BEST Artificial Intelligence Software
NeuroShell Trader Professional
2003, 2004, 2005, 2006, 2007, 2008
2009, 2010, 2011, 2012, 2013, 2014
We have been rated #1 TWELVE years in a row!

Interviews with Real Traders   More interviews...
Standard and Poor's Andre Archambault

Andre Archambault, BA, JD, MBA, is the Director of Quantitative Strategies at Standard and Poor's Investment Services, and he is also the developer and genius behind Standard and Poor's extremely popular Neural Fair Value (NFV) System for ranking stocks and building portfolios. NFV, which uses neural nets made with NeuroShell, is among the more popular features of Standard & Poor's MarketScope, an electronic investment intelligence service for financial advisors, brokers, and money managers with approximately 77,000 subscribers worldwide.

After briefly practicing law in California, Andre focused his career on investments, moving to New York and starting work as a special situation analyst for a brokerage firm. After 2 years, he was brought to Chicago by the Continental Bank to work as an analyst/portfolio manager in its Trust department. Seventeen years later, Andre left Continental to buy a seat on the CBOE. After about two years of making markets on the floor of the exchange, Andre yearned to return to the more traditional investment world where his long term planning abilities served him best. He accepted an offer by Standard & Poor's to return to New York and head MarketScope’s research department.

Andre Archambault purchased his first copy of NeuroShell in 1993, and has purchased just about everything we have sold since then. "NeuroShell is a great program," he says, "and one of the essential reasons for the success of NFV portfolios. It has been a real performance booster to the original Fair Value System."

The Fair Value system started in 1995, but Andre decided to boost its performance with neural networks in 1999. After a year of work preprocessing data and evaluating inputs, the Neural Fair Value system was introduced at the end of September 2000. The idea was to pick the best performing stocks for a portfolio in which the subscriber was always invested. The first NFV-based portfolio -- called NFV20, as it always held exactly 20 stocks -- became incredibly popular among MarketScope’s subscriber base, and has become the basis for S&P’s model portfolio management. NFV20, from the end of October 2000 to the end of April 2005, had a compounded annual growth rate of 10.9%, compared to -4.6% for the S&P 500 index. A portfolio of $10,000 invested in 9/31/2000 would now be worth $16,253 - pretty good, especially considering how the market as a whole has been during those years.

NFV20 was replaced on MarketScope by a 25-stock portfolio, the NFV25, on October 5, 2004. Since then to the end of April 2005, NFV25 is up 10.41% compared to the S&P 500 index itself, which is up only 1.97%. S&P has developed several other NFV model portfolios that, like NFV20, are used to manage the portfolios of major institutional investors.

Here is how Andre describes NFV25:

"The Neural Fair Value 25 Portfolio, a quantitative model proprietary to Standard & Poor's, seeks to outperform the market by buying undervalued issues with the potential to deliver superior returns over the next six months. The NFV 25 is based upon the Fair Value Portfolio. The Fair Value Portfolio chooses the most attractive stocks based on where stocks are trading in relation to their Fair Value, or the price at which a stock ‘should’ trade. That price is based on such fundamental data as corporate earnings, earnings as projected by Wall Street analysts, price-to-book, return on equity, and current yield relative to the S&P 500. The Neural model is based on "Neural Networks" theory, an artificial intelligence concept designed to replicate the human brain's ability to learn. During a neural model's training period, prediction errors are reduced by adjusting inputs. In Standard & Poor's Neural model, we input the factors that produced better market performance during the most recent 6 months. The Fair Value universe is then filtered through the Neural model creating the Neural Fair Value universe of equities. In other words, Fair Value stocks that are more sensitive to current market movements. The NFV25 portfolio is composed of stocks from the Neural Fair Value universe considered to have superior price appreciation potential.”

In addition to NFV, Andre also publishes other market predictions each day on MarketScope using other NeuroShell models.

Advanced Neural Network Software for Financial Forecasting and Stock Prediction

Ward Systems Group, Inc
Executive Park West
5 Hillcrest Drive
Frederick, MD  21703

Phone : (301) 662-7950
Fax : (301) 663-9920
Email :

Skype (Sales Only) : wardsystems

Copyright© 1997- 2014 Ward Systems Group, Inc. All rights reserved
Copyright Information
Privacy Statement

IMPORTANT INFORMATION AND DISCLAIMERS Investments and trading involve risks, including possible loss of principal and other losses. The Software is designed chiefly to provide tools with which you can learn to build your own systems applying artificial intelligence to the financial markets. The Software and the examples therein are not intended to be working trading systems. The Software is licensed to you with the understanding that Ward Systems Group is not engaged in rendering any investment or other professional advice. If investment or other professional advice is required, the services of a competent, licensed professional should be sought. No Ward Systems Group employee, agent or representative is authorized to provide any such advice of any nature whatever, and any such advice, if given, is in violation of Ward Systems Group policy, is unauthorized and may not be relied upon. The use of any trading system or strategy, including any system or strategy included as a sample in, or that is or was developed using, the Software, does not and cannot guarantee that you will make profits, increase profits or minimize losses. Any popular or other strategies or systems included in the Software are intended merely as examples of technical ideas that can be incorporated into a personally designed trading system. None is recommended. You must use your own judgment or consult a professional for advice on such matters. Additionally, trading results based upon hypothetical or historically-tested trading systems do not necessarily compare to results of actual trading. No hypothetical or historical trading record can completely account for the level of risk present in actual trading. Numerous factors relating to market conditions, human error, human/emotional reaction to losses during actual trading, and inherent limitations of certain hypothetical or historically-tested models can account for these differences. There is no guarantee that your hypothetical trading results, even if tested against historical data, will produce comparable actual trading results. In fact, there are frequently sharp differences between hypothetical or historically-tested performance results and the actual results subsequently achieved by any particular trading system or strategy. Also, you should be aware that certain commonly-used trading "jargon," including trading terms used in the Software, such as, for example, a "limited risk" position, should not be taken literally. For example, so-called "limited risk" positions in certain options trades are not in fact limited as may be expected. Again, with respect to matters such as these, a licensed professional should be consulted.

Click here to read the entire license agreement