# Moving averages and MACD

The main drawback of using moving averages in share price analysis is that moving averages tend to give accurate signals of turning points in price too late for traders to act on them.

It is maybe not so surprising, therefore, that technical analysts should have tried to devise indicators that get around this objection. In the 1960s, US technician Gerald Appel devised that has become known as the Moving Average Convergence Divergence indicator - or MACD.

It belongs to series of indicators that use a cyclical indicators with a relatively long amplitude combined with one of a shorter amplitude that acts as a 'trigger' for buy and sell signals.

In the case of the MACD, rather than use moving averages covering two different period lengths (14-day and 42-day averages for example, or 12-day and 26-day in some formulations) to perform this function, it does something rather more subtle. It calculates the moving averages for two time periods like this, but then computes a series of values for the difference between them, by subtracting one from the other. This is then plotted on a chart as the so-called 'main line'.

What it then does is to compute a moving average of the main line over a relatively short period (perhaps seven days or nine days) to act as a trigger, with buying and selling signals being generated when the two lines cross under specific circumstances.

The MACD is plotted on a chart where zero is the equilibrium and buy signals are generated when the main line crosses up through the trigger line from below, with sell signals generated when the main lines crosses down through the trigger line from above. Some schools of thought also read the crossing of the main line through the zero axis from below as a buy signal and the crossing through zero from above as a sell signal.

Confused signals

One problem with MACD is that when a particular market or share is directionless, the signals become confused. As an indicator it is best used to spot a change of direction after a sustained trend has been established for a significant length of time. The MACD can also be used in conjunction with the underlying share price. A new high for the share price that is not accompanied by the same thing for the main MACD line is sometimes construed as bearish, for example (and vice versa when a share hits a new low). It perhaps shows that while a share may ostensibly be making new highs, its directional momentum is fading.

MACD was regarded as an extremely reliable tool in the 1980s, but in recent years there has been increasing scepticism about its reliability. Some studies show that over time it tends to generate a surprisingly high number of loss-making trades when used in a standardised way.

One problem is that, with the advent of greater computing power, the routine anomalies that were present in the system in former times are no longer there, and that traders can scrutinise - and act on - a range of patterns almost instantaneously.

Peter says

While most acknowledge that MACD is an improvement on the simple use of crossover points of raw moving averages, the fact remains that MACD essentially remains a lagging indicator that is prone to unreliability, particularly if there are sharp changes in the market or in market volatility.

Volatile conditions can increase the likelihood that a signaled change in trend will reverse extremely quickly leading a so-called 'whipsaw' effect from a false signal and short lived change in direction.

At best, therefore, MACD is best used - as are many technical indicators - as one of a series of tools and used for confirming a signal given in the first instance by other, more reliable methods.

I must admit to being something of a sceptic when it comes to believing in the reliability of many technical indicators and the way in which the usefulness of this indicator has declined over time and alters with differing market conditions seems to me to suggest a more general problem that applies, to a greater or lesser degree, to all analysis of this type.