Chapter 02 – Greatest Hits of Computation in Finance



Computational Finance, Stock Market Analysis, and Investment Trading

“A computer does not substitute for judgment any more than a pencil substitutes for literacy. But writing without a pencil is no particular advantage.” – Robert McNamara

The Journal of Portfolio Management (JPM*) is one of the more upscale investment management publications around. For $500 a year, you get four issues, nicely bound like oversize paperbacks, without any advertising. It’s a crossover between rigorous academic publications, like the Journal of Finance, and trade magazines, like Wall Street & Technology, that have shorter staff-written articles and lots of ads. On significant anniversaries, JPM assembles special issues with invited pieces from both academics and practitioners on relevant topics.

JPM did this for the thirtieth anniversary issue in 2004. Some touched on the changes brought about in investment management by computation and where we can expect significant progress in the future, but only one was explicitly about technology in finance.(1) The editor invited me to expand on this topic for an upcoming issue. This chapter is based on that article.

It is illuminating to look back on past technological breakthroughs of the century to see which ideas have proven valuable, which have slipped into the zone of “further study,” and which will get you into trouble if mentioned in the polite company of financial professionals. I write this as a long-term self-confessed nerd on Wall Street, having witnessed a number of technological charges up various hills and a slightly smaller number of retreats back down.

Many of the greatest hits in this chapter existed in some form for a long time but were impractical because the roomful of mainframes needed for them was so large and expensive. The small number of people and firms with the resources to do these limited the market incentives for them to grow and succeed. Two basic technological laws, which are really more like notions on their way to concepts, are Moore’s and Metcalfe’s laws. Moore’s law is the well-known doubling of computational power every 18 months. Metcalfe’s law is the less well-known maxim that the utility of a network grows as the square of the number of users.(2) These are as close as we come to laws of nature for technology. The magnitude of the progress as we move out to the further reaches of Moore and Metcalfe space can’t be overstated. There’s more computational power on your desk than existed in the world 30 or 40 years ago. This may even be true for your wristwatch if you have a sufficiently fancy model.

There is no need to belabor the analogies about computer progress compared to automobiles and airplanes.(3) A novel way of looking at what we have at our disposal for technological innovation in finance is provided by Hans Moravec of the Carnegie Mellon Robotics Institute and author of Mind Children: The Future of Human and Robot Intelligence. He compares artificial and natural “computing machines” by storage capacity and processing speed. The book includes a remarkable chart that shows by his scoring that Deep Blue, the IBM machine that beat world champion Garry Kasparov, is edging up to the monkey zone on the evolutionary ladder. The ordinary machines used for more conventional tasks (like portfolio optimization) are comparable to insect brains.(4)

Financial Technology Stars

There are some clear nominees for the Financial Technology Hall of Fame, which have brought widespread, nearly universal, changes in the way people view and participate in markets. Topping my list are the following:

• Electronic market access
• Market data graphics
• Spreadsheets
• Databases and Internet information

For quantitative investors, I would add portfolio optimization despite the barriers to wider acceptance of this technology such as sensitivity to errors and unintuitive results (unless they are constrained).

Electronic Market Access and Electronic Trading Systems

Thirty years ago, the Designated Order Turnaround (DOT) and NASDAQ electronic systems were concepts on their way to notions. A so – called program trade literally involved wheelbarrows of paper trade tickets physically distributed to floor traders.

Today, trading floors from London to Tokyo have been replaced by machinery. Purely electronic markets, the electronic communication networks (ECNs), have come from nowhere to claim a significant portion of volume. Increasingly, the discussion of the future of exchanges is a discussion of technology. The August 16, 1999, cover of the Industry Standard, an information technology trade magazine, proclaimed “Stock Exchanges: RIP.” The photogenic trading floors, such the one at the New York Stock Exchange (Figure 2.1 ), are being closed, downsized, and relegated to use as media backdrops for financial reporters around the world.

Figure 2.1 Physical trading floors like the one at the New York Stock Exchange are an endangered species.

Figure 2.1 Physical trading floors like the one at the New York Stock Exchange are an endangered species.

The ongoing refinement of the techniques of electronic market access has also been one of the stealth success stories of the past decades and, as will be discussed throughout this book, is a likely prospect for future innovation.

Market Data Graphics and Stock Market Data Visualization

Throughout the early 1980s, market data terminals – almost all called Quotrons irrespective of their corporate origins – looked like computer screens. However, they were really keyboard-controlled cable television. They provided a picture of the data, the last trade, and the quote. It was the same information one would find by grabbing the end of a paper ticker tape and running it back to find the newest information for the stock of interest. This was literally what people did for a long time to get current quotes.

Figure 2.2 -- Market data systems: then and now.

