RSI vs Stochastic
The relative strength index compares recent gains with recent losses over a chosen length. The stochastic oscillator compares the latest close with the recent high-low range instead. Both produce a nought-to-one-hundred reading from the same bars, and the second one is considerably twitchier.
Two bounded oscillators, both scaled from nought to one hundred, both read for the same things. They are almost always shown together, and the case for doing that is weaker than it looks.
What each one is
The relative strength index compares recent gains with recent losses over a chosen length and expresses the result on a fixed scale. The relative strength index covers it.
The stochastic oscillator compares the latest close with the recent high-low range. A close near the top of the range reads high; near the bottom reads low. Stochastic covers it.
Both are bounded, and that is their shared virtue. A reading means the same shape of thing on any instrument, which an unbounded indicator cannot offer.
Where they differ
What is being compared. Gains against losses in one; the close against the range in the other. Those are different questions that produce similar-looking answers most of the time.
How twitchy each is. The stochastic moves far more, because the recent range is a smaller and more volatile denominator than a run of gains and losses.
How many parameters each has. One length on the strength index. The stochastic has a lookback and two smoothing settings, so there is considerably more to fit and more to get wrong.
How each behaves at an extreme. Both pin. The stochastic reaches its extremes far more often, which means an extreme reading on it carries even less information than on the other.
Where they agree
Both read the same bars. Neither adds information from outside the price series, so agreement between them is a property of the arithmetic.
Both pin in a trend. A strong move holds either at an extreme for many bars, and reading that as overbought is the single most common error with both.
Both fail in a range. On this site’s shared series direction runs average 2.01 bars with a longest of 11, and short runs send either back and forth across any threshold you set.
And neither contains a stop. The ninetieth percentile bar range here is 1.101, and the invalidation belongs at structure rather than at a threshold.
Which one to use
Run the relative strength index when you want the steadier reading. Fewer extremes means each one is worth slightly more, and one parameter is easier to leave alone than three.
Run the stochastic when you want speed and will accept noise. It turns earlier at genuine reversals and also at a great many non-reversals, which is the same trade every faster tool offers.
Run one, not both. Two bounded oscillators on the same bars are one measurement displayed twice, and the confidence produced by seeing them agree is not supported by anything.
And when you want a genuine second opinion, use something that is not price — volume, or the structure of the chart. Anything computed from the same closes will keep agreeing with you.
Why the extremes mislead
Because an extreme means the recent bars were one-sided. That is a description of what already happened, and one-sided bars are exactly what a trend is made of.
And because both can stay pinned far longer than a position can be held. The reading has no time limit and your account does.
What to fix before either is useful
Write the signal down. Which threshold, crossed which way, held for how many bars. A rule you can state is a rule you can review; one you recognise on sight is one you will recognise after the fact.
Fix the parameters and leave them. Three settings on the stochastic is three opportunities to fit the past, and changing any of them after a losing run resets your sample to zero.
Decide what an extreme means to you. If it means a reversal is due, you will be wrong through every trend; if it means the move is strong, you have a different and more defensible tool.
And count the signals. On this site’s shared series direction runs average 2.01 bars, so any threshold rule fires far more often than a market actually turns, and the count is what tells you whether the rule can pay for itself.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, no title compares these two
directly — this pair is constructed from two subjects the corpus covers separately. Separately, the
relative strength index appears in 820 titles at a median of 5,021 across 529 channels, and the
stochastic in 184 at a median of 11,915 across 136. The counts come from site/corpus_count.py.
820 videos on one at 5,021 and 184 on the other at 11,915. Four and a half times the coverage for the steadier tool and less than half the audience per video — the less-taught oscillator draws substantially more interest per upload.
The answer to the question on that chart is that both are at an extreme because the same bars were one-sided. That is what a strong move looks like from inside — and fading it means betting against the only thing either indicator has actually measured.
When it fails
The failure is fading extremes in a trend, and both oscillators encourage it every bar. The reading reaches its threshold, which is presented as overbought, so a short is taken. The move continues and the reading stays pinned, which looks like an even stronger case. It is not a stronger case — it is the same observation repeated, because a sustained move keeps recent bars one-sided by definition. The position is added to as the evidence appears to mount, and the mounting evidence is the trend itself.
The second failure is running both for confirmation. They read the same closes.
A third is optimising the stochastic’s three settings. That is fitting.
A fourth is treating a threshold as an entry. It is a description.
A fifth is finding divergence by looking. Any two series diverge somewhere.
And a sixth is expecting either to lead price. Both are computed from closed bars.
Related
The relative strength index covers gains against losses. Stochastic covers the close against the recent range. And stochastic RSI covers what happens when one is run on the other.
These two are on more beginner charts together than any other pair, and the reason given is always the same — confirmation. They read the same bars. If they disagree it is a difference in smoothing, and if they agree it is arithmetic.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.