Banking, Finance & Investing

(a subtopic of Applications)

Contents Good Places to Start Readings Online Related Web Sites Related Pages More Readings

"Money is just a type of information, a pattern that, once digitized, becomes subject to persistent programmatic hacking by the mathematically skilled. As the information of money swishes around the planet, it leaves in its wake a history of its flow, and if any of that complex flow can be anticipated, then the hacker who cracks the pattern will become a rich hacker." - from Cracking Wall Street

Good Places to Start

HAL 9000-Style Machines, Kubrick's Fantasy, Outwit Traders. By Jason Kelly. Bloomberg.com May 3, 2007). "Way up in a New York skyscraper, inside the headquarters of Lehman Brothers Holdings Inc., Michael Kearns is trying to teach a computer to do something other machines can't: think like a Wall Street trader. ... The programs they're writing are designed to sift through billions of trades and spot subtle patterns in world markets. Kearns, a computer scientist who has a doctorate from Harvard University, says the code is part of a dream he's been chasing for more than two decades: to imbue computers with artificial intelligence, or AI. ... A third of all U.S. stock trades in 2006 were driven by automatic programs, or algorithms, according to Boston-based consulting firm Aite Group LLC. By 2010, that figure will reach 50 percent, according to Aite. AI proponents say their time is at hand. Vasant Dhar, a former Morgan Stanley quant who teaches at New York University's Stern School of Business in Manhattan's Greenwich Village, is trying to program a computer to predict the ways in which unexpected events, such as the sudden death of an executive, might affect a company's stock price. Uptown, at Columbia University, computer science professor Kathleen McKeown says she imagines building an electronic Warren Buffett that would be able to answer just about any kind of investing question. ... One day, a subfield of AI known as machine learning, Kearns's specialty, may give computers the ability to develop their own smarts and extract rules from massive data sets. Another branch, called natural language processing, or NLP, holds out the prospect of software that can understand human language, read up on companies, listen to executives and distill what it learns into trading programs. ... Rex Macey, director of equity management at Wilmington Trust Corp. in Atlanta, says computers can mine data and see relationships that humans can't. Quantitative investing is on the rise, and that's bound to spur interest in AI, says Macey.... AI proponents are positioning themselves to become Wall Street's hyperquants."

A "Neural" Approach to the Market - This S&P portfolio uses a computer model that "learns" from its mistakes -- and has handily beaten its benchmark index. By Will Andrews. BusinessWeek Online (May 8, 2006). "For investors, experience is the best teacher -- even for a computer-driven stock-selection strategy. That's the basic approach of Standard & Poor's Neural Fair Value 25 portfolio, which employs the investment research outfit's proprietary quantitative stock ranking system. The Neural Fair Value (NFV) concept, which was created by Andre Archambault, S&P's director of quantitative strategies, starts with S&P's Fair Value stock valuation system, which uses earnings estimates and other metrics to determine whether stocks are trading above or below their fair value. The 'neural' part comes into play when Archambault's model, updated weekly, combs the 3,000 stocks in that group for the 25 names it thinks have superior price appreciation potential. ... The NFV approach, Archambault explains, is based on 'Neural Network' theory, an artificial intelligence concept that seeks to replicate the human brain's ability to learn from mistakes. ... [Q:] The Fair Value concept is familiar to many investors, but the part that makes this unique is the neural overlay. How does the artificial intelligence concept come into play? [A:] Neural nets are kind of like 'black boxes,' and they're being used in all kinds of industries. ..."

How a Computer Knows What Many Managers Don't. By Zubin Jelveh. The New York Times (July 9, 2006). "If movies like '2001: A Space Odyssey' and 'The Matrix' are any indication, humans are not comfortable with the idea of artificial intelligence controlling their fate. So why ever trust a computer model to run your investments? Because, in the real world, it seems to pay off. Many mutual funds that make their trades based on the recommendations of a proprietary computer model, known as quantitative or quant funds, have outperformed their benchmarks in the last three years. And investors have noticed."

Artificial intelligence applied heavily to picking stocks. By Charles Duhigg. The New York Times / available from The International Herald Tribune (November 23, 2006); appeared in The New York Times on November 24, 2006 (A Smarter Computer to Pick Stock). "Studies estimate that a third of all stock trades in the United States were driven by automatic algorithms last year, contributing to an explosion in stock market activity. Between 1995 and 2005, the average daily volume of shares traded on the New York Stock Exchange increased to 1.6 billion from 346 million. ... [I]nvestment firms have increasingly begun exploring mathematics' furthest edges and turning to people like [Ray] Kurzweil, who became an expert in pattern recognition while he was building a reading machine for the blind. ... Wall Street has rushed to mimic the techniques. Because arbitrage opportunities disappear so quickly now, neural networks have emerged that can consider thousands of scenarios at once."

Combating money laundering with IT - Artificial Intelligence helps banks and institutions detect anti money-laundering by providing enterprise-wide transactional analysis that analyse multiple situations, detect unusual patterns in the data and flag-off suspicious transactions. By Dr Kaustubh Chokshi. domain-B. "KPMG, a leading management consulting firm, has estimated that funds worth anywhere from $590 billion up to a staggering $1.5 trillion are laundered annually through the global economy, amounting to 2-5 per cent of global GDP. Closer home, estimates on the amount of black money in the Indian economy vary from 20 per cent to 40 per cent of India's total GDP. ... To counter money-laundering operations, banks and financial institutions need solutions that take an enterprise-wide view of transactional analysis. These solutions must have the ability to analyse multiple situations, detect unusual patterns in the data and flag off suspicious transactions when things seem amiss.In recent times, solutions powered by artificial intelligence (AI) technologies have emerged as a powerful and effective option for rising above the shortcomings of traditional AML solutions. AI-based AML solutions are far more effective, as they use techniques like Bayesian inferencing and neural networks, which learn by example and experience in a non-linear mode (similar to how the human brain works), rather than merely being programmed to rigidly perform specific tasks."

Almost Human? Artificial Intelligence is back in the hearts and minds of technology gurus. By Robert J. Derocher. Insight (The Magazine of the Illinois CPA Society, September 2001). "In their tutorial, 'Introduction to Artificial Intelligence and Expert Systems,' Brown and Daniel O'Leary, an accounting professor and AI expert at the University of Southern California, say that AI, from an intelligence perspective, is 'making machines 'intelligent' -- acting as we would expect people to act.' From a business perspective, it is 'a set of very powerful tools and methodologies for using those tools to solve business problems.'"

• Access their tutorial via our Expert Systems page.

Smart Tools - Companies in health care, finance, and retailing are using artificial-intelligence systems to filter huge amounts of data and identify suspicious transactions. By Otis Port, with Michael Arndt and John Carey. Business Week's 2003 edition of The BusinessWeek50. "Banks, brokerages, and insurance companies have been relying on various AI tools for two decades. One variety, called a neural network, has become the standard for detecting credit-card fraud. Since 1992, neural nets have slashed such incidents by 70% or more...."

Artificial intelligence and the future of finance - Artificial intelligent agents are about to move beyond the realms of sci-fi horror and into the finance sector. Comment by Graham Whitehead, principal consultant, BT Exact. finextra.com (February 9, 2005). "What we all need as we journey down the Information Super Highway is help. This is at hand, in the form of agents - artificial intelligent agents rather than the James Bond variety. Just like the secretary or personal assistant of days gone by, these new agents will sort, sift and digest the mountains of information that will surround consumers. ... These agents offer great potential for the finance sector. They can help consumers source the best financial options, allowing them to manage their money more efficiently. Equipped with details of present financial products such as interest rates of savings accounts or mortgages or loans, agents will be able to identify the products that best meet a user's requirements."

How bots can earn more than you - Software robots can already outperform people on the stock markets, and that is just the beginning. By Duncan Graham-Rowe. New Scientist article preview (August 20, 2005; Issue 2513: subscription req'd.). "One morning this month, David Pardoe earned himself $4.7 million without lifting a finger. All the hard work was done by a robot. True, it was a robot without a body - a software robot, in fact - but almost a century after the word 'robot' was coined, the vision of automaton slaves is at last becoming reality. Software robots - also known as bots or software agents - can earn hard cash in the real world. They can even outperform people in some tasks, so it makes sense to let them do all the hard work."

Mimicking fraudsters - If your card use has been queried, it's probably because more banks are now using artificial intelligence software to try to detect fraud. By Ken Young. The Guardian (September 9, 2004). "Credit card fraud losses in the UK fell for the first time in nearly a decade last year, by more than 5% to £402.4m, according to research by the Association of Payment Clearing Services (Apacs). The fall has put a spotlight on the increasing use of neural networks that have the ability to detect fraudulent behaviour by analysing transactions and alerting staff to suspicious activity. As commercial applications of research into artificial intelligence, these systems give the impression of mimicking human abilities for recognising unusual activity. Karina Purang, a financial analyst at Datamonitor in London, says the use of neural networks is growing: 'These systems are very important to banks trying to reduce fraud, and are becoming standard across the card industry to detect unusual spending patterns.' She says Barclays reported that after installing Fair Isaac's Falcon Fraud Manager system in 1997, fraud was reduced by 30% by 2003. The bank attributed this mainly to the new system. ... Nick Sandall, head of retail banking at Deloitte, says that banks also use other technologies. 'The artificial intelligence community is constantly bringing us new solutions. ...'"

Locallending company soars after flying "under the radar." By CydneyGillis. The Seattle Times (October 6, 2004). "Layne Sapp startedin 1984 as a private mortgage broker. This year, his company expectsto provide $3.1 billion in mortgage loans and bring in $120 millionin revenue. ... Since 1995, Sapp said, MILA has invested $50 millionin a software system that relies on artificial intelligence to automatethe underwriting process. Mortgage brokers use the program to entera borrower's income, length of employment and other financial data-- up to 250 parameters in all -- to get a commitment on a loan within10 seconds. ... The software not only finds a loan specific to theborrower's needs, Sapp said, but tells the mortgage broker what documentsthe borrower will need to present, how to present the loan paperworkand how long the process will take. That removes surprises for boththe borrower and the broker, ensuring MILA gets repeat business,he said. 'The mortgage business is very painful,' Sapp said. 'Ifyou can deliver accurate answers up front, the loan officer dealingwith the customer is now more effective.'"

Brokers Will Spend Big on Anti-Money Laundering. By Jessica Pallay. Wall Street & Technology (May 1, 2003). "The brokerage industry will spend almost $700 million in the next three years on anti-money-laundering technologies, according to a recent report by Massachusetts-based consultants, TowerGroup. The 2001 USA Patriot Act requires financial institutions to establish anti-money laundering programs. ... Complex solutions include technology systems that offer artificial intelligence, [Robert Iati] says, using rules-based analysis, such as Mantas or Searchspace. For example, if an investor suddenly changes investing behavior, and that investor uses a bank that has been known to transact terrorist funds, the technology would post an alert for the situation to be investigated."

Can technology build a better Buffett? Valuable lessons from artificial intelligence, investors. By Carla Fried. Business 2.0 (February 13, 2004) / available from CNN.com. "If ever there were a field in which machine intelligence seemed destined to replace human brainpower, the stock market would have to be it. Investing is the ultimate numbers game, after all, and when it comes to crunching numbers, silicon beats gray matter every time. Nevertheless, the world has yet to see anything like a Wall Street version of Deep Blue, the artificially intelligent machine that defeated chess grand master Gary Kasparov in 1997. Far from it, in fact: When artificial-intelligence-enhanced investment funds made their debut a decade or so ago, they generated plenty of media fanfare but only uneven results. Today those early adopters of AI, like Fidelity Investments and Batterymarch Financial, refuse to even talk about the technology. Still, artificial intelligence has steadily improved in sophistication and quietly made itself indispensable on Wall Street. According to Andrew Lo, director of the Laboratory for Financial Engineering at MIT, every investment firm embracing a math-driven strategy uses some form of AI in its research, and Lo expects the terminology to appear again soon in promotions for retail investments like mutual funds and privately managed accounts. Before the hype machine cranks up this time, however, it would be smart to figure out just what AI can and can't do for investors."

