Stochastics and turning points

Share this

Technical analysis is the art (some would say science) of forecasting future share price movements from analysis of past movements. It is not all about finding arcane and improbable patterns that might only be in the eye of the beholder. Much of it is based on rigorous statistical analysis.

One example of this is a group of indicators that examine momentum trends in share prices.

Momentum indicators

Momentum indicators are used to spot turning points in share prices. They measure the increasing or decreasing rate of change of individual share prices. They work on the principle that momentum is highest early in a trend and then steadily decreases, rather like a ball thrown up into the air. At first the rise is rapid but its upward momentum declines as gravity reasserts itself, and then it begins to fall, slowly at first and then with increasing speed.

There are many technical indicators of share prices, but many of them have a crucial drawback. While they are quite good at signaling turning points, some of them tend to do so only after the event, which means that some of the benefits of a movement in the share price may have been lost.

Stochastics

One of the key momentum indicators that attempts to get around this problem is a group of what are known as stochastic indicators (or simply stochastics). These can be of use in certain circumstances. They tend, for instance, to be more effective in establishing turning points in shares which show strong cyclical tendencies.

The idea behind a stochastic indicator is based on the premise that if a share is in an uptrend, then prices tend to accumulate in each successive period at or near the top of the range of prices for that period. The same thing happens in reverse during a downtrend. If we can work out a way of representing this on a chart, then, so the theory goes, the turning point in the stochastic value should coincide with, or even lead, the turning point in the underlying security.

Calculating

The stochastic value is calculated by taking the difference between today's price and the low point for the period (say, 15 days) being considered, compared with the difference between the highest and lowest price for the same 15 days. It can therefore vary between zero and 100, the 100 point being reached when today's price is the highest price for the period under consideration, and zero being when it is at the lowest point.

Why does the stochastic work the way it does? The simple point is that after a strong rise or fall, prices tend to consolidate. During a period like this, if prices begin to close away from the extreme, the stochastic value will tend to change direction, indicating a potential turning point in the 'underlying'.

So, rather as is the case with the moving averages of share prices, stochastics are of particular significance where divergence occurs in what appear to be major overbought or oversold areas. This might occur, for instance, when the share price rises to a new high but the stochastic fails to rise in sympathy. This would be taken as a sell signal. Similarly when the price falls to a new low but the stochastic doesn't, this could be taken as an imminent buying opportunity.

The way this is highlighted in a chart is for a stochastic value to be smoothed using the average of several period, with the unsmoothed stochastic value then acting as a 'trigger' when it reverses course by sufficient to cross back over the smoothed average.

Peter says

I would be the first to admit that I am not - in my own investing - a great devotee of using technical analysis. That's essentially because I am not a short term trader and because I think I can get more of an edge using other approaches.

As a sometime statistician, however, there is something in the statistical logic of stochastics that appeals to me. The calculations are complicated and to make best use of them you need to subscribe to one of the chart software services like Sharescope or Updata, which handle the calculation and graphing of these and other relevant indicators with ease.

Most technical analysts would, however, suggest that stochastics, though useful, should be employed in conjunction with a range of other indicators in order to confirm a decision suggestion by them. Alternatively, identify an opportunity using stochastic analysis, and then look to other indicators to confirm this view.