Pink Iguana

Home » Uncategorized » Check in with Sergey Aleynikov

Check in with Sergey Aleynikov


Jef Feeley, Bloomberg, Goldman Sachs Wins Fight Over Ex-Programmer’s Legal Fees, here. This is the High Frequency Equity Trading developer who was thought to have stolen super fast trading code written in Erlang. He gets off, but has to pay his own legal fees defending himself against the charges of stealing High Performance Erlang code.  It’s like he has to pay his legal fees for defending himself against stealing Euclidean Circle Squares. The weird part is nobody seems to know that Euclidean Circle Squares do not exist much the same way High Performance Erlang is an impossibility unless you are really  massively uninformed.

Aleynikov may now have to pay his own fees stemming from criminal charges and lawsuits over the trading codes, estimated to total about $7 million, including about $550,000 at issue in Delaware. Aleynikov’s federal theft conviction was overturned in 2012. Three years later, a judge threw out state-court charges he pilfered the bank’s intellectual property. Prosecutors are appealing.

Oscar Williams-Grut, Business Insider,Ex-Google engineers have spent 2 years secretly trying to solve ‘the hardest problem in banking’, here. Morgan Stanley still runs their Swaps Trading in APL As of 5 years ago, BofA needed old Next machines on Ebay to run their Credit trading business. The London Whale  failed to Calibrate the Gaussian Copula for CDX tranches running on a Dataflow implementation on an FPGA “Supercomputer.” Goldman loves high performance Erlang and Dave Korn’s kid runs the technology. You could not make up a less believable set of stories. These guys think that the banks’ problems will be solved by a new operating system. They are going to a gun fight with a plastic spoon.

Bank trading has way to much money to make it more or less independent of compute technology. Certainly, there are examples like 1980’s CMBS where technology enabled trading that could not otherwise be printed on the books.  But past a point, call it 1990, there are no real needs for more compute cycles. The trading businesses were so flushed with cash  and the easy prospect of making more profit that compute power became a non critical part of the evolution of trading. How non critical?  Firm risk can run in BigDecimal and not cause a single followup question. Swaps Trading can run in an APL  interpreter and no one laughs.  The compute tech is so bad because it is not in focus and is allowed to linger on the fringes of pet projects because it does not hit the profit. London Whale hit the profit but  there is no one  in any of the congressional testimony who can even formulate a tough question and follow up .

Taylor discovered the problem with core banking while helping a high street lender with the problem of automatically categorising transactions.

Warrick told BI: “They were looking at the banking infrastructure and they were just amazed. It was like having a computer science history lesson walking through the bank, but the weird thing was all of those systems were still in place and running, from 1970 onwards.”

He added: “Last time Paul had cracked what people thought was impossible: speech technology. Now he wanted to crack the biggest problem for the banking system. You need a proper banking operating system at the heart.”


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: