by Ray Rondeau
Today’s markets are more sophisticated, faster moving and intercorrelated than at any time in history. In this “new” investing environment, an increasing number of investors are looking for ways to compensate for these phenomenon and the accompanying increased levels of volatility. One popular approach that is often considered is using technical analysis. In our first segment of the “Classic Technical Indicators” series of articles, we are going to look at one of the more popular indicators for stocks and overall markets analysis: the relative strength index (RSI). This installment covers the basic aspects of this indicator; in future articles, we will also look at some of the more esoteric, or less known but equally insightful, applications for the more advanced investor.
As most experienced investors are well aware, what theory (by definition) forecasts to happen is often very different than what really happens. So perhaps the most useful aspect of these articles will be the insights derived from the extensive amount of backtesting performed to illustrate the specific examples. The goal of these articles is to reveal not only the “historical” returns with the traditional indicator settings but, more importantly, the return statistics and the specific parameters covering an array of optimized settings over a variety of important categories.
RSI is a pure price-based indicator (closing price), with a momentum oscillator classification. Investopedia defines the term oscillator as “a technical analysis tool that is banded between two extreme values and built with the results from a trend indicator for discovering short-term overbought or oversold conditions. As the value of the oscillator approaches the upper extreme value, the asset is deemed to be overbought, and as it approaches the lower extreme, it is deemed to be oversold.”
RSI focuses on the characteristics of price momentum (change and velocity of price) and the likely continuation or change of momentum from labeled overbought or oversold levels. It is formulated from measurements of the speed and size of past price moves, plotted as a single line within an oscillating zone. Securities with high RSI values can be interpreted as having overbought positions, while stocks with low RSI values can be interpreted as having oversold positions. RSI can and has been effectively used on major market indexes, exchanged-traded funds (ETFs) and individual stocks, typically with a user preference toward charts covering longer time periods.
The RSI indicator (lower pane) is applied to four different charts (bar charts) on the Dow Jones industrial average covering four time frames: weekly, daily, 60-minute & 15-minute. Although, RSI can be modified and applied for use on shorter time periods, such as for swing and day trading analysis, it is usually applied to longer term charts for position trades and longer term investment-based analysis.
Origin & History
J. Welles Wilder Jr., the originator of several popular indicators including average true range (ATR) & average directional index (ADX), first featured the relative strength index in 1978 in his book “New Concepts in Technical Trading Systems” (Trend Research). In that same year, RSI also appeared in a June 1978 article in Commodities Magazine.
Displayed here is the original calculation method that was first proposed and defined by Wilder. This RSI formula creates an oscillator that compresses the RS into a zone of 1 to 100 for identifying extreme readings.
- RSI = 100 – (100 / (1 + RS)) or 100 * (RS / (1+RS))
RS = Average Gains (over the period) / |Average Losses| (over the period)
Losses are treated as positive numbers: |absolute value|
The traditional RSI formula uses a simple straight average to calculate average gain and loss.
Wilder then developed a second form of the RSI calculation, often referred to as Wilder’s RSI; this figure is calculated using Wilder smoothing average (WSA). This is also sometimes referred to as a modified moving average (MA).
Wilder Smoothing Average (WSA):
Average Gains over the Period (smoothed) =
(((Average Gains MA) * (n-1)) + Current Bars Gain) / n
(((Sum of Gains MA – (Sum of Gains MA / n)) + Current Bars Gain) / n
Average Losses over the Period (smoothed) |Absolute value| =
(((Average Losses MA) * (n-1)) + Current Bars Loss) / n
(((Sum of Losses MA – (Sum of Losses MA / n)) + Current Bars Loss) / n
- n = (number or RSI periods)
RSI = 0
RSI is 0 when there are no gains over the referenced time period.
RS = 0 / X = 0 RSI = 100 - (100 / (1 + 0)) = 0
RSI = 100
RSI is 100 when there are no losses over the referenced time period. i.e. “dividing by 0”. (See RSI’s Positive Calculation Bias below)
RSI values can be calculated intra-bar by using the last price of the bar as opposed to waiting for the closing price of the most recent price bar being printed. Price bars in technical analysis are representations of a stock’s price movement that features the open, high, low and closing prices for a specific period of time or particular set of data. For instance, a one-minute bar represents the price data for one minute. Likewise, a one-day bar is represented by the set of quotes from that one day.
RSI Construction & Layout
The vast majority of investing and trading platforms have the RSI indicator incorporated in their software. The RSI indicator is normally strategically plotted under the main panel’s price plot for more intricate studies (which will be covered later). If volume is included in the chart, RSI is usually plotted above volume, which is traditionally in the bottom pane. Construction of the chart usually has an “insert indicator” selection and setting of the parameters.
