Most Wall Street firms' desktop computers run on a version of the Microsoft Windows operating system. Critical staff, from chief financial officers to traders, have long used Microsoft's Excel spreadsheet application to analyze market and transaction data, build financial models and track results.

But the software giant wants to move securities firms beyond viewing it principally as a provider of office productivity software and number-crunching worksheets. The firm has bolstered Excel so it can tackle at least 100 million rows of data at one time, and it has enabled securities firms' sophisticated analysis and trading applications to preside and run on its Azure cloud-computing service. The moves are examples of how the Redmond, Wash., firm is priming itself to be part of the digital foundation for the financial industry's moves toward consolidating technology systems and the flow of work between business units.

In fact, Microsoft's efforts to move serious number crunching to the desktop may finally impress some of the biggest naysayers of off-the-shelf software: traders. They have traditionally customized their technology products to digest ever-growing data streams, and build the best trading models and risk-management tools.

John Jacobs, chief operations officer at Lime Brokerage, which caters to high-frequency-trading firms, says high-speed traders are especially apt to develop their own risk-management software, in part because what constitutes risk remains open to interpretation. "Risk management is really the only way to manage the large amount of order and trade-flow activity generated by a high-frequency trader," Jacobs said.

Traders have necessarily become adept programmers in Excel, given the ubiquity of the spreadsheet application and the increasingly analytical nature of their jobs in the data-driven financial markets. Microsoft's bolstering of spreadsheet capacity over the last few years to a million rows of data came in direct response to Wall Street requests, the company has said.

"On every single desk, especially those with quantitative analysts doing modeling, Excel is one tool that is always used," said Lloyd Altman, a senior director in the capital markets practice of Accenture, one of Microsoft's official "partners" that helps tailor software vendor's general applications to clients' needs.

Altman noted Excel strengths, such as its graphical front end, its built-in tools to perform regression analysis and other analytics as well as its ability to pull in and use third-party calculation tools, such as Fincad, which develops derivatives pricing models.

But at a certain point, he says, Excel is no longer enough. Those traders typically have to migrate their work to more robust platforms, especially when performing data-intensive risk analysis or they want extreme speed in trading.

"Algorithms don't want a human involved," said Altman, adding, "You might want Excel to create the algos, but you don't want to use it to deploy them, since it would just slow you down."

Starting May 12, Microsoft began offering business customers the 2010 version of Microsoft Office that includes a more powerful version of Excel. It can be turbocharged by a new application called PowerPivot, launched at the same time, that effectively bolsters Excel's capacity to 100million rows or more of data.

This could change the nature of developing trading models. In the past, trading prioritizing speed has required specially designed software and hardware to send orders to market centers at speeds in which microseconds determine whether a trade results in a profit or loss. But the analysis fueling each order adds to trade latency as well. With 100 million rows of data, Microsoft could score some big points with traders by enabling them to develop more-complex trading and risk models that tap a vast store of data-all in Excel.

"To deal with speed in high-frequency trading, you have to keep all the data in memory for you to even have a chance to do calculations fast enough," said Matt Meinel, vice president of market development at 29West, a high-speed messaging provider acquired by data-integration specialist Informatica in March.

Meinel calls the hundred million rows of data enabled by PowerPivot "key" to supporting advances in high-frequency trading, if not the actual trading itself.

"Basically, Microsoft is providing a way for Excel to access a very large memory space" for data, Meinel says, adding, "That gives you a much faster response time ... and it gives you the ability to build much more complex models."

Meinel said he told Microsoft executives, while serving on an industry advisory council for Microsoft Office in 2001 while in a previous role at UBS, that half the world's capital was going through the Excel calculation engine. But he asserted that the product suffered too many failures in difficult calculations.

Microsoft has vastly improved Excel in the past decade, says Meinel. But financial services firms still tend to rely on platforms comprising hundreds or even thousands of servers to essentially create a giant spreadsheet to aggregate very large quantities of data and analyze it.

"You do all the slicing and dicing on that platform, and then you download it to Excel to view," he says.

Meinel says most traders can write spreadsheet macros, but at a certain point must pass off the model to a programming team.

With PowerPivot, Microsoft is enabling more of that work to be done at the trader's desktop, or the desktop of any user juggling large quantities of data.

Microsoft is also enabling traders and other finance executives to host their applications, such as trading or risk-analysis models, at the software firm's data centers, to boost their capacity and automatically benefit from technology upgrades.

This is what Microsoft's Windows Azure cloud-computing platform offers, said Ben Narey, director of Microsoft's banking and capital markets groups.

Azure, in production since the start of the year, enables users to develop software applications and have Microsoft host them. "We think by building these highly scalable data centers, we can actually improve on speed. Most companies are investing in this infrastructure themselves and so have certain limitations," Narey says.

New York-headquartered RiskMetrics Group, which analyzes risk for sell-side and buy-side financial firms, recently adopted Azure to gain database capacity, especially to support large bursts of activity in today's volatile markets.

Crunching clients' data in risk models such as Monte Carlo simulations requires huge capacity. Rather than investing in a system to handle burst volumes itself, RiskMetrics is instead devoting those resources to developing new services for customers, according to a case study by Microsoft and RiskMetrics.

