Marc Chaikin started his career in trading as a stockbroker in 1966. By 1969 as a new bear market was in full swing, he began doing research into technical analysis as an alternative to fundamental research. He started working with relative strength and volume analysis, following the work of Larry Williams and Joe Granville. “I wanted to develop a multi-indicator approach to trading using relative strength, momentum, and volume analysis,” Marc told us. Soon thereafter Marc started his first hedge fund, but he was back as a broker by 1971. By 1977, he had become head of the options department at Tucker Anthony & R.L. Day, Inc. where he applied the technical analysis he had developed to options trading.

It wasn’t long before Marc purchased a seat on the New York Futures Exchange (NYFE) so he could trade right on the floor. He started using standard trading software in 1980 to build his own indicators – still centered around the work of Williams, Granville and David Bostian. He developed the now famous Chaikin Oscillator, Chaikin Accumulation/Distribution indicator, and the Chaikin Persistence of Money Flow indicator. His work was devoted to determining whether stocks were under accumulation or distribution and then applying momentum indicators for timing. However, he left the floor in 1981. “I always wanted to be on the floor.” he said, “but it turned out to not be such a great place to make money.”

Next Marc joined Drexel, Burnham, Lambert, Inc. in 1982 where he began doing press pieces and even a segment on the Financial News Network (FNN), which was the predecessor to CNBC, and continuing on at CNBC. “At Drexel I became interested in neural networks, just as my son was studying the technology at Brown University. Look at your records and you will see that I was one of the earliest users of NeuroShell 1,” he told us proudly. “I used it to build trading models for stock index futures.”

Marc’s next adventure was to create the very successful Bomar Securities, where he built a technical analysis workstation for portfolio managers and traders. He sold the company in 1992 to the Instinet subsidiary of Reuters. However, he continued on building equity related analytics products for Instinet and Reuters. He continued his neural network research too, but never really had enough data to build robust models until he used S&P cash index data back to 1962, after which he developed successful models to predict S&P futures 3 days in advance.

During this time, Marc was a lecturer on neural nets at places like the New York Society of Security Analysts. He began using genetic algorithms and later on fuzzy logic in conjunction with John Bollinger. “I developed some very strong opinions on how to build neural nets,” he added, “feeling that you had to feed significant and proven reversal and continuation patterns to the network in order to build profitable neural networks.”

Marc left Instinet in 1998 to focus on day trading as a business. “I wanted disciplined models to deal with intraday volatility,” he explained. “I became interested in candle charts as a way to explain and deal with some of the extreme volatility, but the problem was that there were too many formations and the interpretations were too subjective.”

Then in 1999 he began reading the articles of Viktor Likhovidov in Technical Analysis of Stocks and Commodities magazine (TASC). “The TASC articles discussed how to measure candlestick patterns by assigning a ranking to each candle. I was intrigued. Unfortunately, I didn’t have programming resources to implement and test Viktor’s work, so instead I developed rule based systems based on intraday momentum to trade stocks and ETFs (QQQ and SPY).” Later Marc was to return to candlesticks.

In April 2002 Marc spotted a TASC press release about the NeuroShell DayTrader Professional. “I didn’t realize you guys had a day trading product, so I picked up the phone and talked to Marge Sherald and later Steve Ward, whom I hadn’t spoken to in about 12 years! I purchased the DayTrader and the first month I was able to build a short term trading system better than anything else in my career! Success came from using the optimizer as opposed to the neural net aspects of the software. I brought in my year old rule based system and combined it with the candlestick systems you provided in the TASC tips. One of the best things about your systems is that you regularly post the charts from the TASC articles on your website, making it easy for non-programmers to implement and test new ideas. It is very valuable. The optimizer then helped me extend Viktor’s work.”

We asked Marc to explain his system a little further. “I incorporate candlesticks in a disciplined way using an optimized version of Likhovidov’s CandleCode along with momentum indicators such as MACD in two time frames, one a five minute and the other a longer term MACD, as well as two time frame moving averages, again one a 5 minute and the other based on a longer term time frame. The icing on the cake was the ability to incorporate money management discipline in the form of multiple stops, and either test them by trial and error, or with the optimizer. Using the back testing capability, I was able to zero in on a system which models the extreme volatility of the QQQ. I built some SPY systems which were also profitable, but they weren’t as profitable as the QQQ system. Actually, I determined that I could apply the QQQ signals to the SPY with much better success.”

Marc further explained that he used 2 years of hourly data and tested on one year out-of-sample. “With the power of NeuroShell and the speed of modern processors, I did all of this very quickly compared to my previous efforts!”

Marc Chaikin’s current goal is to become a commodity trading advisor, implementing trades for ETF’s and Nasdaq and S&P futures. His new company is Chaikin Capital, LLC, in Washington, CT. It is scheduled to start money management in October 2002. “I like the real time monitoring capability in the chart and portfolio views,” he told us. “I can show prospective investors how I monitor the signals and the stops. It gives them confidence that a disciplined system is being used. The back test is also very valuable in marketing. Anybody who wants to market their system has a valuable tool here – you can compare the actual to historic trade by trade to show how you are tracking your system in real-time trading.”