Robots beat humans in trading battle. The BBC (August 8, 2001). "In the first ever test of its kind, a team of robots has beaten humans in simulated financial trading. Computer giant IBM pitted robotic trading agents, known as 'bots', against humans in trading commodities such as pork bellies and gold."

WARREN. From the Software Agents Group at Carnegie Mellon University. "The WARREN system builds on current investment practices to deploy a number of different, semi-autonomous software agents. These heterogeneous agents acquire information from and monitor changes to stock reporting databases, interpret stock information, predict the near future of an investment, and track and filter relevant financial news articles. The WARREN system is designed to monitor the ongoing portfolio mangagement process, and thus to function under conditions of extreme uncertainty." [P.S. Yes, it really is named for Warren Buffet!]

Computers Are Learning The Business - Advances in computer processing power open the way for wider use of so-called artificial intelligence, at the same time that the self-serve aspect of online processes has increased the need for systems that "think." By John Hackett. Bank Technology News. (April 2001).

Automated Mortgages. Get fast loan approval. Gary R. Crum. Special to The Christian Science Monitor. (July 3, 2000) "Chances are, when you apply for a mortgage, your loan will be underwritten immediately by a computer. While automated underwriting will not solve every loan problem, it certainly makes buying a home faster, better, and cheaper for most homebuyers."

Terminator technology predicts markets. Reuters News Service; available from Forbes.com. (April 29, 2001). "'Financial markets are not efficient. Markets are volatile and do not move smoothly but exhibit catastrophic jumps from time to time. They behave in a chaotic fashion,' [Dave] Jubb told Reuters. Indeed, volatility has been the norm rather than the exception for stock markets in recent years. ... According to Jubb, artificial intelligence tools are better at analysing these market movements than traditional techniques. 'Conventional quantitative methods can't model these complex relationships. Neural nets can learn non-linear relationships in the markets,' he said."

Readings Online

Monitoring NASDAQ for Potential Insider Trading and Fraud. AAAI Press Release (September 17, 2003). "NASD has developed an intelligent surveillance application -- the Securities Observation, News Analysis and Regulation (SONAR) system -- that automatically monitors the NASDAQ, OTC, and futures markets for suspicious patterns. ... SONAR includes several AI techniques, such as data mining, natural language processing for text mining, intelligent software agents, rule-based inference, and knowledge-based data representation."

SciFinance: A Program Synthesis Tool for Financial Modeling. By Robert L. Akers, Ion Bica, Elaine Kant, Curt Randall, and Robert L. Young. AI Magazine 22(2): Summer 2001, 27-42. This paper is based on the authors' presentation at the Twelfth Innovative Applications of Artificial Intelligence Conference (IAAI-2000). "The SciFinance software synthesis system, licensed to major investment banks, automates programming for financial risk-management activities-- from algorithms research to production pricing to risk control."

UBS Investment Bank Launches Automated Mortgage Underwriting Engine. Press release available from Business Wire via Yahoo! Finance (November 10, 2005). "UBS announced today that its Investment Bank has launched a new Automated Underwriting Engine (AUE) for mortgage loans on its Whole Loan Conduit platform.Using a sophisticated artificial intelligence system, the UBS AUE is capable of instantly analyzing a mortgage loan application to provide correspondent loan sellers with a consistent and reliable underwriting decision. The engine allows increased flexibility in the analysis of loans, taking into account loan characteristics that compensate for minor shortcomings."

Making Brain Waves. By E.B. Baatz. CIO Magazine. January 15, 1995. "Neural network technology has saved member companies of MasterCard more than $50 million."

Shouting fades out as software moves in. Barron's / available from The Financial Post (October 25, 2004; subscription req'd.). "Calm, cool and comparatively colourless, electronic trading has all but extinguished the open-outcry mayhem of U.S. futures exchanges. About a third of screen-based transactions rely on artificial intelligence, once the stuff of science fiction, and now a popular tool in Wall Street's expanding arsenal. As one financial-futures broker describes it, a computer, not unlike the mythical HAL of 2001: A Space Odyssey fame, can 'see the algorithms and execute the trade faster than most people can point and click.' ... Computer-trading technologies range from traditional 'black box' programs based on entry and exit prices to advanced programs that learn from their own mistakes. And quite often, it is difficult for other market participants to tell whether a person or a machine has entered a trade. Strategy Runner and Black Box Development both have developed advanced artificial-intelligence trading platforms for the futures market."

Man vs. machine. By Jeff Benjamin. Investment News (December 3, 2007). "When computers manage the portfolio, the first challenge is often preventing the human brain from getting in the way. This has become the mantra of Shashi Mehrotra, chief investment officer at Legend Advisory Corp., an asset-management firm based in Palm Beach Gardens, Fla., which hands over responsibility for more than $2 billion in investor assets to an artificial-intelligence computer program. ... 'I used to second-guess her, but I was wrong nine out of 10 times,' he said. The use of artificial intelligence, also known as neural networks or genetic algorithms, has been described by some as the second generation of quantitative investing because it has the flexibility to get smarter through the expansion of input data."

Duo-Mining -Combining Data and Text Mining. By Guy Creese. DMReview.com (September 16, 2004). "As standalone capabilities, the pattern-finding technologies of data mining and text mining have been around for years. However, it is only recently that enterprises have started to use the two in tandem - and have discovered that it is a combination that is worth more than the sum of its parts. First of all, what are data mining and text mining? They are similar in that they both 'mine' large amounts of data, looking for meaningful patterns. However, what they analyze is quite different. ... Collections and recovery departments in banks and credit card companies have used duo-mining to good effect. Using data mining to look at repayment trends, these enterprises have a good idea on who is going to default on a loan, for example. When logs from the collection agents are added to the mix, the understanding gets even better. For example, text mining can understand the difference in intent between, 'I will pay,' 'I won't pay,' 'I paid' and generate a propensity to pay score - which, in turn, can be data mined. To take another example, if a customer says, 'I can't pay because a tree fell on my house;' all of a sudden it is clear that it's not a 'bad' delinquency - but rather a sales opportunity for a home loan."

