance & Reinsurance industries in the US. This includes proposed legislation for federally charted Insurance firms as well as an attempt to get US and non-US Reinsurance firms on par with each other when it comes to collateral requirements. It sounds like Lloyd's has been pushing for change for a long time.

The Economist Blogosphere

2006-08-13T18:08:00.000-04:00

The Economist recently had an article on the use of blogging to flatten the ivory tower of economic thought. Here are the blogs specifically mentioned in the article:

http://delong.typepad.com/

http://becker-posner-blog.com/

http://rgemonitor.com/blog/setser/

DMA Risk and Intraday VAR

2006-08-13T17:48:00.000-04:00

Also in the latest FI are thoughts on risk in the increasingly common world of DMA ....... with notes on Prudential's use of Intraday VAR.

After the Trade is Made - Perspectives @ FIA Roundtable

2006-08-13T17:39:00.000-04:00

Some thoughts on back office automation (and other post-trade topics) in the latest issue of FI......

Several broker representatives who participated in the roundtable asked if the endusers would welcome a move towards the use of FIXML as a standard message format. The cost-benefit was not clear to all end users. One end user warned that moving to FIX was a big burden in cost terms for both end-users and brokers. The initial cost of a full FIX implementation is so high—he estimated $500,000—that many users are balking.

Still another end-user commented that his firm relies on direct connections to the exchanges. "I don't care how the data comes in from the firms," he said. The more critical issue for his firm, he added, is getting an intraday rack. "We can't reconcile intraday in back-office clearing."

Credit Derivatives Training Oct 2006 in NYC

2006-08-13T17:24:00.000-04:00

Check out the syllabus for this PRMIA event....cool stuff

Buy-Side Priorities

2006-08-11T20:27:00.000-04:00

Thanks to Bob for alerting us to the 2007 Buy-Side Priorities for UK firms......I would venture a guess that OTCs will continue to occupy the thoughts of fund managers on this side of the pond as well.

More Hedge Fund Valuation Chatter

2006-08-11T20:19:00.000-04:00

To go along with our previous post, here is a new GARP article that aggregates some current thought on independent valuations.....

Intraday VAR

2006-08-11T18:31:00.000-04:00

This paper on Intraday VAR won the Best Paper in Risk Management award from PRMIA......and here are a few more from the conference.

FPGAs on Wall Street

2006-08-11T18:25:00.000-04:00

The High Performance on Wall Street conference is happening in NYC on September 18 2006.

Nice to see FPGAs on the Agenda.....

For additional reading on High Perf, check out Tom Coleman's recent FEN article too.

FIA & OIC Equity Options Conference

2006-07-30T08:43:00.000-04:00

The program at the upcoming FIA & OIC Equity Options conference in NY sounds interesting, especially the talks on BestEx and Liquidity.

Isn't it about time that Liquidity was treated like a full-fledged risk citizen? We do have some of the Greek alphabet not accounted for yet and Jarrow makes an attempt here, as does Duffie.

Independent Valuations

2006-07-23T07:59:00.000-04:00

I didn't expect to see the hedge fund valuation topic appear in Bottom Line Personal, although it was a good segue into the plug for Citco and GlobeOp and similar service providers.

VaR Tail Fat

2006-07-22T07:36:00.000-04:00

The fat is in the tails ...... Risk Metrics was talking about this 10 years ago .... FEN has republished an article recently too as part of their 10th anniversary reprinting of classic articles ....

CDS Confirmation Backlog

2006-07-22T07:28:00.000-04:00

80%, that's an impressive number.... Novation really takes on new meaning this way .... much different than the old studio days :)

Hedge Fund Systemic Risk

2006-07-17T19:24:00.000-04:00

Interesting commentary on how hedge funds may create systemic risk by attracting certain types of talent ...

And some good background on systemic risk ...

The End of the Credit Rating Duopoly As We Know It?

2006-07-16T09:54:00.000-04:00

It looks like Bucks County isn't just for coffee anymore....

Rep. Fitzpatrick from that district has recently passed Bill that "removes the Securities and Exchange Commission (SEC) from the process of approving certain rating agencies as nationally recognized statistical rating organizations".

I'm really curious to see what the impact of this will be on CDS spreads and the overall CDS market in general. Hull has done some interesting research of rating agency events on spreads. I'm also curious to see the impact of this on Basel II, because the BIS clearly pays close attention to rating agency event impact on credit.

Hedge Fund Regulations

2006-07-01T17:52:00.000-04:00

The Economist is getting in on the Hedge Fund Regulation discussion .......

