6.12

📈 Indicator Combos & Market Internals

Confluence over single signals: proven indicator combos plus market internals (TICK/TIKI, McClellan, TRIN, NH-NL), sentiment (VIX, Put/Call) and AROON.

1. 🎯 Confluence Over Single Signals

In the basic course you learned each indicator individually — RSI, MACD, Bollinger Bands, Stochastic, ATR. That is the vocabulary phase. The professional phase is about the grammar: how you combine these tools so that a robust picture emerges from individual, often contradictory signals. That is exactly what confluence is — the convergence of several independent clues at the same spot.

A single signal is always suspect. An RSI below 30 screams "oversold", but in a strong downtrend it stays there for weeks and you catch a falling knife. A MACD cross looks good in a backtest, but in a sideways market it produces an avalanche of false signals. The solution is not the better indicator — there is no such thing. The solution is the layering of independent sources of information.

📖 Builds on: Chart Analysis · Module 6 — Indicators (MACD, RSI, Bollinger, Stochastic, ATR).

The three information layers

A professional setup answers three different questions — and each question needs its own type of indicator:

LayerQuestionTypical tools
TrendIn which direction am I even allowed to trade?Moving averages (SMA 50/200), ADX, Aroon, market structure
MomentumDoes the move have power — or is it running out?RSI, MACD, Stochastic
VolatilityHow large is the fluctuation — where do I set stop and target?ATR, Bollinger Band width, Squeeze

These three layers are orthogonal: they measure different things and do not automatically confirm one another. When the trend filter, the momentum trigger and the volatility timing all point in the same direction, you have real confluence. If one layer does not agree, you stay out.

The anti-stacking rule

This is where most beginners make the decisive mistake: they pile three oscillators on top of each other — RSI, Stochastic and CCI — and rejoice when all three show "overbought". That is not confluence, that is an illusion. RSI, Stochastic and CCI are all calculated from the same price action following the same basic idea (recent prices relative to the range). They are highly correlated — three voices, but only one piece of information.

⚠️ Anti-stacking rule: Never stack several indicators of the same type. Three oscillators on top of each other give you the same information three times and feign a certainty that does not exist. A redundant signal is not a confirmed signal. Take from each of the three layers — trend, momentum, volatility — exactly one representative.

The rule of thumb is: at most one indicator per information layer. One trend filter, one momentum gauge, one volatility measure. More lines on the chart do not mean more clarity, but more noise and more reasons to rationalize a bad trade. Less is measurably more here.

💡 Practice test: switch off one indicator for a moment. If your decision does not change, it was redundant — get rid of it. A setup that only works because five lines agree at the same time is usually just fitted to the past (more on that in the practice section).

2. 🧩 Proven Combos

From the three layers only a few, but very robust combinations can be built. Three of them have proven themselves over decades and across all asset classes. They are no holy grails — but they all follow the same sound principle: a slow indicator filters, a fast one triggers.

1. MA trend filter + RSI pullback

The classic for trend following. The moving average (e.g. SMA 50 or EMA 21) defines in which direction you are even allowed to trade — above the MA only long, below it only short. Within this trend you do not wait for an overbought RSI, but for the pullback: in an uptrend the RSI briefly falls below 40–50, the price holds the MA as support — and you enter when the RSI turns up again. So you buy weakness in an uptrend, not strength.

2. Bollinger Bands + Stochastic (mean reversion)

The combo for sideways markets. Here you deliberately seek no trend following, but the return to the middle. When the price touches the lower Bollinger Band (a statistical extreme, 2 standard deviations) and the Stochastic is simultaneously in the oversold zone (below 20) with a %K/%D cross to the upside, that is a mean-reversion long back to the middle band. Important: this combo belongs in calm, sideways-moving phases — in a strong trend the price "rides" along the band and the setup becomes a trap.

3. MACD + ADX (trend strength)

The MACD signal tells you the direction of the momentum, the ADX tells you whether there is any trend at all worth following. ADX measures trend strength (not direction) on a scale of 0–100: values above 25 indicate a sustainable trend, values below 20 a directionless market. Only when the ADX rises above 25 do you take the MACD cross seriously. This is how you filter out exactly the sideways phases in which the MACD would otherwise deliver false signal after false signal.

ComboSetupSignal (trigger)Filter / confirmation
MA trend + RSI pullbackTrend following, pullback entryRSI turns up out of the 40–50 zone (in an uptrend)Price above SMA 50 / EMA 21; MA held as support
Bollinger + StochasticMean reversion, range%K crosses %D upward below 20 (oversold)Price at the lower band; market sideways (no trend phase)
MACD + ADXTrend following, momentumMACD line crosses signal line upwardADX > 25 (sustainable trend present)

Note the pattern in each row: the filter column always contains a different indicator type than the signal column. Trend filters momentum, trend strength filters momentum, range context filters mean reversion. That is the anti-stacking rule put into practice — no two oscillators parroting each other.

7030 Price + EMA(21) · Trend filter RSI(14) · Momentum Pullback to EMA RSI turns up from 40–50
Combo 1 in the picture: the EMA trend filter only allows longs above the line; the entry (green) only comes when the price pulls back to the EMA and the RSI turns up out of the 40–50 zone. Slow filter, fast trigger — two different indicator types.

💡 Which combo when? In trends, Combos 1 and 3 work (trend following), in ranges Combo 2 works (mean reversion). The ADX from Combo 3 is at the same time your best switch: ADX high → trend-following combos, ADX low → mean-reversion combo.

3. 🌐 Breadth, Sentiment & Market Internals

Up to now you have only looked at one chart — that of the instrument you want to trade. Professionals additionally look at the state of the entire market. Because an index level alone conceals one decisive question: is the index rising because many stocks are rising — or because just a handful of heavyweights are pulling it up while the breadth is already crumbling? This question is answered by market breadth (breadth).

Breadth — how many stocks carry the move?

Breadth indicators count how many individual stocks of an exchange participate in the index move. They are pure internals — they do not come from the price of the index, but from the underlying stocks.

IndicatorWhat it measuresUse / threshold
$TICK (NYSE tick)Number of NYSE stocks rising on the last tick, minus those fallingReal-time intraday sentiment; extremes around +1000 / -1000 show buying or selling panics — often short-term counter points
$TIKIThe same tick logic, but only across the 30 Dow stocksVery short-term program-trading signal; values of +/-20 indicate concerted index arbitrage
Advance-Decline LineCumulative difference of rising minus falling stocks (summed up day by day)Trend confirmation: if the index rises along with the AD line, the uptrend is broadly carried and healthy
McClellan OscillatorDifference of two EMAs (19 and 39 periods) of the daily advance-decline dataMomentum of market breadth; above 0 = breadth improving, below 0 = breadth deteriorating; extremes around +/-100 signal overextended states
McClellan Summation IndexRunning summation (cumulative sum) of the McClellan OscillatorLong-term breadth picture; a crossing of the zero line marks major trend changes in market breadth
TRIN (Arms-Index)(rising/falling stocks) divided by (volume of rising/volume of falling stocks)Above 1 = selling pressure (volume flows into falling stocks), below 1 = buying pressure; extremes above 2 often short-term bottom signals
New Highs / New LowsNumber of stocks at a 52-week high minus those at a 52-week lowA healthy bull market shows many new highs; an expansion of new lows while the index rises is a warning signal

Important on the mechanics: the McClellan Oscillator is nothing other than the difference of two exponential moving averages of the advance-decline figures — structurally related to the MACD, only calculated on market breadth instead of a single price. The Summation Index is its integral (the running sum) and thus smooths the daily ups and downs into a long-term breadth trend. The TRIN in turn couples the numerator (stocks) and volume: only when the volume flows disproportionately into the falling stocks does it rise above 1 — so it measures whether the money follows the price.

Sentiment & volatility — what does the market expect?

While breadth measures the current state, sentiment indicators measure the expectation and the fear of market participants.