"Trust me, I'm an expert. Would you allow a machine to play the stock market with your life savings? Well, the people in the know are doing just that." By Clive Davidson. New Scientist (December 6, 1997).

Tracking money trails with technology. By Sandeep Junnarkar. CNET News.com (November 20, 2001). "While it may be impossible to spot all questionable financial activity, smaller measures can be taken to assist banks in mining financial data, according to Konrad Feldman, the chief executive of Searchspace, a company backed by HSBC Bank that develops and uses artificial-intelligence software. ... CNET News.com recently talked to Feldman about the pitfalls and strengths of technology in the investigation and about the accompanying concerns regarding consumer privacy."

The Penn-Lehman Automated Trading Project. By Michael Kearns and Luis Ortiz, University of Pennsylvania. IEEE Intelligent Systems (November / December 2003). "The PLAT Project has developed a trading simulation that merges automated clients with real-time, real-world stock market data."

• Check out the project site.

Cracking Wall Street. "Suppose you could discern market trends, speed up time to see where those trends were going, then bet on what you discovered. Geeks in suits are doing that today." By Kevin Kelly. Wired (July 1994; Issue 2.07).

AI Fights Money Laundering. By Jason Kingdon. IEEE Intelligent Systems (May/June 2004). "Artificial-intelligence-based software from Searchspace monitors banking customer activity to identify unusual behavior and detect potential money-laundering situations."

Tech, and the Future of Finance - Futurist James Canton offers predictions on how technology will impact CFOs in 2003 and beyond. By Marie Leone. CFO.com (December 13, 2002). "CFO.com: When will AI-based decision support systems hit the mainstream? Canton: Within five years we'll witness the rise of the neural net, genetic algorithm, and expert systems that provide advice for CFOs and treasurers -- such as what is the best play to make for an overnight investment. The systems will create 'expert behavior' rules from massive databases that are filled with previous transaction data and outcomes. Eventually CFOs will use financial software agents to 'clone' their expertise for true multi-tasking."

Asian Investors Seek Profit in Neural `Karma'. Commentary by Andy Mukherjee. Bloomberg News (March 23, 2004). "Using Paradigm's Forex DayTrader, which predicts movements in major currencies over a 24-hour time frame, the punter made a $46,000 profit in two days. ... DayTrader is one of more than 100 trading systems based on so-called neural networks that are supposed to mimic the way billions of brain cells work together to recognize patterns in complex data. Researchers have tried to replicate the human brain's neural circuitry in activities such as predicting energy prices and measuring creditworthiness. Unlike conventional software, systems based on neural networks aren't limited by their programmers' abilities. They learn better ways to analyze data as more information comes along. U.K.-based Retail Decisions uses neural networks to help online retailers prevent payment fraud. For two decades, researchers at universities in Britain and France have tried to build the perfect 'neural nose' that can discern smells. Such a system could alert the authorities to gas leaks, or warn retailers about foodstuff turning stale. Neural networks started appearing in the financial industry in the 1980s."

Zippy agents going for brokers. By Stephen Pritchard. FT.com / FT-IT review (July 13, 2005). "Researchers at HP’s European labs in Bristol, England have found that international financial institutions are increasingly showing interest in their work on automated trading agents - despite the fact that the agents were not originally developed for financial markets. HP Labs’ complex adaptive systems group first started working on trading algorithms in the mid-1990s. But they were originally developed to help large companies allocate computing resources in data centres. ... The Zip algorithm works by calculating the best trading strategy for continuous double auctions (CDAs), the trading basis of most financial markets. ... Zip traders have the ability to 'learn' from their actions, using simple machine learning rules. This function allows the trading algorithms to improve their own behaviour. As a result, Zip algorithms succeed in trading where zero intelligence algorithms fail."

• Additional information about the ZIP autonomous adaptive trading agent algorithm:
  • Archived version of Dave Cliff's ZIP website from his days at Hewlett-Packard Labs.
    • a May 2007 update about ZIP60 with links to papers.
  • News Articles:
    • ZIP's collection
    • Gordon Gecko makes way for trading software. By Robert Matthews. New Scientist (Issue 2606: pages 42 - 45; issue date: June 2, 2007; subscription req'd).
  • Syritta Algorithmics Ltd, a trading technology company.

Credit Card Companies Turn To Artificial Intelligence. By Margaret Webb Pressler. The Washington Post / available from the Tampa Tribune (September 29, 2002). "With billions of dollars at stake, and more clever crooks, credit card companies have become very smart about protecting themselves with astonishingly sophisticated network computers and software programs. 'We're at a level whereby we can understand with artificial intelligence ... the potentially fraudulent transactions,' said Raf Sorrentino, vice president of risk management for First Data Corp., one of the biggest providers of credit card processing and payment services. Credit card fraud costs the industry about a billion dollars a year, or 7 cents out of every $100 spent on plastic. But that is down significantly from its peak about a decade ago, Sorrentino says, in large part because of powerful technology that can recognize unusual spending patterns."

The Financial Crimes Enforcement Network AI System (FAIS). Identifying Potential Money Laundering from Reports of Large Cash Transactions. By Ted E. Senator, Henry G. Goldberg, Jerry Wooton, Matthew A. Cottini, A. F. Umar Khan, Christina D. Klinger, Winston M. Llamas, Michael P. Marrone, and Raphael W. H. Wong. AI Magazine 16(4): Winter 1995, 21-39. "The Financial Crimes Enforcement Network (FIN-CEN) AI system (FAIS) links and evaluates reports of large cash transactions to identify potential money laundering. The objective of FAIS is to discover previously unknown, potentially high-value leads for possible investigation. FAIS integrates intelligent human and software agents in a cooperative discovery task on a very large data space. It is a complex system incorporating several aspects of AI technology, including rule-based reasoning and a blackboard. FAIS consists of an underlying database (that functions as a black-board), a graphic user interface, and several preprocessing and analysis modules. FAIS has been in operation at FINCEN since March 1993; a dedicated group of analysts process approximately 200,000 transactions a week, during which time over 400 investigative support reports corresponding to over $1 billion in potential laundered funds were developed."

Branch 'Bots' Make Banking Better. By Jerry Silva. Bank Systems & Technology (September 28, 2004). "The next time you walk into a bank branch, look up and smile ... there may be someone (or something) looking back and watching your every move. No, it's not the normal security cameras located over the teller window. These new robot eyes track your every movement, from the moment you come through the doors, walk around the lobby and conduct your business, until you leave. In fact, there may be many such robots in the branch, assisting you and the branch during your visit to ensure that the branch is operating as effectively as possible."

Countrywide Loan-Underwriting Expert System. By Houman Talebzadeh, Sanda Mandutianu, and Christian F. Winner. AI Magazine 16(1): Spring 1995, 51-64. Countrywide loan-underwriting expert system (CLUES) is an advanced, automated mortgage-underwriting rule-based expert system. The system was developed to increase production capacity and productivity, improve the consistency of underwriting, and reduce the cost of originating a loan. The system receives selected information from the loan application, credit report, and appraisal. It then decides whether the loan should be approved or whether it requires further review by a human underwriter. If the system approves the loan, no further review is required, and the application is funded. CLUES has been in operation since February 1993 and is currently processing more than 8500 loans each month in over 300 decentralized branches around the country.

Plain talk by machine could give investors unbiased help. By Jon Van. Chicago Tribune (May 8, 2004; no fee reg. req'd.). "You could do worse than to seek financial advice from a machine. A lot of people have. Sageworks Inc., a company based in Raleigh, N.C., uses artificial intelligence to create reports in plain English that describe a company's financial performance. It does this by taking numbers from the firm's balance sheet and financial statements and converting them to a narrative. 'Our reports don't recommend buying or selling stocks,' said Brian Hamilton, Sageworks chief executive. 'We just present an objective analysis that's easy to understand.' The computers tend to ignore Wall Street's various enthusiasms."

Related Web Sites

Advanced Financial Systems Research specialises in state of the art technologies and artificial intelligence techniques applied to finance. This includes natural language processing and expert systems as tool for analysing qualitative information in finance and producing investment decisions, fuzzy logic, neural networks and genetic algorithms applied to finance.

• "This page contains a list of links suggested by Advanced Financial Systems Research in the fields of artificial intelligence and natural language processing applied to finance."
• another list from Advanced Financial

AI in Finance and Investment Gems. "This WWW page, maintained by Robert R. Trippi at the UCSD Graduate School of International Relations and Pacific Studies, is intended to provide short cuts for visiting places of interest to persons involved in applying AI to finance and investment problems."

AIStockBot. According to the SourceForge project summary, "AIStockBot aims to become the greatest Technical and Fundamental Stock Analysis program written. Using brute-force Artificial Intelligence, it should pick stocks better than you. We focus on security/market analysis, indicators, statistics, human mind..."

FatKat. "Quant investing is a significant phenomenon. It is applied today to an estimated trillion dollars of market funds. In the near future, Quant technique are expected to play a major role across the broad scope of capital markets (a 20 trillion dollar arena). FatKat's goal is to combine the best in financial knowledge with the best in mathematics and computer science to create a self-sustaining Quant system. Pattern recognition techniques (the hallmark technology of Ray Kurzweil firms) play a central role in finding predictability in the markets of today and the future." - from the Overview

Heuristics and artificial intelligence in finance and investment. Maintained by Franco Busetti. As part of the of topics that include neural networks, genetic algorithms, simulated annealing, the site provides wonderful cartoons and links to lots of software.

Mavent Expert System: "an automated system that collects and processes electronic loan data against the various Rule Libraries available in the Mavent Rule Libraries."

VIP - Advisor: an European IST project.

• Also see this related article: Your virtual assistant for personal financial advice. IST Results May 10, 2005). "Added usability and intelligence has been brought to virtual assistants thanks to technology developed by European researchers, offering online users an entertaining, yet competent professional financial service. 'The interaction of natural language, online translation, 3D-avatar technology and artifical intelligence creates a powerful instrument that will find a wide acceptance among users,' says Ulrich Thiel at Germany’s Fraunhofer Integrated Publication and Information Systems Institute (IPSI) and partner in the European IST project VIP ADVISOR."

Related AI Topics Pages

• Expert Systems
• Fraud Detection & Prevention
• Industry Statistics
• Machine Learning

More Readings

_____. 1990. Expert Systems in the Insurance Industry : 1990 Survey Report. New York: Coopers & Lybrand.

Adhami, Ebby, Michael Thornley, and Malcolm McKenzie. 1992. Pharos - The Single European Market Adviser. In Proceedings of the Fourth Annual Conference on Innovative Applications of Artificial Intelligence, 109 - 126. Menlo Park, Calif.: AAAI Press. "The formation of the single European market (SEM) will create a new business environment in Europe. The competitiveness and, indeed, the survival of many United Kingdom businesses depends on how well they understand and react to the threats and opportunities presented by opening up Europe and associated industry restructuring. Expertise in single-market issues and legislation is scarce and expensive, making it difficult for many organizations to obtain. In addition, the recession has prompted many organizations, particularly the smaller ones, to concentrate their efforts on improving short-term profitability. Most of these organizations cannot afford the resources needed to assess the impact of SEM on their business. This chapter describes pharos, an expert system designed to assess the impact of SEM legislation on businesses in the United Kingdom. pharos was developed by National Westminster Bank (NatWest) and Ernst & Young Management Consultants. It will be used by 70,000 medium-sized businesses, resulting in millions of pounds of savings for the United Kingdom business community yet offering the bank a competitive advantage. This section discusses the importance of SEM. This importance is assessed in relation to business in the United Kingdom in general and NatWest in particular. How pharos was conceived is also discussed."

Bynum, Sue, Robert Noble, Cheri Todd, and Ben Bloom. 1995. The GE Compliance Checker: A Generic Tool for Asessing Mortgage Loan Resale Requirements. In Proceedings of the Seventh Conference on Innovative Applications of Artificial Intelligence, 29 - 40. Menlo Park, Calif.: AAAI Press. "This paper describes the GE Compliance Checker [GECCO], a knowledge-based application for use in the home mortgage industry. GECCO is a tool for automating the information-intensive processes of underwriting and reselling mortgage loans. GECCO was initially designed and deployed for one business component of GE Capital Mortgage Corporation [GECMC], and then successfully integrated into three other GECMC businesses. Its first application was for third-party underwriting. This was followed by the use of GECCO in wholesale pricing and registration, and in direct loan origination. Most recently, GECCO has evolved into a commercial product offered for purchase to mortgage lenders. GECCO has significantly improved the underwriting and resale process: quality control has become much more effective, adding consistency, completeness and robustness to the decision making process; the quantity of loans processed has increased; customer service has been enhanced; and a once subjective process has now been standardized. The successful use of AI has also permeated GECMC business application software to the extent that AI has become a requirement rather than a remote technology used in an isolated application."

Byrnes, Elizabeth, Thomas Campfield, and Bruce Connor. 1989. Innovation and AI in the World of Financial Trading. In Proceedings of the First Annual Conference on Innovative Applications of Artificial Intelligence, ed. Herb Schorr and Alain Rappaport, 71 - 80. Menlo Park, Calif.: AAAI Press. "An expert system called TARA, for Technical Analysis and Reasoning Assistant, has been built and deployed to assist foreign currency traders in their decisions to buy, sell, or hold market positions. Trading with the system began in May, 1988. Since then, TARA has paid back all of its development costs plus an attractive return on investment. This achievement represents a major breakthrough for Al in financial services. Trading is a high risk, high reward profession that at first glance appears unsuitable for an expert system solution. The knowledge is fuzzy, no two experts seem to agree and, for a system to offer significant benefit, a large volume of data must be analyzed in real time. TARA’s success demonstrates that expert systems technology can be effective in the financial trading environment."

Byrnes, Elizabeth, Thomas Campfield, Neil Henry and Steven Waldman. 1990. Inspector: An Expert System for Monitoring Worldwide Trading Activities in Foreign Exchange. In Proceedings of the Second Annual Conference on Innovative Applications of Artificial Intelligence, ed. Alain Rappaport and Reid Smith, 15 -24. Menlo Park, Calif.: AAAI Press. "Inspector, an expert system, was implemented to assist foreign exchange management in the monitoring of trader activity and the compliance of risk management policies. The knowledge of senior foreign exchange managers, traders, controllers, and auditors is contained in an expert system shell, Nexpert Object. The foreign exchange deals from all Manufacturer’s Hanover Trust (MHT) branches are recorded every day in a relational database, Oracle. Combining these systems with local area networks (Lans), global communications, and resident C programs provides a daily review of all worldwide MHT foreign exchange activity. The success of Inspector is directly attributable to the successful combination of traditional technologies, worldwide coverage, and a robust knowledge base. This success demonstrates that expert system technology, coupled with traditional technologies, can be effective in monitoring transaction-oriented financial activities."

Dzierzanowski, James, and Susan Lawson. 1992. The Credit Assistant: The Second Leg in the Knowledge Highway for American Express. In Proceedings of the Fourth Annual Conference on Innovative Applications of Artificial Intelligence, ed. Scott, A. Carlisle and Phillip Klahr, 127-134. Menlo Park, Calif.: AAAI Press. "This chapter describes the development and deployment of the credit assistant (CA), a knowledge-based system to support credit operations for Travel-Related Services (TRS) of the American Express Company. CA was developed using art-im, a rule-based programming environment from Inference Corporation. American Express developed this application under a unix environment (sun and rs/6000) and deployed it into a high-volume real-time mainframe environment. ca was designed to be fully cooperative across business operation units with the authorizer’s assistant (aa) and other knowledge-based systems currently under development. ca also reflects advances in technology and general trends in the AI industry that have taken place since aa was implemented in 1989. This chapter also introduces the knowledge highway concept, the design and construction of a series of cooperative knowledge-based systems to support a global operational strategy of authorizations, credit operations, fraud detection, new account processing, and customer service at American Express. CA was designed to support online credit analysis of card members within the credit operations environment of American Express and to synergistically interact with aa (Dzierzanowski et al. 1989). Credit Operations reviews accounts for the Personal, Gold, Platinum, and OPTIMA card products for credit risk and potential fraud situations. The review process is driven by internal American Express risk management statistical models, which set up risky accounts to be reviewed by analysts. Accounts in question could be set up for many reasons, for example, those showing a delinquency or a history of past due balances. When an account is queued to an analyst for inspection, CA is invoked to support the review by denoting interesting features on the card member’s account and recommending actions. Previous to the implementation of CA, a case required, on the average, 22 transactions to achieve resolution. With CA, one transaction can review data, synthesize information, annotate an account, and provide advice and recommendations to a credit analyst. Advice ranges from setting the account up to be reviewed again in several weeks to recommending the cancellation of a card in serious situations. To support continuous training, scripts are also generated if interaction with the card member occurs. In addition, CA ensures that credit policies are consistently enforced. For example, state laws vary on permissible collection activities. Collection procedures allowed in Minnesota might be illegal in Maine. CA takes all the different statutes into account and guarantees that the analyst is in compliance. As scheduled enhancements are rolled out, the system autonomously makes decisions on some cases, composes letters to card members, orders additional information when necessary, and routes accounts into queues for specific actions by analysts. Designed as components in the knowledge highway of cooperative expert systems, CA and aa have been built for compatibility and allow for the real-time exchange of data. Results from one system can be incorporated into the decision process of the other. aa is part of American Express’s front line of service in credit authorization at a point of sale. It interfaces with the TRS credit authorization system (cas), which is based on a transaction-processing facility-based system built for high volume and transaction rates. aa handles transactions referred by cas. Those charges that are not resolved by aa automatically are then sent to an authorizer with supporting advice and recommendations, thus serving as a decision support tool."

