Talk:Relative strength index

Overbought
I am going to let this stay for now, because it is what the amateur investor commonly believes. However, really, overbought is complete nonsense. Just when a strong trend is starting is not the time the speculator should be considering selling. This is when a speculator should look at the strength of the trend and see about buying. There is no such thing as overbought. Prices can always go higher.

As I told you before, stocks are never too high to buy or too low to sell.
 * &mdash;Jesse Lauriston Livermore

Ranges
The levels 70 – 30 or 80 –20 are purely arbitrary. The RSI indicator like most indictors smoothes out prices so that the analyzer can more clearly see the trend.

Increasing the length of periods RSI encompasses, smoothes out the volatility, shortening it increases it. To make it useful adjust it so that it fills your graph during normal times. During extreme moves the indicator will want to break out of your monitoring window. During these extreme trends the rules for congestion areas do not work. A reading near 100% does not mean the move will not continue. It fact it may indicate just the opposite. Simularaly a reading near zero is not a buying sign.

"When I buy stocks for a rise I like to pay top prices and when I sell I must sell low or not at all.
 * &mdash;Jesse Lauriston Livermore

False buy or sell
There is also no such thing as a false signal. The RSI signals are correct; it is just the practitioners opinion on what the reading means that may be in error. When the reading is at 70 or above in normal market conditions many poeple assume that the equity is overbought and should be sold or when its at 30 or below the reverse. But this is not true. The readings are subjective to every user. Many equities will actually climb much higher when levels of 70 or above are reached or continue to go much lower when the indicator reads 30 or lower.

Alternative intepretation using reversal patterns
In his book, Trading for a Living, Alexander Elder suggested looking for reversal patterns in technical indicators such as RSI. This is an interesting variation on the 'magic number' overbought and oversold application. For example, a breakout from a double bottom on the RSI chart will often precede the breakout of a double bottom on the price chart. —Preceding unsigned comment added by 125.27.4.157 (talk) 01:15, 29 April 2008 (UTC)

Division by Zero
The equation for RS looks like incorrect in case of continual growth, when denominator becomes a zero:


 * $$RS = { EMA[N] \; of \; U \over EMA[N] \; of \; D }$$

Starkest (talk) 10:33, 13 July 2009 (UTC)

Hmm, Yes you are right, but as denominator tends to zero, RS tends to infinity. And when RS tends to infinity, RSI tends to 100. So it can be said it is well defined. Just my opionin. --83.18.219.251 (talk) 00:31, 3 January 2011 (UTC)

Don't understand
OK, I'll readily admit to being not great at math, but I do have a minor in mathematics, and I'm wondering how this

$$ RSI = 100 - 100 \times { 1 \over 1 + RS } $$

can be rewritten as

$$ RSI = 100 \times { EMA[N]\;of\;U \over (EMA[N]\;of\;U) + (EMA[N]\;of\;D) } $$

The stockcharts RSI has it written as

$$ RSI = 100 - { 100\over 1 + RS }$$

I guess maybe I'm probably missing something obvious here, as usual. 1 + RS can be rewritten as

$$ { (EMA[N]\;of\;U) + (EMA[N]\;of\;D) \over EMA[N]\;of\;U  } $$

so $$ 100 \times { 1 \over 1 + RS }$$

can be rewritten as

$$ 100 \times EMA[N]\;of\;U \over(EMA[N]\;of\;U) + (EMA[N]\;of\;D)$$

I don't see how the current equation got rid of the 100 .... 67.189.104.78 (talk) 00:57, 25 July 2008 (UTC)

You're right, the rewritten version is wrong, so I've deleted it from the article.93.96.236.8 (talk) 13:06, 10 August 2010 (UTC)

different formulation
I find different formulation of RS, and U and D:
 * $$U = max(0, close_{now} - close_{prev})$$
 * $$D = max(0; close_{prev} - close_{now})$$

Or even simpler as:


 * $$U = +max(0, close_{now} - close_{prev})$$
 * $$D = -min(0; close_{now} - close_{prev})$$

It is of course the same as in article, but shows, that U and D, are just positive and negative parts of close_new - close_prev, which makes it easier to visualise and understand. —Preceding unsigned comment added by 83.18.219.251 (talk) 00:29, 3 January 2011 (UTC)

Requested move I

 * The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section. 

