Strategy and Technical

Relative Strength Index [RSI]

The Relative Strength Index was devised by J Welles-Wilder in 1978 and continues to be one of the most widely used indicators.  Because the RSI number oscillates (or ‘moves’ in plain English) between 0 and 100 it is described as belonging to a group of indicators known as Oscillators. Its basic job is to describe the momentum (or strength of movement) in a particular direction. Specifically the RSI is a Momentum Oscillator. In basic terms it tells us if a security has tended to move more consistently and strongly in an upwards or downwards over the period being considered.

It is important to understand that the term ‘relative strength’ in this context does not compare the strength of a particular financial instrument or index against another. It actually compares, over a set period, for a single instrument the times where the price closed up against those times where it closed down. If the price has been generally moving upwards in a sustained manner it would indicate that the momentum has been in this direction and would be reflected by a higher relative strength number.  In contrast, if the market has been moving consistently downwards a low number for relative strength would be expected.

The RSI itself is an index of this Relative Strength number. It effectively calculates all of the possible outcomes as a number between 0 and 100. Because this index is range-bound it makes it far easier to use as a comparison tool.

Interpretation

The RSI is widely used to identify situations where a security is overbought or oversold. The RSI is considered overbought if the RSI is above 70 and a price fall could be expected. The RSI is considered oversold if the RSI is below 30 and an increase in buying and corresponding price increase could be expected.

As with all indicators interpretation of these numbers must be taken within context. If a particular security is experiencing a long sustained uptrend the RSI will probably be above 70 but is probably not overbought. Therefore this should not in this instance be taken as a signal to sell.

Where a security is generally travelling sideways with no particularly strong ongoing trend the RSI becomes a far more reliable indicator and an RSI oversold or overbought signal should be taking into account when making a decision on a trade.

The RSI becomes far more relevant when used in conjunction with other indicators, notably moving averages (especially utilising cross over points between moving averages of different time periods) and momentum.

Divergences

A commonly used technique with RSI charts is the analysis of divergences. This is where the graph line on the RSI is moving upwards whilst the price is moving downwards. (and vice-versa). A common interpretation of this is that momentum is gathering pace in the opposite direction to the price and a reversal of the price direction could be expected.

Let’s use a simple example of a pendulum representing price. It is as if the RSI is saying that, although the pendulum is still moving in a particular direction, it is reaching the end of its swing and there are now forces acting on it which means that we can soon expect it to turn around and head in the other direction.

Calculation

Nearly all trading software will calculate and chart the RSI but a simplified version of the formula and a basic example are given below.

Average Gain    =    Total of the gains for the past X periods/X

Average Loss     =   Total of the losses for the past X periods/X

X                =           The number of periods over which the RSI is being calculated

RS              =            Average Gain/Average Loss

RSI            =           100   -   [ 100 / (1-RS)]

To demonstrate let us take the following simple example of a share for which we have 14 days of closing price data and for which we want the 10 day RSI.

The first RSI number we can calculate with this data will use the information from Day 1 to Day 10 (10 period RSI). The days where the share closed up are isolated and the amounts by which it closed up on each of those days is then added together. The total is then divided by 10 to give an average. The same thing is then done for the days where the share closed down. Note that the absolute value for the movements down is used.

Using the formulae above, we can then calculate the relative strength on Day 10 and then the RSI itself. The 10 day RSI can then be calculated for day 11 using the movement in closing prices from Day 2 to Day 11. This continues day by day and the results can then be plotted on chart.

10 Day RSI	Closing Price	Gain	Loss	Average Gain	Average Loss	Relative Strength	RSI
Day 0	100
Day 1	101	1
Day 2	100		1
Day 3	102	2
Day 4	99		3
Day 5	103	4
Day 6	104	1
Day 7	105	1
Day 8	104		1
Day 9	103		1
Day 10	104	1		1	0.6	1.67	62.5
Day 11	106	2		1.1	0.6	1.83	64.71
Day 12	109	3		1.4	0.5	2.8	73.68
Day 13	102		7	1.2	1.2	1	50
Day 14	95		7	1.2	1.6	0.75	42.86

Note that if there are no losses in a period being considered, the RSI will be considered 100. If there are no gains, it will be considered zero.

Figure 1 shows a practical example of a 10 week RSI being displayed in conjunction with a weekly candlestick chart. Note the overbought signal in April 2010 where the RSI rises above 80. This is folllowed by a strong sell off.

Figure 1

What Length of time to use?

The standard length of time for calculating the RSI is 14 periods. The example above used 10 which would be slightly more sensitive – i.e. the most current closing period’s trade will have a bigger impact on the RSI. Conversely, if we took a 20 day RSI the calculation becomes less sensitive.

The calculation above used a day as a single period but any unit can be used (months, weeks, days minutes etc).

Summary

RSI is a very useful and widely used momentum oscillator which can identify potential reversals in price patterns or confirm ongoing trends. The commonest setting is that of a 14 period RSI.

It is best used in conjunction with other tools such as moving averages or momentum indicators.