Egan, Jack. Artificially Intelligent Investing. In U.S. News and World Report: March 15, 1993. "Cynics might wonder whether there's intelligent life to be found on Wall Street, but within the financial community there's plenty of life in artificial intelligence. AI, as it's commonly known, is the programming of computers to approximate the workings of the human brain. Far more than mere number crunching, the focus is on sophisticated tools like neural networks that simulate thought processes." Available for purchase from the magazine.

Freedman, Roy S., Robert A. Klein, and Jess Lederman, editors. 1995. Artificial Intelligence in the Capital Markets: State-of-the-Art Applications for Institutional Investors, Bankers & Traders. Chicago: Probus Pub.

Golibersuch, David C., Rebecca Towne and Cheryl A. Wiebe. 1995. GENIUS(TM) Automated Underwriting System: Combining Knowledge Engineering and Machine Learning to Achieve Balanced Risk Assessment. In Proceedings of the Seventh Innovative Applications of Artificial Intelligence Conference, 49-61. Menlo Park, Calif.: AAAI. "The GENIUS Automated Underwriting System is an expert advisor that has been in successful nationwide production by GE Mortgage Insurance Corporation for two years to underwrite mortgage insurance. The knowledge base was developed using a unique hybrid approach combining the best of traditional knowledge engineering and a novel machine learning method called Example Based Evidential Reasoning (EBER). As one indicator of the effkacy of this approach, a complex system was completed in 11 months that achieved a 98% agreement rate with practicing underwriters for approve recommendations in the fist month of operation. This performance and numerous additional business benefits have now been confirmed by two full years of nationwide production during which time some 800,000 applications have been underwritten. As a result of this outstanding success, the GENIUS system is serving as the basis for a major re-engineering of the underwriting process within the business. Also, a new version has recently been announced as an external product to bring the benefits of this technology to the mortgage industry at large. In addition, the concepts and methodology are being applied to other financial services applications such as commercial credit analysis and municipal bond credit enhancement. This paper documents the development process and operational results and concludes with a summary of critical success factors."

Goonatilake, Suran, and Philip Treleaven, editors. 1995. Intelligent Systems for Finance and Business. Chichester, UK & New York: Wiley.

Hart, Peter E. 1989. Application-Driven Architecture: A Case Study of Syntel. In Proceedings of the First Annual Conference on Innovative Applications of Artificial Intelligence, ed. Schorr, Herbert and Alain Rappaport, 62-70. Menlo Park, Calif.: AAAI. "The special, requirements imposed by important classes of financial risk assessment applications have driven the development of a new expert system architecture. We discuss the relation between the demands of these applications and features of the architecture, and describe our experience with large-scale, deployed products that have been built using this new approach."

Hottiger, Steve, and Dieter Wenger. 1992. MOCCA: A Set of Instruments to Support Mortgage Credit Granting. In Proceedings of the Fourth Annual Conference on Innovative Applications of Artificial Intelligence, ed. Scott, A. Carlisle and Phillip Klahr, 135-150. Menlo Park, Calif.: AAAI. "Mortgage credit granting has to be supported by task-oriented instruments (highly interactive support tools that enable cooperative problem solving) because of increasing competition and unfriendly economic circumstances. mocca (mortgage controlling and consulting assistant) provides instruments for decision support, customer consultancy, and management. Most of the underlying models in mocca are innovative in terms of being either completely new or operative for the first time. The implementation of the application is based on a combination of different knowledge-based techniques, such as agent-based processing, generic application building blocks, fuzzy measure, data-driven paradigm, and a two-layer windowing system. mocca (mortgage controlling and consulting assistant) is now being used in the major branch offices of Swiss Bank Corporation (SBC). The development of the application required two person-years."

Kindle, Kyle W., Ross S. Cann, Michael R. Craig, and Thomas J. Martin. 1989. PEPS: Personal Financial Planning System. In Proceedings of the First Annual Conference on Innovative Applications of Artificial Intelligence, ed. Schorr, Herbert and Alain Rappaport, 51-61. Menlo Park, Calif.: AAAI. "PFPS is an expert system developed over the last five years at Chase Lincoln First Bank, N.A., to provide objective, affordable expert financial advice to individuals with household incomes ranging from $25,000 to $150,000, and up. PFPS is an integrated personal financial planning system encompassing the following areas of financial expertise: investment planning; debt planning; retirement savings and settlement of retirement plans; education and other children’s goal funding; life insurance planning; disability insurance planning; budget recommendations; income tax planning and savings for achievement of miscellaneous major financial goals (purchase of house, extended travel, etc.). PFPS has supported the personal financial planning service of Chase Lincoln First in upstate New York since late 1987. The PFPS system components are implemented as an embedded system on an IBM 4300 series mainframe under VM/CMS and on a Symbolics 3600 series LISP processor. The Symbolics system is connected to the IBM mainframe in a master/slave relationship using an application -level protocol defined on top of RS232C. All of the inferencing and planning is done on the Symbolics 36xx using a system architecture based on a blackboard framework, an object-oriented database, and a goal-directed generate and-test search paradigm with extensive search-space pruning."

Kingdon, Jason 1997. Intelligent Systems and Financial Forecasting. New York: Springer Verlag.

