Apple IncEdukacja
Are you using RSI correctly?
You know I love doing little stock market math experiments. Here is one with RSI!
Have you over shorted a stock that was “overbought”, only to incur losses? And longed a stock that was “oversold” only to, you guessed it, incur more losses?
Yeah, same. In fact, I have completely abandoned the use of RSI and Stochastics in my day to day trading, only using them on a theoretical level to fuel my probability models. But the truth is, its not that RSI isn’t useful. And its not that it doesn’t work. In fact, it works very well. The problem stems from a lack of understanding of this indicator.
I happened to stumble across an odd phenomenon with Stochastics and RSI. If you read my ideas, you know I trade based on probabilities. And how my probabilities are formatted, are they look at multiple different statistical based technical indicators, such as RSI, Stocahstics and Z-Score, and compare these values to historical incidents of similar values. I do this to plot out the behaviour of equities when they were at these like values, and make determinations of what is likely to happen based on what previously has happened in these instances.
One thing that I noticed and picked up on is that, things that traditional wisdom said were ‘bearish’ indications, such as a stock being “overbought” (as evidenced by an RSI over 70%), actually tended to correlate to extremely bullish sentiments and bullish stock behaviour. Things that were considered more bullish, such as a stochastic being oversold, actually tended to correlate to a stock being extremely bearish.
What my assessment of probabilities made me realize was that, instances where conventional wisdom tells you to short a stock, actually are misguided. The same goes for where wisdom tells you to long a stock. And so it made me wonder, what is going on? And me, being a logical person, decided to look into this. And the answer really was pretty simple and extremely obvious. In fact, you just google “What is RSI” and “Stochastics” and the answer slaps you right in the face. These are “momentum” indicators. And momentum indicators aim to assess the Strength or weakness of a stock’s price. The higher the RSI or Stochastic, the more strength and the more momentum. Inverse for the lower RSI.
This makes more sense when we consider how RSI is calculated. RSI is a function of average gain divided by average loss. This yields the RS factor, or relative strength of a stock. As stock price increases, so does the RS factor as its gaining in value. This signals an influx of buying and thus “high momentum” as buyers step in to keep up this momentum. As buyers fade, RS declines and thus the strength declines.
Its very obvious if we simply look at RSI on a chart. Let’s use AAPL
We see in this example, the RSI plateau’s in “overbought” territory despite the fact that the stock price continues to rise. Why does this happen? Because buyers are stepping in to keep up or surpass the average loss. There is high “momentum” as buyers are stepping in. So whereas conventional wisdom tells us to short on RSI’s over 70, this is actually contrary to what RSI is telling us. In this example, RSI is telling us that there is a lot of frenzied buying. There is high momentum and buyers are keeping up with sellers. And this is an ideal time to get long to keep up with momentum. However, we can see that later some bearish divergence occurs:
Why does this happen? Because the average gain is not keeping up with the average loss despite what we may be seeing in the chart, and thus we have a declining Relative Strength signaling weakness. Buyers are weakening and sellers are starting to predominate.
Let’s look at a short example:
In this chart, we see RSI is plateauing in “oversold” territory despite the continued decline in stock price (shown in the white box). We then see a characteristic “bullish divergence” being created as stock price remains relatively static (shown in the blue box with pink trendline). Why is this happening? Well, in the white box, sellers are keeping up or surpassing the average gain. Thus, this is an area that you would short, to keep up with the bearish momentum. The bullish divergence tells us that sellers are starting to fall behind (average gain is starting to surpass average loss) and signals us to either exit our short positions or wait for the buying momentum to pick up for us to get in long.
Well more conventional trading rhetoric says that oversold RSI should lead to an aggressive snap up and an overbought RSI should lead to some dramatic selling, there is actually no statistically significant relationship between RSI and a stock’s high and low. What that means is that RSI alone is insufficient to affect or have any meaningful impact on how high or low a stock goes. Using the ETF SPY (for the sole reason that I have a math model available to me to test this), the Pearson Correlation between SPY’s High and its 14 day RSI reading is 0.062, meaning absolutely no significant relationship exists. The relationship between RSI and SPY’s daily low is 0.069. Again, meaning no significant relationship exists.
