Capital Market CEP Fantasy Land
In Tech Spending Hit by Subprime Mess, Jeffery Schwartz says,
“According to Tabb, spending on development is being refocused on projects that can help firms improve their margins and, not surprisingly, do a better job at risk management. As such, investments in capabilities such as algorithmic trading and complex event processing (CEP) are likely to be pivotal in some firms’ efforts to become more competitive and improve their efforts at mitigating risks.”
“But for some banks that have deployed such technologies — the now-defunct Bear Stearns, Lehman Brothers, Citigroup and Merrill Lynch — the question is: How did these companies fail to mitigate the risks that have slammed their businesses if their development teams were developing and deploying sophisticated systems?
“There is definitely an awareness that perhaps the systems that existed in place to assess the value of portfolios or judge risk [are being scrutinized],” said Stevan Vidich, an industry architect in Microsoft’s financial services group. “
He added that there is strong interest in CEP and other risk management methodologies. A growing number of shops have started deploying such solutions based on the .NET Framework, Vidich said, and he believes such investments will continue.
“Clearly, there’s a lot of need to deal with the immense influx of data and being able to analyze data in a timely manner,” Vidich said. “It also drives need for systems like business intelligence, or BI, applied to a near-real-time scenario, which is a very attractive proposition.”
What are these guys on Wall Street smoking?
This is the precise “over hyping” problem I have warned about repeatedly. Folks selling rule engines that perform basic calculations over a time window of streaming data have been marketing their wares as “superbrains” that can solve very complicated problems and, at the same time, save Wall Street and The Planet.
Let me be perfectly clear here Wall Street. Listen very carefully.
There is nothing in any of the so called CEP products in the market place that is going to stop losses related to the subprime meltdown effecting the “now-defunct Bear Stearns, Lehman Brothers, Citigroup and Merrill Lynch,” as Jeffery Schwartz implies.
To imply that the risk management (and corporate governance) required to mitigate the current crisis on Wall Street can be foreseen, solved, or even mitigated, by a rules engine (or any software) is complete and absolute fantasy.
I think the fever created by the subprime flu is putting folks on Wall Street, or at least the vendors and the analysts pandering to them, in a Capital Market CEP Fantasy Land.
Filed under: Analytics, Business Activity Monitoring, Business Event Processing, Business Events, Business Rules, CEP News and Events, Complex Event Processing, Cyber-Trading Technologies, Cybersecurity, Event Processing, Event Stream Processing, Event-Driven Architecture, Predictive Business, Risk Management
Tim,
This is *precisely* why I said at that Wall Street conference about two months ago … “CEP is not a Golden Pill”.
Hi Marc,
Thanks for stopping by the blog.
Yes, lately when the weekly Google Alerts for “complex event processing” hit my inbox I read it as more of “the comics” versus anything of value.
Selling snake oil as the answer to all our problems is nothing new; but the degree of hyperbole and fantesy in the CEP market is becoming laughable.
Anyone who would bolt a CEP engine on their network and tell the CEO their problems are solved, or will be solved soon, does not understand what they are doing.
Much like SOA and component-based architectures have promised the moon and stars and barely delivered; CEP is going down the same path - a path of “jump-on-the-bandwagon” technology worship characterized by false promises and hopes.
Yours sincerely, Tim
Bingo.
If the underlying financial models don’t get the financial risk framework right, all that complex event processing can do is to help implement a fundamentally poor framework.
Portfolio managers that invested in the subprime market very likely outperformed other, more prudent managers with similar guidelines through 2006. And these revenue generators likely took home paychecks that many of us could retire on. The cost center of risk management activities usually carries far less clout in such environments.
We Monday morning quarterbacks of the subprime meltdown can see a very simple attribution.
• Subprimes yielded more than the less credit-risky alternatives
• Any reasonable market participant understood the sensitivity of these instruments to home price appreciation (or depreciation)
• While much work went into the financial modeling of defaults under various scenarios, no empirical data existed for a previous experience of significant home price depreciation
• A significant amount of the data describing borrower characteristics was fradulent. While there were certainly anecdotal evidence of this pre 2007, any financial decisions to avoid would have been based on forward-looking intuition, not algorithms based on empirical data.
• Finally, holders of pooled subprime mortgage securities almost certainly did not give anywhere near the weight to the complete evaporation of liquidity. Subprime prices have been absolutely clobbered, not only by the quantum leaps in the expectations for defaults, but also by the defacto market demands for much higher returns (vulture-like behavior), even after the quantum leaps of default expectations have been estimated. Market partipants as a whole completely missed that one.
Now how are the technology guys gonna stop that???
[...] vendors and software! Note: I address this topic back almost exactly one year ago in this post, Capital Market CEP Fantasy Land. Share and [...]