IndicatorWhat it measuresUse / threshold
VIX (fear gauge)Expected 30-day volatility of the S&P 500, derived from option pricesHigh = fear, low = complacency; as a risk-on/-off switch: a falling VIX favors long setups, a rising VIX calls for caution
VIX term structureRatio of shorter-term to longer-term volatility futures (Contango vs. Backwardation)A normal upward-sloping structure (Contango) = calm risk-on; an inverted structure (Backwardation) = acute stress, risk-off
Put/Call-RatioTrading volume of puts divided by that of callsContrarian indicator: very high values = extreme fear (often a bottom near), very low = greed (often a top near)

Both the Put/Call-Ratio and the VIX are at their core contrarian extreme indicators: it is not the middling values that trade, but the swings. When everyone flees into puts and the VIX explodes, most of the fear is usually already in the price — statistically that tends to be a bottom rather than a beginning. Conversely, deep complacency (low VIX, many calls) is historically the most dangerous moment.

The key concept: divergence price ↔ breadth

The real added value of the internals lies in the divergence. It is one of the few genuine early-warning systems of technical analysis:

⚠️ Breadth divergence: The index makes a new high — but the Advance-Decline Line, the McClellan Oscillator or the number of new 52-week highs no longer confirms this high. Fewer and fewer stocks carry the rise, a few heavyweights pull the index. This is the classic distribution phase before a top forms: the facade rises, the foundation is already crumbling.
S&P 500 · Index (HTF) McClellan Oscillator · Market breadth Price: higher high ↗ Breadth: lower high ↘
⚠️ Bearish breadth divergence: the index (top) marks a higher high, yet the McClellan Oscillator (bottom) does not confirm it — fewer and fewer stocks carry the rise. Not a timing signal, but a clear warning sign of waning trend quality.

These divergences are not a timing tool — they do not tell you the day of the top, but that the quality of the trend is declining. That is exactly why breadth, VIX and Put/Call belong consistently in the toolbox for index and broad-market trades: you never trade an index blindly on its own chart, but always with the internals filter at your back. If the breadth confirms the trend, you step on the gas. If it diverges, you reduce size and trail stops.

💡 Rule of thumb for broad-market trades: first the green light from the breadth (AD line and McClellan confirm the index), then the risk environment from the VIX (no acute stress), and only then the individual setup. A long idea against a falling AD line is a trade against the market — even if the index is still rising.

4. 📡 The Aroon Indicator

The Aroon indicator (Sanskrit for "dawn", developed by Tushar Chande in 1995) answers a very concrete question: How long ago did the price make a new high or a new low? Unlike most indicators, Aroon does not measure the price itself, but the time since the last extreme. That is precisely what makes it an unusually early trend detector.

Aroon Up and Aroon Down

Aroon consists of two lines that fluctuate between 0 and 100 (the standard period is usually 25):

  • Aroon Up measures how many periods have passed since the highest high within the observation window. If the high is right at the very front of today, Aroon Up stands at 100. The longer ago the last high was, the closer the value falls toward 0.
  • Aroon Down measures the same for the lowest low. A fresh low = near 100, no new low for a long time = near 0.
ConstellationMeaning
Aroon Up above 70, Aroon Down below 30Strong, intact uptrend (continuously new highs)
Aroon Down above 70, Aroon Up below 30Strong, intact downtrend (continuously new lows)
Aroon Up crosses Aroon Down upward📈 Early confirmation of a beginning uptrend
Aroon Down crosses Aroon Up upward📉 Early confirmation of a beginning downtrend
Both lines below 50 and intertwinedConsolidation / sideways market — no trend

The Aroon Oscillator

The Aroon Oscillator combines both lines into a single one: Aroon Up minus Aroon Down. It swings between -100 and +100 around a zero line:

  • Clearly above 0 (toward +100): uptrend dominates.
  • Clearly below 0 (toward -100): downtrend dominates.
  • Near 0: directionless market.

The zero-line crossing of the oscillator is the most compact form of the Aroon signal — a single value that expresses both direction and rough strength.

7030 Price · Uptrend Aroon(25) Aroon Up Aroon Down Up crosses Down → trend start
Aroon measures the time since the last high or low. When Aroon Up crosses above Aroon Down (yellow dot), an uptrend begins early; as long as Up stays above 70, the trend is intact.

