F# for single dealer platforms

Matt shares his thoughts on F# for SDP implementation here. I don’t know as much about F# as he does, but I have dabbled with VS2010 and played with OCAML on Ubuntu at home. I’m sure the stateless functional approach will be a big help in structuring asynchrony and concurrency in server processes. So I agree with the general thrust of his argument here.

I would take issue with one of Matt’s statements: “If you think about an SDP, its primarily an integration build”. I have heard the same view espoused by managers who think an SDP build is just a trivial matter of hooking up some vendor servers like Caplin’s Liberator to internal middleware like the TIB, and hey presto, quotes and executions will flow. It’s nowhere near that simple. I prefer to say that building an SDP is like building an ECN. Think of it as building you’re own Bloomberg or TradeWeb.

Max on broker APIs

Max Dama’s posted a couple of times recently on broker APIs. In his most recent post he drops an intriguing aside when commenting on IB, Wex and Lime: “I haven’t used each of the three’s FIX APIs”

Which prompts the question: why not ?  Why code to a proprietary API rather than an industry standard protocol. I’ve had less than happy experiences with FIX for Fixed Income myself, so my guess at the reasons would be…

• Convenience: going straight to a Java or .Net API is quicker than selecting a FIX implementation and/or protocol implementation like quickfix.org, and then doing an integration.
• Statefullness: AFAIK FIX doesn’t address state like “order submitted”, “order on the book”, “order cancelled” etc. A well sorted API with a good callback interface makes this kind of thing earlier. FIX is just a protocol, leaving you to build this stuff yourself.
• Speed: less infrastructure means less latency.

I wonder whether FIX is more popular with sell side firms offering connectivity than the buy side. It seems from Max’s posts that he’s doing connectivity work for hedge funds building automated trading systems using the broker APIs he mentions. I’d be interested to know more about why those firms choose to use the proprietary APIs…

DynaTracing Caplin Trader

DynaTrace sounds pretty cool – gotta try running our Caplin Trader based GUI through it…

MicroMen

My first computer was a ZX81, so MicroMen on BBC4 was a powerful blast of nostalgia for the early 80s. Funny, sad and exhilarating – the Acorn hack against the clock to build the BBC prototype was fantastic, and the Clive Sinclair Mensa sequence is hysterical. I was an undergrad at Cambridge in the late eighties, so I was pleased to see them get loads of detail right: the Baron of Beef pub, the Computer lab. I used to see Sinclair jogging on the Madingley Road on the rare occasions I was out and about before midday…

I built my ZX81 from a kit. I later added all kinds of peripherals, including a real keyboard and 16K RAM pack. Another great point of detail in MicroMen was the reference to the dreaded RAM pack wobble that afflicted ZX81s. After the ZX81, I went on to a Camputers Lynx, also mentioned in MicroMen, and then an Atari 520 that I took to college in 86.

Visual Studio 2005 property bolding

VS dev trivia: if entries in your C++ project property pages show up as bold it means they have been given explicit overrides in the .vcproj XML. This broke my build today as the overrides stopped inherited values from .rules and .vsprops files being pulled in.

Joel on duct tape developers

As usual Joel hits the nail on the head in contrasting pragmatic, delivery focused “duct tape programmers” with perfection obsessed abstraction astronauts who never ship because they can’t stop polishing and honing. It’s those duct tape devs we’re always looking to hire for front office trading systems, cos time to market is everything in our game…

Web Workers

One of the hard restrictions that JavaScript GUIs like Caplin Trader run up against is the single threadedness of the browser’s JavaScript engine. Organising the asynchronous execution of code to service both user interface and network events in a responsive and timely manner from a Comet server is fiendishly difficult.

So I was heartened to read about Web Workers today. In future they would enable a GUI like Caplin Trader to handle network events on a background thread…

Wolfram Alpha: another nail in Bloomberg’s coffin ?

Wolphram Alpha is billed as a computational knowledge engine. My questions is: could it be another nail in Bloomberg’s coffin ?

Some time ago I blogged about how Google might provide some of what Bloomberg currently provides for traders. Another part of the picture that I didn’t mention then is analytics. In Bloomberg’s book he describes how they started out with market data in direct competition with Reuters, then added news and email. Another big value add was price histories, analytics and charting. Wolfram Alpha’s scope seems boundless, so a fortiori it could offer something similar. Certainly, it’s reported as offering charts in its results. To me, the crucial  questions are…

• Is financial data among the curated data sets being gathered for Wolfram Alpha ?
• Will Wolfram Alpha understand concepts like DV01, accrued interest, PE ratios, the greeks, moving averages ?
• Is Wolfram Alpha aiming to be a platform too ?
• Will it be scriptable ?  With Mathematica scripts ?
• Will it have social features to encourage the sharing of scripts and library building ?

Open Source Data Integration

Talend is claiming to be the first provider of open source data integration software. I suspect the OpenAdaptor guys may dispute that claim…

Xmas reading

Christmas is coming, and I’ve been figuring out my reading list for the break. I’ve pulled my old copy of Godel, Escher, Bach off the shelf, and have started the first re-read for 20 years. A couple of things hit me right off the bat: how could I have missed the GEB/EGB (Godel,Escher,Bach/Eternal Golden Braid) sequencing pun ?  And it seems obvious that GEB is a big influence on briarpig’s style. I’m also looking forward to reading Hasbrouck’s Empirical Market Microstructure. Looks like a natural successor to Harris’ Trading and Exchanges.
Thought for the day: to be is to be the value of a bound variable.