It’s Sunday morning. Do you sit in front of your computer, coffee cup in hand, and look for stock trading candidates? Maybe, you input technical conditions into a scan engine. Then it spits out a bunch of stocks with oscillators entering or leaving overbought or oversold zones, or with classic MacD crossovers either bullish or bearish. Perhaps the scan engine suggests candidates flirting with a Fibonacci level or Fibonacci time frame, or Fibonacci arc or Fibonacci angle, there are any manner of Fibonacci ingredients, and there are any manner of general technical conditions to screen for. Many of them work. Often these scans reveal some very good trading material. Sometimes they don’t and that’s all part of the trading game.
My main objection to scanning via indicator or oscillators is that the people who use them often have no understanding of how they are constructed. For instance, the data used in the Relative Strength Index, is similar to the data used to compute the Stochastic indicator. “Technicians” have to be careful then not to use a like indicator to confirm another like indicator. This problem is called multicollinearity. It’s comparing apples to apples by using similar input data.
Back in the days before computer charting, I had to update weekly charts sent by mail. I would simply draw a new daily candle as the trading week progressed. But, if you chose to, you could also update the technical indicators, like the MacD and the RSI, which were included on the charts. This involved following the formula for the indicator and doing the math. One positive result was that you became familiar with the information that went into the construction of the indicator or oscillator. This helped avoid the danger of multicollinearity and it gave me a better understanding of technical indicators and oscillators. Today there are so many or more accurately too many, indicators that can be attached to your computer charts. Again, “technicians” (I am using this word lightly) fill their charts with studies and they have no idea what those studies represent. To them it’s this simple: a moving line crossed over or under a particular horizontal line, and that is either a good or bad thing for the stock. If you don’t know what input produces the output of your indicators you are not a technician. But, I digress.
Reverse engineering a chart is the opposite of imputing data into a scan engine to find a trading candidate. It involves first identifying a stock chart that provided a profitable trade in the recent past. Then disassembling the chart by looking at the technical conditions just prior to the initiation of the trade. This includes not just the indicators and oscillators, but any candle patterns and volume levels, all using multiple time frame analysis. Basically, you’re looking at a chart that represents a previous profitable trade and you’re trying to identify the early technical particulars that made it a winner. Breaking it down and deconstructing the conditions just prior to the initiation of the trade.
Often traders just move on if they miss a winning trade. They immediately start scanning and scrolling for another opportunity. But reverse engineering, even a missed trade, is an opportunity to gain more trading insight. Take, for instance, a chart that has a clear stair-step pattern of higher highs and higher lows. It may be too late to enter that trade but you can go back to where the trend began, and notice the technical conditions that presaged the start of the pattern. The same applies, of course, to bearish trend reversals. What happened to momentum and money flow ahead of the reversal?
Reverse engineering is a simple and unique way to analyze a stock chart, and it can prevent weekend chart list highway-like hypnosis. Always use your time productively. Don’t just blindly scroll through reams of charts. Pick prime candidates and analyze them thoroughly and in a unique way. It keeps you sharp and on the lookout for new trading ideas.