Right at the center of it all is CT Attorney General Blumenthal who recently testified to the Senate Judiciary Committee....

Thankfully most of what we're hearing sounds like the regulators want to be sensitive to the important economic benefits of hedge funds ........ and balance that sensitivity with continued diligence in keeping systemic risk under control....

Quant Congress & Faster Numerics

2006-06-26T20:43:00.000-04:00

Check out the keynoters @ Incisive's Quant Congress : July 12-13 in NYC

Also, if you happen to be in Ohio on July 12-14, check out the interesting discussion on using FPGAs for faster numerics. (Thanks to Wilmott for the heads up on this one).

Good Timing Pixar

2006-06-26T20:35:00.000-04:00

Pixar must have timed the release of Cars and its message of what's important in life to coincide with the 50th anniversary of Ike's crowning achievement: the Interstate Highway System.

Before the Trade Is Made

2006-06-06T09:36:00.000-04:00

Could this be the year that buy-side pre-trade analytics and pre-trade due-diligence apps start to take advantage of post-trade analytics technology like the trusty Microsoft Analysis Services OLAP solution?

http://www.it-analysis.com/business/regulation/content.php?cid=8412

http://www.wbresearch.com/tradetechusa/full.html

You've Been Metcalfed

2006-05-21T07:15:00.000-04:00

I'm still advocating the use of LinkedIn for all. A few of my colleagues and I have even begun to use LinkedIn as a way to gain Metcalfe-sized returns from our social networks.

So if you haven't signed up already.....or you've deleted previous invitations.........

$1000 Grid?

2006-04-27T12:15:00.000-04:00

Could China eventually be the answer to the $1000 grid challenge?

Don't Test Everything

2006-04-27T12:12:00.000-04:00

An Agile version of the 80/20 rule ...

Volatility by any other name...

2006-04-26T07:59:00.000-04:00

Being classically trained in Black-Scholes-Merton, I never could get my mind around why Vega is sometimes called Kappa or Lambda.....

Could it really be a puritanical Greek Alphabet thing?

Beantown Capital

2006-04-24T08:18:00.000-04:00

Boston's presence in the hedge fund world....

Go Yankees!!!

Hedgestock

2006-04-12T08:14:00.000-04:00

For those with or without fond memories of Woodstock, out of the ashes comes Hedgestock, promised to be the UK networking event of the year.....

I even heard Alvin Lee may show up.....I wonder if he can play as fast as he did in 1969

RedHat buys JBoss

2006-04-10T14:24:00.000-04:00

http://www.redhat.com/about/news/prarchive/2006/jboss.html

Are You LinkedIn Yet?

2006-03-29T11:17:00.000-05:00

Still waiting to see more folks on Linked In....

The Assimilation Sprint

2006-01-20T09:55:00.000-05:00

I've always been a big fan of the "Integrate Early" and "Integrate Often" approaches to project execution and planning sprints around integration points. The current project I'm working has forced me to delve a little more into the importance of integration. We're working in an environment where applications need to use centralized, shared libraries of code and need to be deployed in a standard way across the enterprise.

Why is this library dependency and shared service integration different than intra-project component integration? Well, for one, shared service capabilities are important institutional knowledge that consultants don't always understand or use. We get competitive advantage from being able to assimilate into the development and deployment culture of the organization by embracing these shared libraries and deployment mechansisms on Day 1 and making them an important part of Sprints, just as important as the Functional bits we're delivering.

Excel - The Trader's Data Integration Portal

2006-01-20T07:36:00.000-05:00

When you hear something like this, it sounds almost too good to be true:

Give traders an application where they can integrate all the data they need in whatever way they need to integrate it. It doesn't matter what data it is: market data, position data, risk data, operational data, news.... just let the traders loose with it and let them build the relationships between the data without waiting for the IT department to figure out how to do it.

Hmmm....sounds like Excel to me.

Property Derivatives

2006-01-15T11:42:00.000-05:00

What will they think of next?

Property Derivatives is a new and illiquid market, as most derivatives markets are when they first develop :)

Useful Tools for 2006

2005-12-21T08:12:00.000-05:00

Here is an annotated list of tools that will likely become more prominent in 2006. In the quest to build more functionality faster, tools like this can help.

VMWare Workstation is a great concept. Use one machine but create a bunch of virtual machines that resemble your target deployment environments. This could work for testing code on the client side, server side and everything in between. It's also useful for training.

Windows Workflow Foundation answers the need for .NET infrastructure to support workflow development. Most of the front-office apps we build (and even some of the middle-office apps) usually support some sort of workflow support between users to accomplish some task. Historically, toolkits to support workflow development have been light to non-existent so having infrastructure like this can help speed development. An article on this technology can be found here.