Distinction: Aroon vs. ADX

Aroon and ADX are often confused because both revolve around trends — but they measure fundamentally different things, and that is precisely where their complementarity lies:

PropertyAroonADX
MeasuresTime since the last high/low → trend direction + early phaseTrend strength (directionless)
Directional statementYes — Up vs. Down shows the directionNo — ADX alone only says "strong/weak", not where to
Strength of reactionReacts early, already at the first new highs/lowsReacts more sluggishly, only confirms established trends
Typical useDetect the start of a trend, spot range breakouts earlyConfirm whether an already visible trend is sustainable (threshold 25)

Memo: Aroon tells you whether and where a trend is forming; ADX tells you whether it is strong enough to trust it. Whoever wants to be there early looks first at Aroon. Whoever wants to filter out false breakouts puts ADX in front. In the practice section we combine exactly these two perspectives.

5. 🛠️ Practical Setup: The Confluence Long

Now we put it all together. The goal is a confluence long that is only triggered when all three information layers plus the market context agree. No single element is enough — and that is exactly the strength: you trade less often, but every trade has the market at its back.

The setup — step by step

LayerCondition (all must be met)
🌐 Market filterIndex above SMA 200, Advance-Decline Line confirms the index high, VIX not in stress (term structure in Contango)
📈 TrendIndividual stock above EMA 50; Aroon Up above 70, Aroon Down below 30 (intact uptrend)
MomentumRSI turns up out of the pullback (40–50); MACD histogram turns positive again
📏 Volatility / riskStop-loss = entry minus 2× ATR; position size such that this stop costs at most 1 % of the account

If all four rows are met, it is a trade with confluence: the broad market carries it (filter), the instrument is trending (trend), the entry catches a pullback with turning momentum (momentum), and the risk is cleanly sized via the volatility (ATR). If even one row is missing, you skip the trade. There is a new setup every day — but never the same account again if you gamble it away.

The trader who maintained his system to death

How a good setup became a worthless one — through too much optimization.

Jonas started with exactly the confluence long from above — market filter, Aroon, RSI pullback, ATR stop. Over three months it ran decently: solid, not spectacular, but profitable. Then he began to "improve". He noticed that a few losing trades would have been avoidable if he had additionally required the Stochastic below 30. So he added it. Then he spotted a trade that would have gone better with an RSI threshold of 45 instead of 50. So he changed that. Then a CCI filter. Then a second MA. Then a rule that there is no trading on Mondays, because two Mondays brought losses.

After three months his system had eleven conditions. In the backtest the curve looked dreamy — almost no losing trades anymore. In real application the system practically never triggered anymore, and when it did, it lost anyway. Jonas had fitted his system exactly to the past — to coincidences that would never repeat. He had stacked several oscillators (violating the anti-stacking rule) and hung every rule on individual trades.

The lessonEvery additional rule that you fit to individual past trades makes your system better on paper and worse in reality. This is called curve-fitting or over-optimization. A robust system has few, logically founded rules — not many that coincidentally explain the past. If a filter drastically lowers the number of trades without there being a good reason, it is usually over-fitting.
⚠️ Warning about over-optimization: More filters are not more safety. Every rule you add only to optimize away past losing trades is curve-fitting. Stick to the three layers plus market context — and resist the urge to invent a new condition after every loss. A few losses are not a defect of your system, they are the cost of doing business.
🎯 At a glance
  • Confluence instead of a single signal: layer trend + momentum + volatility — exactly one indicator per layer.
  • Anti-stacking rule: never stack several oscillators — they give the same information twice.
  • Three proven combos: MA + RSI pullback (trend following), Bollinger + Stochastic (mean reversion), MACD + ADX (trend strength).
  • Market internals: breadth ($TICK, McClellan + Summation, AD line, TRIN, New Highs/Lows) and sentiment (VIX, Put/Call) as a filter for index trades.
  • Divergence price ↔ breadth is the early-warning system: an index high without breadth confirmation = distribution phase.
  • Aroon detects trends early (direction + time), ADX confirms their strength (threshold 25).
  • Confluence long only when market + trend + momentum + ATR risk all agree.
  • Avoid curve-fitting: few logical rules beat many over-fitted ones — every loss-driven additional rule makes the system worse.