Understanding True Concepts of Overbought and Oversold

forex market overbought oversold concepts

The term overbought and oversold is heard on the wires and commentaries and is an integral part of any trader’s armoury, but it is a common theme that when I question traders or present at seminars, the common most tools that are used to quantify this are the Slow Stochastic and RSI. This re-enforces a common theme in technical analysis, in that the long established mantras continue to feed through as each new set of traders enter the markets. The common philosophy is that once value goes beyond 80 and 20 for the Stochastic and 70 and 30 for the RSI, the market is overbought or oversold.

However, when these concepts are actually put to the test the laziness and inaccuracy of this concept highlights how, not only can this cause problems by exiting good trend following trades too early, it can also encourage the ruinous strategy of trying to catch tops and bottoms with such a theory. If the common perceptions of what constitutes divergence are also applied the results can be even more damaging. The theory here states that negative divergence is being created if price makes new highs for the trend but the indicator of choice fails to do so and subsequently turns down (a failure swing). The reality is that all but the most explosive of trends that move into blow off tops or capitulation bottoms, normally diverge, and in many cases can continue to do so throughout much of the trend.

The inherent flaw when thought about is somewhat obvious. The indicators of choice have limits of scale and the market in theory can move anywhere and often can do it very quickly. This flaw becomes apparent when looking at the Slow Stochastic in particular. The calculation looks at the relationship between the close and its vicinity to the high and low. In any trend, by default closes will tend to err towards the high or low depending on what the trend is. Therefore once the indicator gets into its overbought or oversold zone, it often takes just one or two bars of price movement in the opposite direction to make the indicator cross over. This is highlighted in the recent Eurodollar fall as shown in Figure 1. Note how a two day correction makes the indicator cross up before the market slumps once more and then on the next subsequent up bar, the indicator posts divergence. This divergence then continues on 4 more occasions in quick succession before finally the market does post some modest gains. This highlights how a trader can feel that the market remains irrational longer than they can remain solvent.

forex market overbought oversold concepts highlighted

The RSI presents a somewhat different dilemma as typically there is no crossover feature meaning the trader is simply left with 3 alternatives. Track divergence, track any move out of the extreme, or simply a change of direction when in that extreme. Again looking at the Eurodollar in Figure 2 none of these scenarios would have been satisfactory as price moves out of the extreme and changes direction on the same bar just before the downtrend resumes, and then in a similar vein to the Stochastic, posts multiple divergences that eventually come good. In fact on the day new lows are posted for the trend, the Rsi is very close to 50 in value. In my testing experience the lesser of the evils is to reference a simple change in direction when the indicator has been beyond 85 or 15.

forex market overbought oversold concepts Eurodollar

So what are other solutions? One obvious alternative is place a non linear calculation on non linear data. The preferred method of choice for many traders is the Macd. Whilst this is an improvement on the others there are still problems associated with how extremes are qualified. Normally it is necessary to create a point of reference from previous trends and the extremes that they created. However, this is a somewhat haphazard approach and on historical data can provide relatively few sample sets. However, the concept of divergence normally produces a lower number of false signals, and if the trader is prepared to be patient and have higher risk, waiting for the Macd to either crossover or switch back through the zero line is another option. Figure 3 shows how, although the timing was not perfect the trade may have survived. Certainly there were less false dawns.

forex market overbought oversold concepts

My preferred solution is to take a similar concept to the Macd but in a more sophisticated way. This uses what is called the Kase Peak Oscillator@. In Cynthia Kases excellent book, Trading with the Odds she explains her complete trading methodology more comprehensibly, but on a more simplistic basis, this study takes a multiple number of moving averages, creates an in analysis of trend cycle length, and includes a calculation for volatility skew. What this does is create an oscillator that has superficially similar properties to the Macd. However, in addition it places a line which represents 2 standard deviations around that. Her theory states that once momentum has moved beyond 2 standard deviations of its own momentum, mathematically we are now at an extreme. If a trend cycle is being fulfilled (65 bars), then what she calls a Peak Out occurs (the purple line). This mean that this is the extreme or there is one more extreme due in this trend cycle.

forex market overbought oversold concepts KPO@ and the KCD@

Returning to the chart and figure 4 we can see the KPO@ and the KCD@. This latter indicator tracks the Momentum change in the oscillator itself. The Peak Out or oversold situation is flagged a little early and is followed shortly afterwards by the KCD switching to over zero. This means that the trace of momentum is now positive from negative. We are now in the situation where this is the low or there is one more low in this trend cycle of 65 bars. The 2 vertical lines represent that figure. As can be seen in figure 4, whilst its exact point is slightly off balance, the trader is now in the position of linking this analysis together and looking for other timing indicators to confirm the trade. There are two black arrows under the candles. These use a very simplistic and my original method of quantifying divergence. It uses a standard Rsi and states that price is making a 9 bar low but the Rsi is making a 3 bar high. There is no value placed on what the value of the Rsi actually is. Two days later the cycle is complete with the trade being false if price breaks the absolute low of the trend.

Shaun Downey