Figure 2.2 -- Market data systems: then and now.

With the advent of true machine-readable market data, the text frontier, it was not long before these basic, simple text “pictures of data” were replaced by screens with an increasing variety of market charts showing both common and customized analytics. The tedious work once performed by analysts with pencils, calculators, and rulers essentially disappeared. (See Figure 2.2 .)

>>>>>> READ MORE HERE < <<<<<<

Notes:

* An earlier version of this article appeared in the Fall 2005 issue of the Journal of Portfolio Management (”If You Had Everything Computationally, Where Would You Put It Financially? Thirty Years of Computation in Finance”). Reprinted in “Nerds on Wall Street” with permission.

1. Sergio Focardi, Peter Kolm, and Frank Fabozzi, “New Kids on the Block,” Journal of Portfolio Management 30, no. 4 (Fall 2004): 42 – 53. The first article in the issue, by Andy Lo, “The Adaptive Market Hypothesis,” is an excellent in – depth discussion of the modern view of market efficiency.

2. Robert Metcalfe is the inventor of Ethernet (the ubiquitous wiring of the Internet), and the founder of 3Com. His law was demonstrated first with telephones and fax machines.

3. Okay, let’s belabor them a little for Generations X and Y and the Millennials, who may have missed these back when they were everywhere, but may also pick up this book. Gordon Moore, a founder of Intel, was fond of comparing the progress in computer technology to progress in automobiles. He used measures like mileage and cruising speed as analogous to memory size and CPU cycles per second. Every year, the hypothetical “car” corresponding to a current computer got more fabulous – traveling at 100,000 miles per hour, getting 100,000 miles per gallon, and the like. Auto industry people felt dissed by all this, and said, “Yes, but it would crash 12 times a day.” It’s tough to mix metaphors.

4. Hans Moravec, Mind Children: The Future of Human and Robot Intelligence (Cambridge, MA: Harvard University Press, 1988).

5. Steve Snider, Fidelity Pyramis quantitative portfolio manager, personal communication.

6. Emanuel Derman, My Life as a Quant: Reflections on Physics and Finance (Hoboken, NJ: John Wiley & Sons, 2007). This excellent book’s primary focus is on the use of quantitative methods applied to fixed income securities, but many of the insights have wider applicability.

7. “America’s Top 300 Money Managers,” Institutional Investor, July 2008.

8. Robert Schwartz and David Whitcomb, Transaction Costs and Institutional Investor Trading Strategies, Monograph 1988 – 23 (New York: New York University Stern School of Business, 1988).

9. Jonathan Stempel, “Citigroup Buys Automated Trading Desk,” Reuters, July 2, 2007.

10. Try the Wayback Machine for yourself at www.archive.org . Type in a URL and a date, and spin the dial. The live music archive, at the same URL, is also quite a find, especially for fans of the Grateful Dead.

11. For current information and a perspective on the evolution of electronic markets, see the Tabb Group at www.tabbgroup.com

12. A live version is available at http://marketrac.nyse.com/

13. There is a huge collection of past and current HCIL work at the lab’s site: www.cs.umd.edu/hcil/

14. The tree map is still enormously useful for its original purpose – tracking down those files that suddenly take over your disk. A free utility along these lines is Sequoia View, from the computer science department at Eindhove Technical University in the Netherlands ( www.win.tue.nl/sequoiaview/ ).

15. For more on this remarkable story, see Paving Wall Street: Experimental Economics and the Quest for the Perfect Market by Ross Miller (John Wiley & Sons, 2002).

16. See “Delving Deeper” by David Leinweber, Bloomberg Wealth Manager, October 2003. 17. The Harvard Business School e-Information project at www.people.hbs.edu/ptufano/einfo/ has a nice online collection of these studies.

18. Ser-Huang Poon and Clive Granger, “Practical Issues in Forecasting Volatility,” Financial Analysts Journal 61, no. 1 (2005): 45 – 55.

19. Bill Alpert, “Can Computerized Language Analysis Predict the Market?” Barron’s at WSJ.com, November 19, 2001, http://online.barrons.com/article/SB1005962163718064080.html

20. See sites www.sec.gov, www.newsml.org and www.rixml.org for the latest details on XML use at the SEC, in news, and in investment research.

Wall Street Analytics

Easily find all of the
lowest cost no load
mutual funds and ETFs
Available at major ebook stores:

Amazon -- Kindle/MOBI
index funds list investing books

Apple iBookstore -- iPad/EPUB
index mutual funds investment guide

Barnes & Noble -- Nook/EPUB
no load mutual funds investment guide

Smashwords -- EPUB, MOBI, PDF
low cost mutual funds investment guide

no load index mutual funds
by David Leinweber