Peter Redshaw, an analyst at Gartner Group who focuses on financial-institution outsourcing, says it's a misperception that Windows is typically relegated to the desktop level. He notes that financial services companies often run a variety of architectures such as Unix, Linux and Wintel-the latter combining Windows with Intel chips-in their data centers.

Redshaw says one of the biggest risk-management obstacles facing financial services firms is having the right data in the right place at the right time, especially because it comes from multiple sources, often in different formats.

A turbocharged Excel application alone doesn't address that issue. "Firms spend a fortune in time and effort trying to reconcile all this data," Redshaw says, adding, "They certainly need wider and deeper access to data, so something like PowerPivot will be a help. But it's not a silver bullet."

Redshaw also notes that financial firms are increasingly seeking to consolidate their technology platforms to run more applications on fewer servers and a broader array of financial instruments on fewer applications.

Vendors have reacted to that trend by including more functions in their applications, making it harder to categorize them. "The dividing lines between application categories are blurring-major vendors don't want to be excluded from any bidding list," Redshaw says.

Events such as Greece's recent credit crisis and the so-called flash crash May 6 only highlight the need for trading firms to maintain complex models, yet streamline the applications that crunch that data.

"An event like Greece may have a small impact in each part of a financial business, but in aggregate the impact can be much bigger," Redshaw says.

That mandate to view risk enterprise-wide may give Microsoft an edge going ahead. Its products such as Excel and Word are extremely well understood by users of desktop computers-a market Microsoft dominates.

And they are designed to work smoothly with other Microsoft business applications, including .Net, a set of tools to write and manage Web applications in multiple programming languages, and SharePoint, enabling workers in a company to more effectively collaborate online.

Microsoft's biggest challenge may well be overcoming its desktop-provider reputation. To that end, the firm has ramped up the number of employees dedicated to financial services to 600, up from just six eight years ago. Many of those hires have extensive financial industry experience.

For example, Microsoft hired Matthew Bienfang earlier this year to articulate to clients how the firm's more general applications can be adapted, often working with other vendors. Bienfang was a senior research director at consultancy TowerGroup and previously worked at broker-dealers as head of trading and brokerage operations.

Microsoft has worked with a number of very large financial services firms in recent years, including Raymond James, Bank of America and Credit Suisse.

Credit Suisse, for example, completed a four-year project in 2009 to streamline the technology platform and data used by its alternative investments group. The project consolidated numerous existing software programs built by the firm's IT department and trader-developed macros onto a single platform, and adopted a single workflow, rather than the many fragmented ones of the previous system.

Credit Suisse used the .Net tools and SharePoint to develop the platform, whose components now draw on a normalized pool of data.

Narey says Credit Suisse developed the platform to provide real-time computation of pre- and post-trade analytics, and to ensure the information used by its employees is accurate. The Swiss firm consolidated a hodge-podge of tools and data into a single work flow so that "corners couldn't be cut, and to provide greater transparency to management," Narey said

"The platform has really opened up new opportunities to fund managers, analysts and management to use the data in very timely and strategic ways to manage risk," he added.

Now, according to Narey, portfolio managers can see how the pre-trade analytics shaping their trading decisions are changing in real-time as their actual trades and market dynamics shape their portfolios.

PowerPivot's ability to access 100 million rows or more of data should be attractive "for anyone who needs to analyze lots and lots of information and make informed and real-time decisions," Nary says.

But the toughest nuts to crack are likely to be the trading firms who proudly develop trading-software solutions and are highly secretive to boot.

Take the chief technology officer at a Chicago-based proprietary trading firm, which specializes in fixed-income and related futures. He agrees that Excel's data limitations have caused headaches, so PowerPivot will likely be welcomed by anyone "who has to analyze millions of rows of data."

The bigger issue, the CTO says, is Excel's "poor performance when it comes to processing real-time data."

Office 2010 has greatly improved processing power, Narey contended "A lot of this comes down to multi-thread and multi-core workloads that better achieve the low latency and high-trading volumes financial firms are seeking right now," Narey said

Microsoft noted that Excel 2010 can now process larger sets of data, enabling risk calculations that normally take hours of processing time to "be processed in the back-end quickly and then uploaded instantly to a desktop."

The CTO also said the Azure cloud-computing platform may be suitable for more commoditized applications. But his firm would feel "uncomfortable" moving the most critical processing or data to a third party.

Microsoft responded that security measures are built into the platform, which offers a "99.9 percent uptime service-level agreement."

 

ONE-TWO PUNCH

Here's how Microsoft's upgrade of Office and cloud services are intended to benefit trading firms:

OFFICE 2010

Excel upgraded.

-Handles one million rows of data Redshaw says

-Handles 100 million rows of data with PowerPivot feature.

-Processing speed increased.

-Results integrate with Office functions such as picture rditing and Web applications such as SharePoint.


AZURE CLOUD

-Proprietary trading software can be hosted at an Azure data center.

-Data center automatically upgraded and maintained by Microsoft.

-Combined data-center capacity handles big bursts of messages.

-Security measures are built-in

-Service level agreement guarantees 99.9 percent uptime.

-24/7 service and support.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access