McDonald, David W., Charles O. Pepe, Henry M. Bowers, and Edward J. Dombroski. 1997. Desktop Underwriter: Fannie Mae's Automated Mortgage Underwriting Expert System. In Proceedings of the Ninth Annual Conference on Innovative Applications of Artificial Intelligence, ed. Senator, Ted and Bruce Buchanan, 875-892. Menlo Park, Calif.: AAAI. "Fannie Mae, the nation’s largest source of conventional mortgage funds, has made a commitment to use technology to improve the efficiency of processing a loan by reducing the time, paperwork and cost associated with loan origination. The Desktop Underwriter (DU) system which was developed as a result of this commitment, is an automated underwriting expert system that applies both heuristics and statistics to the problem. The system supports both the wholesale and retail mortgage environments and is built to reason and underwrite loans with incomplete, unverified and conflicting data. The system generates a credit recommendation based on the loan’s conformity to credit standards and an eligibility recommendation based on the loan’s conformity to eligibility requirements. DU is already having a major impact on the mortgage industry. The system helps standardize how the Fannie Mae underwriting guidelines are interpreted, reduces discrimination by removing subjective reasoning from the decision process and reduces the cost of manual underwriting for both lenders and Fannie Mae."

Meltzer, S., and D. Sriram. 1990. ReValuator--An Expert System Approach to Actuarial Valuations . In Proceedings of the Second Annual Conference on Innovative Applications of Artificial Intelligence, ed. Rappaport, Alain and Reid Smith, 39-48. Menlo Park, Calif.: AAAI. "Expert system technology has now matured so that task-oriented business programs can be rapidly prototyped, developed, coded, and deployed on desktop and laptop personal computers. This rapid development and deployment is especially true when the task is well defined, and the target user has little knowledge in the specified domain. This paper sketches the successful implementation of an actuarial program designed to assist a nonactuary in detailed actuarial analysis."

O'Leary, Daniel E., and Paul R. Watkins, editors. 1992. Expert Systems in Finance. Amsterdam: North-Holland.

Pethokoukis, James M. Robotrading 101 - Sophisticated computer programs take the human element out of picking winners on Wall Street. U.S. News & World Report (January 28, 2002). "'We all have the same data, and the question is what the hell are we going to do with it,' says Doug Case, chief investment officer at Advanced Investment Technology in Clearwater, Fla. Case sees AI as the key to decrypting high-velocity, information-saturated financial markets. 'AI can deal with that data and handle these disorderly global markets,' says Case, whose $1 billion firm is majority owned by State Street Global Advisors. ... The VirtualHamilton presents a more tantalizing use of the technology. Why not also a VirtualBuffett or VirtualLynch? These digital doppelgþngers might beat the originals by quantifying the unconscious intuition of these fabled investors. Just as an Ichiro Suzuki doesn't run trajectory and velocity calculations before catching a fly ball, many managers probably don't fully understand how they analyze stocks. Digitize a superstar manager's moves, and you might be able to hack his financial mind. 'That's called reverse engineering,' says Yale finance professor William Goetzmann." Also see a related article from the same issue: Investing Tool - You can do this at home.

Sahin, Kenan, and Kith Sawyer. 1989. The Intelligent Banking System: Natural Language Processing for Financial Communications. In Proceedings of the First Annual Conference on Innovative Applications of Artificial Intelligence, ed. Schorr, Herbert and Alain Rappaport, 43-50. Menlo Park, Calif.: AAAI. "This paper describes the Intelligent Banking System (IBS), a family of applications developed for Citibank. New York, by Consultants for Management Decisions (CMD) to increase the productivity and effectiveness of English text message processing. These messages were previously processed manually. Data entry operators would read and analyze the message and then type information at a standard ASCII terminal interface. IBS applies a combination of natural language processing and -rule-based expert system techniques to analyze the message and to generate a formatted equivalent. IBS also provides a sophisticated intelligent user interface which aids users by applying the system’s domain knowledge to the interactive session."

Shell, Pete, Gonzalo Quiroga, Juan A. Hernandez-Rubio, Eduardo Encinas, Jose Garcia, and Javier Berbiela. 1992. Cresus: An Integrated Expert System for Cash Management. In Proceedings of the Fourth Annual Conference on Innovative Applications of Artificial Intelligence, ed. Scott, A. Carlisle and Phillip Klahr, 151-170. Menlo Park, Calif.: AAAI. "Cresus is a unique application of state-of-the-art expert system technology to the real-world financial problem of cash management. By automating the work of company treasurers, it saves substantial amounts of both money and time every day. Real-world test cases show that cresus performs better and much faster than the human expert: In minutes, it generates a combination of operations that efficiently balances all banking accounts in a 15-day period. It uses user-friendly window technology to control the human-machine dialogue and has been integrated into the work environment of a major electric company in Spain. Written in Common Lisp using unix workstations, it was jointly developed by Union Fenosa, Carnegie Mellon University, and Norsistemas Consultores."

Shimokawa, Tetsuya, Tadanobu Misawa and Kyoko Watanabe. Word of Mouth : An Agent-based Approach to Predictability of Stock Prices. Transactions of the Japanese Society for Artificial Intelligence, Vol. 21, No. 4, pp.340-349 (2006).

Trippi, Robert R., and Jae K. Lee. 1996. Artificial Intelligence in Finance & Investing : State-of-the-Art Technologies for Securities Selection and Portfolio Management. Chicago: Irwin Professional Publishing.

Way, Cyril. 1997. STHANA: Profitability Forecast and Situation Analysis for Automated Teller Machines. In Proceeding of the Ninth Annual Conference on Innovative Applications of Artificial Intelligence, ed. Ted Senator and Bruce Buchanan, 926-931. Menlo Park, Calif.: AAAI. "The French credit card system makes it highly profitable for banks to have heavily used Automated Teller Machines (ATM). The goal of the Sthana system is to capitalize the knowledge spread all over a company into a system capable of issuing recommendations for existing ATM's and capable of forecasting a new ATM's activity. Sthana uses Data Mining and Case-Based-Reasoning techniques to extract information from existing data (including economic, geographical and internal bank data) and from the bank's ATM experts. The system builds up classifications on high level descriptors from raw data and eventually indicates a measure of the ATM's activity and profitability, highlights factors which could lead to higher profitability or pinpoints the ATM's vulnerabilities."