My new “Market Timing Knowledge Base” is almost ready to launch, and I’m excited to share my methodology’s documentation. The goal is to further assist traders in applying the real-time intraday timing signals and targets that I describe daily in my Roadmap Alerts and Live chaRTroom… Currently, my developer is waiting for a software patch that can restrict access from non-subscribers. But I’m making this RSI unit available, and look forward to any feedback. A brief video overview of my If Then Timing Windows and Bias Parameters is also available here.
RSI terminology and descriptions
Relative Strength Index (RSI) was created by famed market technician Welles Wilder. I was fortunate to have interviewed him during the earlier stages of identifying my own patterns and indicators — his insights into price behavior were very instructive.
I regard RSI as a gauge of buying pressure or selling pressure. This is not the same as price momentum, which an indicator such as MACD will express. So, RSI is the effort that sponsorship is applying, and MACD is its production. Relatively excessive effort (RSI) with diminishing productivity (MACD) is closer to being reflected by price action reversing direction.
RSI is a lagging indicator, but my applications of it at least narrow the likely price behaviors to follow. First, some terminology:
|Overbought / Oversold||Value above 80 or below 20, respectively.|
|Simultaneously overbought / oversold||Both 1-minute and 3-minute RSIs overbought or oversold during the most recent instance of the current trend’s price extreme.|
|Divergence (negative or positive)||One or both of the 1-minute and 3-minute RSIs not returning to overbought/oversold, despite price extending to a new extreme.|
|Persistently overbought/oversold||3-minute RSI remaining overbought or oversold while price is trending, although 1-minute RSI has left overbought or oversold territory momentarily during the trending|
|RSI correction||Overbought or oversold that retraces back to the midrange of the indicator’s values (toward 50) can fulfill a correction without price also retracing|
And now, examples of predictive RSI behaviors…
Simultaneously Overbought / Oversold
Little is more universally enticing to fade, than an “overbought” or “oversold” market. RSI probing above 70 or below 30 is often interpreted as the trend nearing its end. Which makes sense, because RSI becomes so extended when pretty much all available sponsorship has expended its energy to produce the overbought/oversold reading.
But if we assume the trend’s sponsorship is “strong-handed” — which I do assume, because it was able to RSI into overbought or oversold — then the next reversal will be sponsored by “weak hands.”
Does that matter? Not always. Not unless 1-minute RSI is simultaneously overbought or oversold along with the 3-minute RSI. Then the weak-handed countertrend move is doomed to failure, so its origin at the price extreme will be retraced (see the nearby chart example).
Countertrend moves can become quite extended before eventually retesting their overbought or oversold price extreme. An overnight retest would qualify as neutralizing the outstanding retest, although these are often retested, too. Not all retests are neutralized, but they’ll be a likely attraction if/when a countertrend move returns to within its proximity.
A price reaction can take simultaneously extended RSIs out of overbought or oversold territory, and their price extreme requires a retest. Simultaneously extended RSIs on the retest — even if both 1-min and 3-min RSIs are making higher lows / lower highs — will still require an eventual retest. Be careful, because when the setup doesn’t produce a temporary reversal, the trend tends to extend aggressively.
So, simultaneously overbought/oversold 1-minute and 3-minute RSIs require an eventual retest. Exceptions are the product of weak-handed or diluted sponsorship. These occur during the open and closing action, and during the noon hour. Also, the knee-jerk reaction to a headline is by definition the product of weak-handed sponsorship, so its price extreme would not require retest.
Persistent Overbought / Oversold
Think twice, or 4-5 times before fading “persistently obt/sld RSI.” Specifically, the 3-minute RSI, when price is trending relentlessly, steeply and substantially, taking 3-min RSI deeper and deeper into overbought (or oversold) territory.
Betting against an extended trend is so alluring, and in this case, so wrong. The attraction is made all the more enticing by 1-minute RSI occasionally diverging (leaving overbought/oversold territory, and not returning to it when price makes a new extreme with overbought/oversold RSI). That’s just sponsorship refueling, and it usually precedes a new price extreme.
The 1-min RSI divergence can appear at or near the trend’s end if 3-min RSI, if 3-min is deteriorating the most since becoming overbought/oversold. Even then, be careful betting against the trend too soon. That’s often the second divergence’s recovery. The safest entry would wait for 3-min RSI to leave overbought/oversold — or, at least, allow little time for the entry to become profitable, or else exit.
5/12/2017 — Live example of Double Top break
7/19/2017 –Persistently 3-min RSI beats negative divergence in 1-min
8/8/2017 — Second divergence’s extension is the last
9/11/2017 — Shallow reaction to Persistent Overbought RSI
Divergence in RSI, not necessarily in price.
Positive divergence (or negative divergence) refers to the RSI reading, and NOT to the expected price resolution. Many traders will view a divergence as a reversal signal. It can be part of a reversal signal, with other elements. Similarly, it is a continuation signal in the absence of a reversal signal’s elements.
Think of RSI divergence as sponsorship calling for reinforcements to extend the move immediately. Reinforcements often do arrive, and the trend often extends. Otherwise, the trend’s stagnation becomes vulnerable to reversing direction.
12/19/2017 — Divergence didn’t end the trend
Ignoring consecutive divergences
Fading RSI divergences is enticing, because often it is immediately productive. That’s a very general statement. Its wide net catches a lot of brief reversals that fail and resume the prevailing trend — to a greater degree than the brief reversal. So, we’ll consider a divergence’s reversal to be valid if its interim extreme is probed.
Putting aside whether and how to trade any other setup that may be developing, we should take note of divergences that aren’t productive. More so, notice consecutive divergences that extend to a fresh extreme without first retracing beyond their interim extreme, or only overlap it by 2-3 ticks.
This can often be observed during the first half-hour of what becomes a trending session. That is among the most dangerous times to take a divergence at face value, and to assume a price reversal is imminent. Similarly, it can also be a missed profit opportunity.
Do you have an RSI setup for analysis? Please share it in the chaRTroom.