ER/Datagen for generating test data permutations from an ERWin data model. Interesting concept whose time has yet to come. This ties in nicely to my previous blog entries about Data Integration testing.

Microsoft support for Domain Specific Languages (DSLs) is starting to get some good press. DSLs open up opportunities to give end-users a controlled programming environment like Microsoft's VBA. How "controlled" is up to you, but the value proposition for domain specific languages in front-office trading applications is clearly visible in Kapital. Finetix and Kapital go way back so we're aware of its value described here. An interesting article on DSLs is here.

Middle Office Agile - Test Driven Sprints

2005-12-20T08:30:00.000-05:00

In big middle-office risk organizations there are far reaching business, operational, technical and political implications for "going live" in production. That makes 30-day Scrum sprints of production code nearly impossible.

These middle office organizations usually consist of a heterogeneous mix of systems and technologies, each of which performs some function (eg: data sourcing, risk calculation, risk reporting, etc...) As mentioned in a previous posting, data integration is a big challenge in middle-office technology projects. In fact, all forms of technical and operational integration dwarf any other challenges in middle-office projects. The bulk of my middle-office work to date has been in setting up and executing "integration" projects like these.

A number of Agile principles can be instrumental in making middle-office projects successful.

From a Project Management standpoint, Scrum Sprints make so much sense because they
delivery quickly and consistently and minimize project risk tremendously. The most success
I've had in the middle-office was with early and frequent integration milestones. Almost always these milestones are completed in an integration testing environment where different teams come together for the first time.

The frequent milestones are helpful because they force teams to "Integrate Often". The scope of these milestones can be considered "Unit Tests" of each integration point.

I think we can be more specific about how Agile works in the middle office by using the term "Test Driven Sprint" to describe the goal of a middle-office Scrum Sprint. Please note that I'm purposely playing on the well-known "Test Driven Development" movement here :)

Agile Relationship Management?

2005-12-19T08:27:00.000-05:00

Is there such a thing as Agile Relationship Management? One of the reasons Finetix has thrived in all economic climates is the ability of its people to maintain very strong relationships with past, current and prospective clients. Is there something more fundamental going on behind the scenes? Why do good software people also maintain good relationships with the people they build software with and the people they build software for?

Agile Rescues CMM?

2005-12-16T07:49:00.000-05:00

I finished up my 9 month Regulatory Credit Risk gig last week. We worked about 40 days straight for the UAT phase, including some late night Thanksgiving runs, but it was well worth it. This multi-million dollar project went into production with nearly all of the scope and functionality promised a year earlier.

The last 2 months of the project were critical and a number of important practices made all the difference in the world. Ironically, don't these smell like the best bits of Scrum (with CMM Marathons instead of Sprints) and XP ?

1) Daily defect review meeting with the Business & Technology teams, including all management on both sides of the fence. This meeting was held every day, without fail, rain or shine. The content for the meeting was also exceptionally managed using Mercury Test Director as the foundation, but with lots of Excel Pivot Table manipulation.

2) Blurred distinctions between traditional CMM roles and responsibilities. Business analysts, Developers, Project Managers; everyone got their hands dirty and the fresh perspectives and frequent, candid discussion helped the forest stay very visible above the trees.

3) When it became clear that 6-9 months of requirements analysis and design documentation didn't cover all the nitty gritty details and in some cases were horribly out of date, Business Analysts paired with Developers to drive the code to completion. You know something is working when your Business' Managing Director starts being able to spot issues with J2EE configuration disparaties between testing environments.

Overall, this large project (40+ people) was successful because of very strong project management practices and business-led PM of the technology deliverables. One of the most interesting practices was the superb technology dependency management to help control the implementation across a very distributed, component based credit risk architecture.

Even though 2005 was successful, everyone in the post-mortem agreed that what we accomplished would likely not be repeatable.

How to improve for 2006?

Some key practices could help put this program on a path for repeatable success in 2006 and beyond:

1) Have the Near-Shore development team institute functional unit testing of the credit risk rules. Hundreds of defects were identified late in the integration testing cycle and limited regression testing was in place as I left.

2) Daily Stand-up during all phases of the CMM Level 2 lifecyle, particularly during requirements analysis and design, when the "sign-off" trap is all too comforting and risky.

3) Most middle-office risk projects don't own any of the data they use and they often have to get it from multiple suppliers; sometimes hundreds of suppliers. This project needs to promote data integration to be a full-fledged development effort and treat it with the same care as the Java programming effort, including all requisite artifacts and Unit Testing tools, etc...