Last month, Aite Group published a report on IT spending at capital markets firms for 2009. Completed with the help of Wall Street & Technology magazine and based on surveys of 28 senior technology executives at a cross-section of capital markets firms, including hedge funds, retail brokerages firms, sell-side firms, private client groups, traditional asset managers and full-service firms, the report predicts an average technology budget cut of 5% for 2009.

Additionally, the report identifies cost reduction as the number one business objective for technology spending in the year ahead, followed closely by risk management. So with cost reduction at the top of the list, it's no wonder that the proverbial 'do more with less' is being bandied about in 2009 prediction lists, blog posts and press articles.

What does that mean for market data storage? For years now, data volumes have been growing exponentially and while there is no end in sight to these increases, there are financial and physical limits on a firm's ability to store this vital information especially given the current economic climate. To help Vhayu clients do more with the technology infrastructure they already have, we introduced a hybrid hardware/software solution, Velocity Squeezer that works seamlessly with our Velocity tick database and delivers at least four times compression of all data.

We have received positive feedback from clients who are using the Velocity Squeezer compression solution. One firm in North America has not only achieved the expected reduction in their TAQ data storage but is also experiencing faster simple query performance and now looking at cards for additional production servers. For more information about doing more with less in a market data storage context, please visit the Vhayu website or contact us.

Over the past two years, a trend has been emerging amongst Vhayu's large sell-side customer base to use Velocity as a platform to build customized transaction cost analysis (TCA) tools, both to measure their own trading performance and to provide intraday benchmarks to their many institutional clients who use broker-provided algorithms.

There are many reasons why we're seeing more and more customized TCA, but unbundling of executions from research and other broker-provided services previously paid by soft dollars stands out the most. So while the cost of execution services has dropped like a stone to commodity status, many brokers are cleverly deploying what I call "Ginzu data management technology" (they can slice and dice market data quickly and easily for three easy payments of $29.95! Wait, there's more!...I digress).

Well, actually, it's not so easy. Many of these benchmarks require the ability to process and analyze over 100 gigabytes of intraday quote and trade data (Level 1 and 2) and publish the results into dozens of different trading applications. In the past, brokers could furnish TCA reports to customers on a weekly basis and it was all good. But traders are heavily scrutinized on both sides of the fence. Many fund management firms have put additional pressure on their traders by measuring their performance on a daily basis, and have tied bonus criteria around it. It has forced funds of this sort to take control over TCA at a macro level instead of trusting broker-provided reports, which usually go lockstep with the algos they provide and can be viewed as skewed.

It all boils down to who has control over the data. As a result, we're seeing more large buy-side firms acting like brokers when it comes to data management, implementing raw market data feeds and market data distribution systems, and analyzing and storing all the data for internal purposes. To stay one step ahead and prove their worth, the sell-side is stepping up the pace at which information is available to trading partners, a shift that has gone from weekly to end of day to intraday to real-time.

I spoke in April with the head of execution services at one of our biggest customers in London about his appreciation of how Vhayu allows his desk to quickly build custom benchmarks and attach them to execution reports for customer dissemination. Prior to Vhayu, he said the reports were taking up to two hours after market close (severely cutting into pub time he noted) but now take less than 30 minutes (hooray beer!). He also said that more customers were using their own sophisticated broker metrics, so to stay ahead, they were going to move to real-time soon to keep the orders flowing (pub owners rejoice).

Well, I recently saw an article about how they have indeed moved to real-time with an official stating, "traditional post-trade analysis tools were not well integrated into most client's daily workflow and so could not add optimal value to trading decisions. Providing clients with real-time trade analysis will give them live, continuously updated information while orders are still active, enabling them to adjust orders in the pursuit of better execution." That's the party line at least. We know the real reason :)

As the human species, we tend to think of ourselves--the top of the food chain of carbon-based life--as the key actors in all the surrounds and everything involves us. We fancy ourselves masters of the planet, universe, and our fate. But in the financial markets today, this could not be further from reality.

Until recently, humans traded all commodities, currencies, precious everythings, stocks, and bonds. They made judgments on the value. Humans made judgments on the value of some thing and bought that thing with the objective of selling that thing for more than the acquisition cost. Human traders learned from experience, collected many forms of input that could have bearing on the value of something, and sought out others who valued what they had either more or less. Trading required a human energy that fundamentally underlies many social and economic structures.

Enter the computer. The first uses were as tools to help humans do a better job of collecting input, valuing some thing and finding others interested in trading. Then a seminal event occurred in the 80's when the first computer stopped only helping humans and started trading directly with other computers. For the first time no human energy was required to execute a trade. A new actor that traded in markets autonomously alongside human traders was born.

This new actor in the market made decisions based on the state of the market that it derived by processing market data. Humans wrote the new rules, but the actors executed them. These new actors were called algorithms. There are many case studies on the disasters and spectacular successes of the early algorithms. We will save those for another day. The truth is that the new actors are now responsible for trading over 50% of the volume in a number of markets and that percentage is growing rapidly.

The essence of this new actor was the silicon, the substrate of CPUs and memory. The new actor traded autonomously and its brain was based on the silicon of integrated circuits. It is reasonable to call it a silicon-based trader, and the real interesting part of the new perspective is to take the point of view that it is a new life form.

The silicon-based life form trader has two characteristics worth discussing. One is consciousness, the attribute of sensing and understanding one's surroundings. We could also call this situational awareness. The other is the evolution of this consciousness.

Perhaps we can best describe the new actor's intelligence as reptilian. A reptile is innately programmed to look for movement and snap at it because it may be either food (opportunity) or a threat. By way of analogy, this new actor had a reptilian level of intelligence and acted with reptilian-like responses. It was conscious of the market by ingesting market data and organizing that data to paint an electronic picture of the price and volume of an instrument. The algorithms then looked for some kind of movement in a price, for example, and then acted in a predetermined manner according to a set of rules.

These algorithms have evolved quickly in a significant direction. They are starting to develop and use memory to provide a more adaptive response to the change in the environment. In addition to reacting to changes in the market, the new actors are starting to vary their response based on patterns that they have seen before. This is completely consistent with the cognitive model that the human trader employs. The human trader looks for changes in the market and based on experience, wisdom, and intuition--all parts of the human trader's memory--the response to the movement will vary. The new actors, the silicon-based traders, are starting to utilize memory to adapt their response to changes in the market and relying less and less on the memory of the human trader.

A final thought on the evolution of the silicon-based life form trader. Evolution occurs through a process of natural selection of the most successful approaches. The successful approaches are further evolved in successive generations. One factor in the speed of evolution is time span of a generation. The shorter the time span, the faster changes are processed as either successful or unsuccessful. In the case of a carbon-based life form trader, a generation is approximately 30 years. In the case of the silicon-based life form trader, Moore's Law describes the doubling of processing speed every eighteen months. Although imperfect, a rough approximation is that silicon-based life form traders are evolving geometrically faster than their carbon-based counterparts. The unfathomable aspect of this is that the doubling of capabilities every eighteen months means the capabilities of the silicon-based trader compounds 20 times during a single generation of the carbon-based life form.

In summary, when we take the view that silicon-based traders remain living entities with a consciousness, an evolutionary past, and the capability for autonomous actions in the market, we dispel the notion that computers are merely tools in the hands of humans. This is the reality today, happening as we speak.