Following our tour of techniques, sometimes quite simple, used by some traders. Our desire is always to offer some understanding of market mechanisms, and help you understand how methods used today and very discredited. We recall that many much more complicated techniques are used on the options markets. For example, Black and Scholes, but also the fractal theory, GARCH model, binomial trees, hopping process, Levy, integro-differential equations and many others. Let us now technique called MACD. Moving Average Convergence Divergence. As its name implies, this tool allows to study the differences between two exponential moving averages. It is calculated simply by difference between two exponential moving averages of different periods. It generally uses the moving

average 12 days (MME12) and 26 days (MME26). Comparing the MACD with its own exponential moving average at 9 days (MME9 = Closing Day * 0.09 + MM from yesterday * 0.91). The MACD is a difference, it will be zero whenever MME12 and MME26 be identical. Here is a graph theory to understand the different interpretations that can extrapolate from this model. The most common use for this indicator is to consider the curve MME9 (Exponential Moving Average to 9 days of the MACD) as a signal line. So when the MACD crosses the line up (1), we are dealing with a buy signal and vice versa when it crosses the line down. This practice of the MACD is very successful because the signal is earlier than the signal caused by the croissement of MM20 and MM50 discussed in previous articles devoted to moving averages. But it also makes the danger of this model as "more an indicator signal, the sooner it tends to deceive." To circumvent this problem, experts have decided to enter the market, in the presence of a signal increases, if the MACD and MME9 were on the same side for a minimum of 14 days. They want to ensure that one is in a situation stable enough that the increase makes sense. Another way to forecast is to study the differences between the MACD and the price but also with the line of zeros. When a value is bullish and the MACD rose by deviating the course and the zero line, we can predict that this growth will continue. Signs of turnaround will be felt when the MACD moves closer to these two lines (2). We can finally use the MACD to detect if a value is overbought or oversold. When the curve of the MACD is growing too quickly is to say when the 12 days moving average deviates rapidly moving average 26 days, we are in a position to purchase on the market and vice versa. Here is a representation of the MACD to the Eurodollar during the year. It notes that after a long period in November during which the MACD (blue line) was above MME9 (A), we have witnessed a cross between early December these curves (B) which was a harbinger of a reversal trend mid December and continues into January. This model is certainly more complex to use but is effective for purposes of short-term forecasting. It gives signals "relatively reliable" in a market of "trending", where the trend is significant, but not too pronounced. It can then capture all the increase (or decrease) in eliminating early departure. The purpose of a few lines of this article is not to judge the appropriateness or otherwise of the technical analysis and results in a market context, but to highlight the extreme disconnect that can exist between these methods and the economic reality and business.