5.1 Presentation of indicators

Technical analysis is composed of two parts: graphical analysis and numerical analysis. The first is essentially based on the simple observation of prices and volumes levels, as well as the existence of characteristic graphical figures. The second one, in comparison, uses mathematical constructions. On the basis of prices (especially closing and extreme levels), technical indicators are created.

These latter can belong to two categories, referring to both facets of the investors’ psychology. These are trend indicators, usually based on averages, on the one hand and counter-trends indicators, usually based on derivatives, on the other hand.

The main interest of technical indicators is their predictive capacity, while graphical analysis often has this possibility only when figures are at least partially drawn. This is especially true for counter-trend models.

5.2. Usefull indicators

5.2.1. Moving averages

The most simple trend indicators are moving averages. They simply correspond to an average calculated on an evolving time scale: every day, the oldest value (often taken at the close) in the average calculus is replaced by the value of the new session.

Consequently, the predictive interest of this indicator is nil (since it represents prices evolution with a certain delay). Still, it enables one to determine trends of mid or long term, stronger and stronger as the average direction is steady.

In spite of the simplicity of this indicator, the length of averages used should be handled with caution. Indeed, analysts prefer using two moving averages simultaneously, with quite different lengths to forecast possible trend reversals. Thus, one will often jointly use moving averages calculated on 20 and 50 days, or on 50 and 100 days…

In particular, this simultaneous use makes it possible to determine buying signals. These occur whenever a short term moving average (e.g. 20 days) crosses a longer term moving average (e.g. 50 days) coming from beneath and thus comes above. This expresses the tendency of the stock to have its most recent prices at a level higher than older prices, thus showing a bullish trend.

Reciprocally, a selling signal occurs whenever a short term moving average crosses down (i.e. from above) a longer term moving average and thus comes beneath.

Overall, the interest of moving averages is to avoid going against the market trend when it follows a strong move.

5.2.2. Bollinger bands

These indicators are derived from moving averages and aim at filling in an important gap. Indeed, moving averages give buying and selling signals at punctual levels, which can thus quickly be invalidated should the market reverse on the very short term (during the session or from a session to another).

It can thus be wise, rather than defining precise thresholds, to use zones defined as intervals on both sides of the moving average. Bollinger bands are built on this very principle. This figure is composed of three trend lines: the middle, upper and lower bands.

The middle band corresponds to a simple moving average, often calculated on 20 days.

The level of the upper band, in every point, corresponds to the sum of the level of the middle band and twice the value of the standard deviation associated to the moving average.

Reciprocally, the level of the lower band corresponds to the level of the middle band diminished by twice the value of the standard deviation associated to the moving average.

An envelop of the stock price is thus determined. This makes it possible to then identify the variation margin in which the stock should stay almost systematically. In the case of a stock following a Gauss law, 95 % of the trades will thus occur between these bands.

These latter then constitute very strong support (lower band) and resistance (upper band) levels. These levels respectively represent interesting buying and selling levels, particularly when no real trend appears on the market (and bands are thus stable on both sides of the average), which enables to play with a trading target (short term target).

In opposition, in a trend market, clues given by bands are related to their spread. Indeed, a growing spread of bands means a growing standard deviation, which is the sign of the beginning a strong trend. Then, when bands narrow, variations on both sides of the moving average get smaller, which suggests the end of a trend. It is then possible to use bands as supports and resistances.

5.2.3. The MACD

One of the most commonly studied technical indicators is the MACD. This indicator (Moving Average Convergence / Divergence) reflects a difference between moving averages and refers to the ascendancy or not of the mid-term relative to the short term.

The considered average lengths are respectively 26 days (0.075 exponential coefficient) and 12 days (exponential coefficient of 0.15).

Moreover, to estimate the variations of the trend, an auxiliary indicator (named signal line) is formed. It is based on a new exponential average on 9 days (0.20 coefficient).

The advantage of this indicator is triple: absolute position of the MACD, relative position to its signal line and existence of divergences.

From the first point of view, oversold and overbought situations can be identified. Thus, a strong rise of the MACD indicates that the 12-day moving average is more rapidly rising than the 26-day one, thus showing a stronger volatility in the short term. Then, the crossing of the zero level should be considered with special caution.

From the second point of view, one of the most relevant invitations to buy is the crossing up of the signal line by the MACD, especially when it occurs on up reversing levels for the MACD (cf. graph).

From the third point of view, divergences can be identified between the MACD trend and that of the stock price on a given period. This phenomenon is marked by the more than proportional increase or decrease of the MACD compared to the stock variation (cf. graph).

5.2.3. The DMI

As opposed to other indicators, the DMI (Directional Movement index) does not aim at defining excesses (oversold / overbought areas) or divergences but rather at determining trends, used to identify buy and sell signals.

It is much more complex to form than other oscillators. First, pressure indicators to buy (+DM) and sell (-DM) are identified. These pressures are then expressed as a percentage of the maximum variation of the market on the period, and a moving average of these indicators is calculated, respectively +DMI and –DMI.

Let us take the example of a 14-day period, with +DMI14 = 0.2 and –DMI = 0.36. This means that 20 % of the market range in the past 14 days was done on the upside and 36 % on the downside. On this range, 56 % (0.2 + 0.36) was directional.

Thus, the more directional the market (on the upside AND on the downside), the greater the sum of the DMIs (DMI sum). Still, the difference between +DMI and –DMI is more often calculated (DI diff). The information given on the trend is then different: the higher this difference, the more directional the market in the same direction. The DI diff / Di sum ratio, expressed as a percentage, then gives DX, which is smoothened on the period (often 14 days) to form ADX.

Once these oscillators are built, it becomes quite easy to use them. First, one can compare the respective positions of +DMI and –DMI. If +DMI stands above –DMI, the trend on the upside is strong, which means that buyers win more and more. In opposition, if +DMI stands under – DMI, this is the trend on the downside which is strong, and the sellers then become the winners.

Moreover, signals are also given by the ADX. As this latter bypasses 17, the market is considered as following a trend. It is then possible to buy (+DMI above –DMI) or sell (-DMI above +DMI).

Finally, the ADX enables one, along with moving averages, to determine the validity of the latter. Indeed, as moving averages sometimes have false signals, it is possible to buy only when the ADX shows a trend, and then to follow signals given by moving averages.