The result of the move request was: page not moved, per insufficient support below. - GTBacchus(talk) 19:53, 5 January 2012 (UTC)

Relative Strength Index → Relative strength index –

Per WP:MOSCAPS ("Wikipedia avoids unnecessary capitalization") and WP:TITLE, this is a generic, common term, not a propriety or commercial term, so the article title should be downcased. In addition, WP:MOSCAPS says that a compound item should not be upper-cased just because it is abbreviated with caps. Lowercase will match the formatting of related article titles. relisted --Mike Cline (talk) 16:09, 23 December 2011 (UTC) Tony   (talk)  01:44, 16 December 2011 (UTC)


 * Oppose. I cannot find a single source (in a casual search) that presents it in lowercase. Even the inventor used title case. We should go by how the sources refer to it. ~Amatulić (talk) 01:50, 16 December 2011 (UTC)


 * Support – It's hard to imagine how the editor is above is unable to find it lower case in sources; it is not uncommon. Of course, a naive search will often show snippets from titles and headings, where it is capitalized.  I've cited some sources that have it lower case, in the article, in case there's any doubt (I did that before noticing this RM, after doing similarly on true strength index).  I'll happily point out more book sources, like  and  and  and  and  if anyone needs them; there are tons more.  Dicklyon (talk) 05:08, 16 December 2011 (UTC)
 * WP:WEIGHT applies here. Not only do most sources I see in a naive search use it as a proper noun in prose (not just in titles), but the guy who invented this index does as well. Certainly you can cherry-pick sources that use the lowercase version. The primary source (the inventor) should carry significant weight.
 * See Talk:Bollinger Bands as another example, where John Bollinger himself writes that he deliberately created the term as a proper noun, with uppercase "Bands" in the name. We should respect that choice, and accord other authors similar respect.
 * The same can't be said for everything; for example the term "average directional index" is used (I recall) by its creator Welles Wilder just once, in all caps, and from then on he refers to it as ADX, and indeed most sources spell it out in lowercase. ~Amatulić (talk) 23:46, 16 December 2011 (UTC)
 * That's exactly the wrong weighting. Everyone who invents a name and an acronym capitalizes his own baby.  The test is whether it is treated as a proper noun by reliable sources that don't have that conflict of interest.  Dicklyon (talk) 17:08, 22 December 2011 (UTC)
 * And referring to the author of a creative work as someone with a "conflict of interest" is just silly. You haven't yet offered a rationale for your position that an invention that nobody has thought of before and is given substantial coverage in reliable sources isn't a creative work. We name creative works the way the author named them.
 * An inventor's interest is to get his work well known, recognized, and treated as very special. That's not the same interest as encyclopedia writers have.  And in very many cases we specifically do not capitalize things that their authors capitalized.  I can probably find you a few hundred example wikipedia articles if you care.  Dicklyon (talk) 00:53, 24 December 2011 (UTC)
 * Oppose. The proposed change will only create confusion: The Relative Strength Index, or RSI, as first described by Welles Wilder in 1978, is a normalized comparison of positive and negative price changes for a single time series, while a relative strength index is a relative performance comparison of two time series. I note that the proposer of this change has proposed a number of other changes to Technical Analysis articles that defy common usage and should not be approved. I will take those up on his talk page. MarketTechnician (talk) 15:08, 22 December 2011 (UTC)
 * Evidence – whatever happened to evidence? What good does it do to assert that lowercase will be confusing, or that it's a proper noun, without good evidence.  Take a look at books]; if you use a search with "the" in front to reduce the hits and snippets of titles and headings, and follow with "and" to eliminate the contexts where what follows immediately is "(RSI)" because the caps are for the acronym definition, then you find about 2/3 of the hits are lower case; other searches give similar results.  Can this be reconciled with capitalization in light of MOS:CAPS?  No, it can't.  Compare with Amatulic's inability to find a single source with lowercase, and MarketTechnician's familiarity with what it is but no evidence about usage, and what have we left in terms of opposition?  Nothing.  Dicklyon (talk) 05:23, 24 December 2011 (UTC)
 * Facts? How about checking the source material? J. Welles Wilder uses "Relative Strength Index" in his 1978 book that introduced the indicator and Robert Levy used "relative strength" in his pioneering 1968 book on the topic. The first work refers to the topic of this article, while the second refers to strength in relation to other time series. MarketTechnician (talk) 22:03, 27 December 2011 (UTC)
 * Yes, we accept all that. But authors usually capitalize their own stuff for emphasis.  The criterion for determining whether something is a proper noun is whether other reliabl and independent sources capitalize it.  From the evidence I presented, not even half do.  Dicklyon (talk) 22:08, 27 December 2011 (UTC)
 * All the references on the first four pages of a Google search for "relative strength index" are "Relative Strength Index" except for one that appropriately refers to comparative relative strength. MarketTechnician (talk) 23:07, 27 December 2011 (UTC)
 * That is Flat Out Wrong (FOW). Google search is not the best tool for this (books is better, as more of what it finds are reliable sources), but let's look at them anyway.  Several like this one show up with "Relative Strength Index" capitalized in the snippet, since it's a heading; but when you look, they use "relative strength index" lowercase in sentences.  Others, like this one use "Relative Strength Index (RSI)" once or twice to define the acronym, and then never use the term again; this is zero information on how it would appear in a sentence.  But since these are both in Investopedia, we can count them both as lowercase.  Here is another of acronym-definition only type.  When you go to books, you find lot more in sentences, quite frequency lowercase (about 2/3); did you ignore this book search that I linked above?  Dicklyon (talk) 04:04, 28 December 2011 (UTC)