The calculations and resulting RSI line are plotted automatically along the x-axis under the corresponding prices, within a vertical oscillating scale ranging from 0 to 100. Key horizontal levels often highlighted on the oscillator are traditionally 30 for oversold (green dotted) and 70 for overbought (red dotted). Some technicians will also place a less-conspicuous horizontal line at the midpoint level of 50. For easier recognition, key RSI-based signals can be automatically imposed directly on the actual price chart by a variety of methods such as ShowMe and PaintBar, which highlight a certain area on a chart.
ShowMe (TradeStation) studies enable a technician to quickly and easily identify a specific condition or technical signal right on the actual price chart, without having to look elsewhere or perform further cognitive analysis. The signals are generated automatically and identify each bar that meets a desired criterion by placing a user-selected symbol on the bar. In this example, our bar chart displays dots, indicating an RSI potential overbought (red) or oversold (green) level.
PaintBar (TradeStation) studies also enable a technician to quickly and easily identify a specific condition or technical signal right on the actual price chart, without having to look elsewhere or perform further cognitive analysis. These signals are also generated automatically and identify each bar that meets a desired criterion. With PaintBars, the areas are identified by painting the price bar a defined color. In this example, with our “line on close” price chart, changes in color highlight that the RSI is in a defined overbought (red) or oversold (green) area.
RSI Chart Setup Video
Click on image to play.
The RSI line itself has only one variable parameter, which is the number of periods that the indicator looks back on. (Note: Some advanced platforms will allow you to adjust the number of bars loaded and referenced, which can affect RSI’s values. See RSI Data Length Dependency covered later in this article.) The traditional (default) time period is 14 periods (days, weeks, months, etc.), the amount originally suggested by Wilder. This length was based on Wilder’s belief that 14 periods represented one-half of a natural cycle. Natural cycles were originally established and tracked for simplicity before computers existed. With this method, investors took 28 days (for one month) and divided it in half—for a half cycle (14 days up, 14 days down). This parameter can be easily adjusted by direct entry in a settings area or dialog box, which is illustrated in the video below.
RSI Parameters Video
Click on image to play.
Source: Created using TradeStation. ©TradeStation Technologies, Inc. All rights reserved. No investment or trading advice, recommendation or opinions are being given or intended.
Putting it all together, a basic chart setup with RSI might look something like this, with a lagging indicator, called moving average, plotted over the price in the upper pane and volume plotted in the lowest pane.
In this daily bar chart of Diamond Offshore Drilling, we have price plotted in the upper pane with a 200-period simple moving average (SMA). In the center pane, we have an RSI indicator with color-coded overbought/ oversold levels. In the bottom pane, we have volume, with a gradient color scheme reflecting the actual intensity of the price move displayed in each volume bar.
The most common and perhaps simplest use of RSI is for signaling potential buy/sell signals from overbought/oversold levels. These signals are realized when changes in momentum are observed from predetermined “extreme” levels and are confirmed with a crossing of the RSI line through the user’s accepted levels. Accepted levels are based on an individual’s preferences and can be adjusted and contoured to the volatility of the issue, the time frame and the number of desired signals a user would like to generate. Crossing below or above the predefined accepted levels, signals that the stock is either oversold or overbought, respectively. This concept will be covered in much greater detail in “RSI Advanced Interpretations,” the third part of this relative strength index series, and will include a bevy of examples.
Conventionally, RSI levels below 30 are considered oversold and a bullish signal is triggered when the RSI line crosses up through the 30 level from below.
Similarly, RSI levels above 70 are considered overbought and a bearish signal is triggered when the RSI line crosses down through the 70 level from above.
In this chart, we see a magnified area showing RSI’s line oscillating within its defined overbought/oversold levels. Note how the RSI indicator in the overbought/oversold levels does not change color until after the actual bar closes in these areas. This is where and when we would have a confirmed reading, as a moving price could reverse and pull RSI back out of the zone. This is the reason for the yellow bleeding (blue arrows) in these areas from the previous connecting days RSI value.
RSI levels above 50 can be interpreted as bullish, because (mathematically) average gains are larger than average losses (absolute value).
RSI levels below 50 can be interpreted as bearish, because the absolute value of average losses would be exceeding those of the average gains.
In the area marked “A,” average gains are larger than the absolute value of average losses and RSI is still not in the predefined overbought area. If RSI is still rising, this could be a bullish interpretation. In the area marked “B,” the absolute value of average losses is now larger than the average gains and RSI is still not in the predefined oversold area. In this case, if RSI is still falling, this could be a bearish interpretation.
In this first part of our four-part series on the RSI indicator, we have looked at the indicator’s origin, its basic individual components, how to construct it and plot it and its most traditional and basic interpretations. In our next segment, “Inside the Numbers,” we will take a deeper look at a number of RSI calculation oddities that are not widely published or known to many of RSI’s most advanced users. In the third part, we will look at a number of other advanced applications and how to strategically contour the RSI indicator to an individual issue’s characteristics and your specific methodology. Ending the series will be the much anticipated backtesting and optimizations article. In this article, we will look at a vast array of detailed statistics on where, when and with which parameters RSI has performed best (and worst) on every single stock (and industry) in the S&P 500 over a 13-year time study.