If we look at it a different way and look at RSI in relation to the points moved in a day between open and the daily high, open and the daily low and the distance between the high and low, there again is no statistically significant relationship. The Pearson values are -0.224, 0.184 and -0.293 respectively. It is interesting to note though that the relationship between open to the daily high or low and the RSI have a negative relationship, albeit not statistically significant. This means that, as one increases, the other decreases. So, theoretically, as RSI increases, the move from Open to High would decrease and vice versa (meaning a high RSI could theoretically limit upside movement and a low RSI could theoretically limit downside movement). But this is not statistically meaningful in any way (we look for a value of 0.500 or greater for a significant relationship to occur).
Trading Implications
So what does this mean to you, the trader? Well, the implications can be summarized as such:
An RSI over 70, or “overbought” technically indicates bullish momentum. Its likely best not to short as you are working against the momentum. Its best to wait for the development of bearish divergence followed by confirmation (manifested by actual selling resulting) prior to getting short.
An RSI under 30 or “oversold” technically indicates bearish momentum. Its likely best not to long as you are working against the momentum. Its best to wait for the development of bullish divergence followed by confirmation (manifested by actual buying resulting) prior to getting long.
Plateauing of RSI means that there is high momentum (whether it be in an oversold or overbought range) and the sheer buying or selling volume is keeping that momentum going.
When used appropriately, RSI is an extremely powerful tool to help assess momentum. However, RSI alone is really insufficient to plan longer term or more prolonged trades. RSI has no statistically meaningful impact on stock behaviour and outside of its ability to assess current momentum, its best not to draw concrete conclusions on a stock’s likely trajectory based on RSI alone. RSI is best used by day traders. Its an okay reference to check the overall strength of a stock’s price and to look for possible divergences; however, at the end of the day RSI is reactionary and delayed and is neither prospective nor predictive in nature.
Have you over shorted a stock that was “overbought”, only to incur losses? And longed a stock that was “oversold” only to, you guessed it, incur more losses?
Yeah, same. In fact, I have completely abandoned the use of RSI and Stochastics in my day to day trading, only using them on a theoretical level to fuel my probability models. But the truth is, its not that RSI isn’t useful. And its not that it doesn’t work. In fact, it works very well. The problem stems from a lack of understanding of this indicator.
I happened to stumble across an odd phenomenon with Stochastics and RSI. If you read my ideas, you know I trade based on probabilities. And how my probabilities are formatted, are they look at multiple different statistical based technical indicators, such as RSI, Stocahstics and Z-Score, and compare these values to historical incidents of similar values. I do this to plot out the behaviour of equities when they were at these like values, and make determinations of what is likely to happen based on what previously has happened in these instances.
One thing that I noticed and picked up on is that, things that traditional wisdom said were ‘bearish’ indications, such as a stock being “overbought” (as evidenced by an RSI over 70%), actually tended to correlate to extremely bullish sentiments and bullish stock behaviour. Things that were considered more bullish, such as a stochastic being oversold, actually tended to correlate to a stock being extremely bearish.
What my assessment of probabilities made me realize was that, instances where conventional wisdom tells you to short a stock, actually are misguided. The same goes for where wisdom tells you to long a stock. And so it made me wonder, what is going on? And me, being a logical person, decided to look into this. And the answer really was pretty simple and extremely obvious. In fact, you just google “What is RSI” and “Stochastics” and the answer slaps you right in the face. These are “momentum” indicators. And momentum indicators aim to assess the Strength or weakness of a stock’s price. The higher the RSI or Stochastic, the more strength and the more momentum. Inverse for the lower RSI.
This makes more sense when we consider how RSI is calculated. RSI is a function of average gain divided by average loss. This yields the RS factor, or relative strength of a stock. As stock price increases, so does the RS factor as its gaining in value. This signals an influx of buying and thus “high momentum” as buyers step in to keep up this momentum. As buyers fade, RS declines and thus the strength declines.