 * Oppose per Amatulic and MarketTechnician's comments. There are similar terms thrown around such as IBD's relative strength (completely different) and the generic sense of relative strength. II  | (t - c) 02:16, 28 December 2011 (UTC)
 * Nobody is confusing the relative strength index with relative strength or anything. Do you have any reasoning based on actual guidelines, evidence, or other info?  Because those guys didn't. Dicklyon (talk) 04:04, 28 December 2011 (UTC)
 * What exactly is relative strength index going to be confused with? And I'm curious to know why some editors here seem not to have looked at Dicklyons's linked searches. Tony   (talk)  07:26, 28 December 2011 (UTC)
 * Relative strength is a generic term that means something completely different than Relative Strength Index. The confusion argument, I agree, is weak, and could be solved by a hatnote in each article. As to searches, one can cherry pick them all day long. Dicklyon argues, nonsensically, that one should ignore any usage of the uppercase version to define the acronym rather than using it in a sentence. That's exactly how this article uses the term. ~Amatulić (talk) 14:27, 28 December 2011 (UTC)
 * But the evidence that I cited specifically avoids counting "relative strength". If you agree that the term is most commonly used as in this article just in defining the acronym, then you should be able to see that that's a style issue; our style doesn't do that (MOS:CAPS says: "When showing the source of an acronym, initialism, or syllabic abbreviation, emphasizing the letters that make up the acronym is undesirable" and "do not apply initial capitals in a full term that is a common noun just because capitals are used in the abbreviation".  The lead sentence says how we distinguish proper from common nouns, using sources in the way I have described: "words and phrases that are consistently capitalized in sources are treated as proper nouns and capitalized in Wikipedia."  Which part of MOS:CAPS do you propose we ignore?  Dicklyon (talk) 14:59, 28 December 2011 (UTC)
 * Kirkpatrick and Dahlquist in “Technical Analysis”, likely the most authoritative modern resource on the subject, capitalize “Relative Strength Index” and do not capitalize “relative strength”. Why is it that non-technicians wish to set usage for technicians when market technicians are clear about the usage they desire? You can argue all you want, but preferred usage by market technicians is “Relative Strength Index” and Wikipedia ought to reflect reality. MarketTechnician (talk) 15:42, 28 December 2011 (UTC)
 * Well that gets back to a point that I've made many times: quite often, specialists in a field capitalize their stuff, while reliable secondary sources don't adopt that stuff as proper nouns, and use lower case instead.  At WP, that's our style.  Here, the specialists are the market technicians; theirs is exactly the input that we need to discount in deciding whether the term is consistently capitalized in sources.  Dicklyon (talk) 16:12, 28 December 2011 (UTC)
 * Wikipedia is of the world, not a world unto itself. And this is not a discussion, it is simply a couple of pedants repeatedly asserting their views while ignoring all others. I can’t cite the appropriate chapter and verse, but this can’t be the sort of civilized discourse that Wikipedia advocates. MarketTechnician (talk) 16:48, 28 December 2011 (UTC) -- who is retiring from the fray.


 * Please, MarketTechnician, be mindful of No personal attacks and Assume good faith.


 * That wasn't a personal attack; rather it was my honest observation of what I felt was going on. I linked "pendant" so the intended meaning would be completely clear as it really felt that way. MarketTechnician (talk) 22:44, 28 December 2011 (UTC)


 * Guidelines are called guidelines because they aren't policies, they aren't rigid rules. Rather, guidelines define our best practices, and those should be applied with a modicum of common sense rather than strict adherance.


 * In my view, Dicklyon's two main arguments strain the application of common sense to the breaking point, namely (a) that we should ignore the most reliable sources in a specialized field just so that we can rigidly apply a style guideline that clearly does not fit, and (b) that what an author calls a creative work, which nobody had created before and is innovative and useful enough to have significant reliable source coverage (regardless of how those sources cover it), somehow has zero weight compared with what other authors call their creative works.