Its very obvious if we simply look at RSI on a chart. Let’s use AAPL
We see in this example, the RSI plateau’s in “overbought” territory despite the fact that the stock price continues to rise. Why does this happen? Because buyers are stepping in to keep up or surpass the average loss. There is high “momentum” as buyers are stepping in. So whereas conventional wisdom tells us to short on RSI’s over 70, this is actually contrary to what RSI is telling us. In this example, RSI is telling us that there is a lot of frenzied buying. There is high momentum and buyers are keeping up with sellers. And this is an ideal time to get long to keep up with momentum. However, we can see that later some bearish divergence occurs:
Why does this happen? Because the average gain is not keeping up with the average loss despite what we may be seeing in the chart, and thus we have a declining Relative Strength signaling weakness. Buyers are weakening and sellers are starting to predominate.
Let’s look at a short example:
In this chart, we see RSI is plateauing in “oversold” territory despite the continued decline in stock price (shown in the white box). We then see a characteristic “bullish divergence” being created as stock price remains relatively static (shown in the blue box with pink trendline). Why is this happening? Well, in the white box, sellers are keeping up or surpassing the average gain. Thus, this is an area that you would short, to keep up with the bearish momentum. The bullish divergence tells us that sellers are starting to fall behind (average gain is starting to surpass average loss) and signals us to either exit our short positions or wait for the buying momentum to pick up for us to get in long.
Well more conventional trading rhetoric says that oversold RSI should lead to an aggressive snap up and an overbought RSI should lead to some dramatic selling, there is actually no statistically significant relationship between RSI and a stock’s high and low. What that means is that RSI alone is insufficient to affect or have any meaningful impact on how high or low a stock goes. Using the ETF SPY (for the sole reason that I have a math model available to me to test this), the Pearson Correlation between SPY’s High and its 14 day RSI reading is 0.062, meaning absolutely no significant relationship exists. The relationship between RSI and SPY’s daily low is 0.069. Again, meaning no significant relationship exists.
If we look at it a different way and look at RSI in relation to the points moved in a day between open and the daily high, open and the daily low and the distance between the high and low, there again is no statistically significant relationship. The Pearson values are -0.224, 0.184 and -0.293 respectively. It is interesting to note though that the relationship between open to the daily high or low and the RSI have a negative relationship, albeit not statistically significant. This means that, as one increases, the other decreases. So, theoretically, as RSI increases, the move from Open to High would decrease and vice versa (meaning a high RSI could theoretically limit upside movement and a low RSI could theoretically limit downside movement). But this is not statistically meaningful in any way (we look for a value of 0.500 or greater for a significant relationship to occur).
Trading Implications
So what does this mean to you, the trader? Well, the implications can be summarized as such:
An RSI over 70, or “overbought” technically indicates bullish momentum. Its likely best not to short as you are working against the momentum. Its best to wait for the development of bearish divergence followed by confirmation (manifested by actual selling resulting) prior to getting short.
An RSI under 30 or “oversold” technically indicates bearish momentum. Its likely best not to long as you are working against the momentum. Its best to wait for the development of bullish divergence followed by confirmation (manifested by actual buying resulting) prior to getting long.
Plateauing of RSI means that there is high momentum (whether it be in an oversold or overbought range) and the sheer buying or selling volume is keeping that momentum going.
When used appropriately, RSI is an extremely powerful tool to help assess momentum. However, RSI alone is really insufficient to plan longer term or more prolonged trades. RSI has no statistically meaningful impact on stock behaviour and outside of its ability to assess current momentum, its best not to draw concrete conclusions on a stock’s likely trajectory based on RSI alone. RSI is best used by day traders. Its an okay reference to check the overall strength of a stock’s price and to look for possible divergences; however, at the end of the day RSI is reactionary and delayed and is neither prospective nor predictive in nature.
For real-time updates and premium indicators, consider joining my group at: patreon.com/steversteves
Wyłączenie odpowiedzialności
Informacje i publikacje przygotowane przez TradingView lub jego użytkowników, prezentowane na tej stronie, nie stanowią rekomendacji ani porad handlowych, inwestycyjnych i finansowych i nie powinny być w ten sposób traktowane ani wykorzystywane. Więcej informacji na ten temat znajdziesz w naszym Regulaminie.