 * There's a hole in our style guideline on this topic. That does not mean that we should force-fit all articles to a guildeline that doesn't fit; rather, we should discuss filling the void in the guideline. That's more appropriate for the village pump, however, not this talk page. ~Amatulić (talk) 21:15, 28 December 2011 (UTC)


 * If by "the most reliable sources" you mean primary sources, then yes, it's WP policy to pay them less attention than independent secondary sources. I'm not saying ignore them, just discount them appropriately; that is, give them less weight, especially on things like caps style, where their author clearly has an interest unrelated to our interest as WP editors.  Dicklyon (talk) 02:37, 29 December 2011 (UTC)


 * By most reliable I meant what MarketTechnician referred to above. ~Amatulić (talk) 05:46, 29 December 2011 (UTC)
 * I'm not sure what references we are talking about. If you have figured that out please post links here. Thanks. Joja  lozzo  22:40, 29 December 2011 (UTC)
 * By "primary sources" I meant the ones he mentioned by Welles Wilder.  He has mentioned others by market techs, which are secondary, somewhat independent of the primary, but still represent "insider" usage as opposed to general usage.  I'm not sure which he was referring to either.  Even among insider books, capitalization is not consistent; would he deny that this and this are good sources? Dicklyon (talk) 23:49, 29 December 2011 (UTC)
 * Well, yes I would. Mr. Wilmott is a very bright guy, but he knows nothing about TA; he even gets the formula for RSI wrong, indeed completely wrong. As for the other, I have not heard of it and I have one of the largest TA libraries in existence. It looks like some sort of secondary-market compilation much like the SFO compilation you cited in another discussion. So, no, those are not very good sources in my view. MarketTechnician (talk) 14:33, 30 December 2011 (UTC)
 * The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

Validity testing
Are there no statistical studies of the validity of these putative signals? If there are, it would certainly be appropriate to cite them.

I.E.: What proportion of times that a signal occurs, does the expected response actually occur?

All that are currently cited are the claims of the authors of the indices.

Hsfrey (talk) 22:43, 11 March 2012 (UTC)


 * If it were that simple, then everyone would be using it the same way.
 * I doubt there are any statistical studies because indicators like this do not have a universal interpretation. There are an infinite number of possible input parameters one could use, on an unlimited number of time frames, and countless ways one could define a "signal" &mdash; and even then, the signal may be just one factor in a decision to go long, go short, or exit a position. For me, RSI has a high reliability when used correctly (and if you poll successful traders who use RSI of what "correctly" means, no two answers would be the same). ~Amatulić (talk) 00:56, 12 March 2012 (UTC)


 * So it can't be tested? or it just hasn't been tested? If the first, scientists would consider it to be a pseudo-science.  If the second, you have to ask yourself "Why not"? Smallbones (talk) 12:10, 3 April 2012 (UTC)


 * Sure, it can be tested. My own use of it is based on it passing the tests developed. Pick your input parameters, pick your time frame, define the conditions to set up a trading signal, and test it over 3000 historical trades. That's what traders do all the time.


 * There's nothing "pseudo-science" about a digital filter, which is what most technical analysis indicators are. Interpretation of the output, however, takes some knowledge about market movements. ~Amatulić (talk) 21:20, 4 May 2012 (UTC)

Requested move II

 * The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section. 

The result of the move request was: page moved -- JHunterJ (talk) 02:21, 2 May 2012 (UTC)

Relative Strength Index → Relative strength index – Re-nominating like Tony's previous RM that only had me supporting it. The opposition based on a claim of "proper name" made no sense. The term is far from consistently capitalized in sources. See book sources, like and  and  and  and. Even inventor Wilder's colleague Blau, inventor of the "true strength index", calls it "Wilder's relative strength index". Its capitalization is an outlier in its category. Dicklyon (talk) 23:27, 24 April 2012 (UTC)


 * Support as re-nominator. Dicklyon (talk) 23:27, 24 April 2012 (UTC)
 * Support—obviously, on the basis of sources, WP's own house rules, and consistency with other article titles. Could those who oppose come up with arguments that aren't quite so lame this time? Think harder. The last RM had to be close by Bacchus on the numbers, but shining the flashlight on the qualitative aspects of the opposes makes one feel queasy. Tony   (talk)  01:42, 25 April 2012 (UTC)
 * Support, per Longman Dictionary of Financial Terms and WP:CAPS ("do not capitalize second and subsequent words, unless the title is a proper noun"). Kauffner (talk) 06:35, 25 April 2012 (UTC)


 * The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Some proposed changes
Information to be added or removed: Adding an external link to https://www.nelhajj.com/rsi-indicator/ which contains statistics on Relative Strength Index signal frequency and accuracy. This information is not present anywhere on the web as it presents statistics of RSI over several years and thus is helpful to anyone interested in learning about the relevance of RSI in the present day Naimelhajj (talk) 18:22, 3 July 2020 (UTC)


 * We don't link to self-published sources, as they're not considered reliable enough. – Thjarkur (talk) 17:22, 4 July 2020 (UTC)