← Back to the Academy
Now that you can read raw price, add the instruments that measure it: moving averages, RSI, MACD, and the chart and candlestick patterns that repeat. Tools don't predict — they describe. Use a few well, and look for the moment they agree.
A moving average (MA) plots the average price over the last N candles, sliding forward each bar. It smooths the jagged price into a single line so the underlying direction is obvious. A simple MA (SMA) weights every candle equally; an exponential MA (EMA) weights recent candles more, so it reacts faster.
Two everyday uses: direction — price above a rising MA is bullish drift, below a falling MA is bearish — and dynamic support/resistance, where price pulls back to a key MA (often the 20 or 50) and bounces. When a faster MA crosses above a slower one it's a golden cross (momentum turning up); the reverse is a death cross. Crosses lag, so treat them as confirmation, not a starting gun.
A moving average is arithmetic, not prophecy: the 20 S M A is literally the average of the last twenty closing prices, recalculated each new bar. Because it's built entirely from the past, it lags — it can only confirm a trend already underway, never call it early. The E M A (exponential moving average) fixes some of that lag by weighting the most recent closes more heavily, so it turns faster; the trade-off is it also whipsaws more in choppy price. The famous 200-day MA matters not because the number is magic but because so many institutions watch it that it becomes a self-fulfilling level — a huge crowd defends it, so it holds.
The RSI measures the speed and size of recent moves on a 0–100 scale. Above 70 is traditionally "overbought" — the move may be stretched; below 30 is "oversold." But the rookie mistake is shorting every 70 and buying every 30: in a strong trend, RSI can stay overbought for a long time. Read it as pressure, not a trigger.
The real gold is divergence. When price makes a higher high but RSI makes a lower high, momentum is fading beneath the surface even though price still rises — a warning the trend is tiring. The reverse (lower price low, higher RSI low) hints a downtrend is losing steam. Divergence doesn't time the turn; it tells you to tighten up.
The R S I measures momentum — how forceful recent moves have been — by comparing the average size of up-closes to down-closes over (usually) 14 bars, then scaling it 0 to 100. It's a speedometer, not a GPS: it tells you how fast you're going, never which turn is coming. That's why "sell at 70" fails so badly in crypto — in a powerful bull run R S I can pin above 70 for weeks while price keeps climbing. Speed is high, but the car isn't crashing. Read overbought/oversold as fuel pressure, and only act when it aligns with structure and levels.
The real edge is divergence: when price grinds to a higher high but R S I prints a lower high, the engine is losing power even as the car inches forward. It's an early warning to tighten stops or take profit — not a precise timing tool, because divergence can persist far longer than feels reasonable before price finally turns.
The MACD takes the difference between a fast and a slow EMA (the MACD line), then plots a slower average of that (the signal line). The gap between them is drawn as a histogram. When the MACD line crosses above the signal, momentum is turning up; crossing below, down. The histogram growing means momentum is accelerating; shrinking means it's fading — often before price reverses.
Like all moving-average tools, MACD lags — it confirms a move that's already underway. It shines as a second opinion: when MACD agrees with what structure and RSI are telling you, your read is on firmer ground. Used alone in a choppy range, it whipsaws.
The M A C D ("mack-dee") is elegantly simple: it's the distance between a fast E M A and a slow one (default 12 and 26). When the fast average pulls away above the slow one, momentum is building up; when it collapses back, momentum is fading. The signal line is just a 9-period average of that distance, and the histogram is the gap between the M A C D line and its signal — so the histogram is really the rate of change of momentum, the acceleration. A shrinking histogram means the move is decelerating before price actually turns, which is why it often gives an earlier heads-up than the crossover itself.
But it's still an average-of-averages, so it lags, and in a sideways chop it whipsaws — flashing cross after cross that lead nowhere. M A C D earns its keep as a second witness: when it agrees with structure and R S I, your read stands on firmer ground; alone in a range, it's a trap.
Patterns are just structure with a name — recurring shapes that often resolve a certain way because crowds behave similarly. The reliable handful:
The trap is seeing patterns everywhere. A pattern only counts when it forms at a level that already matters and resolves with volume. The break — and the retest of the broken line — is where the trade lives, not the drawing itself.
Every classic pattern is a snapshot of a crowd making the same decision. A triangle forms as buyers and sellers squeeze into agreement — volatility contracts, orders coil, and the eventual break releases the pent-up energy in one direction. A head-and-shoulders is a trend running out of buyers: each push makes a weaker high until the "neckline" of support finally gives way. The crucial, non-obvious truth is that the pattern itself isn't the trade — the break and the retest are. Amateurs buy the pretty shape; professionals wait for price to break the pattern's edge on volume, then retest that edge and hold, entering with a tight, obvious invalidation just on the wrong side of the line.
Where chart patterns play out over many bars, candlestick patterns show a shift in one or two. The essentials:
Candle signals are location-dependent. A hammer in the middle of nowhere is noise; the same hammer rejecting a major support, on rising volume, is a real tell. Always read the candle and where it printed.
Candlestick patterns compress a change of control into one or two bars. A bullish engulfing is powerful because a big green body swallowing the prior red one means buyers didn't just show up — they erased a whole session of selling in a single candle. A hammer (long lower wick, small body) is the fingerprint of a failed breakdown: sellers drove price way down, then buyers slammed it back up before the close, trapping every late short. The doji is pure indecision. But the decisive lesson is that these signals are location-dependent — the exact same hammer is meaningless mid-range and hugely significant when it rejects a major support on rising volume. Read the candle and the ground it stands on.
No single tool is reliable on its own — each is just one angle on the same price. The edge comes from confluence: when several independent reads point the same way at the same place. For example: price pulls back to the 50 EMA, which sits on a prior support, where a bullish engulfing prints while RSI turns up from oversold. That's four witnesses telling the same story.
And every setup needs its opposite: invalidation. Decide in advance the price that proves the idea wrong — usually just beyond the level or below the signal candle's low — and that becomes your stop. Confluence tells you where the odds are good; invalidation tells you what it costs to be wrong. You need both before you click.
Confluence and invalidation only pay off when they feed the one number most beginners ignore: position size. The professional rule is fixed fractional risk — never risk more than a small, constant slice (say 1%) of your account on any trade. The math is exact: if your account is $10,000 and your invalidation is 5% below entry, then to risk only 1% ($100) you buy a $2,000 position ($100 ÷ 0.05). Move your stop closer and you can size bigger for the same risk; move it wider and you must size smaller. Notice what this kills — you size off the distance to your stop, never off how confident or excited you feel.
This is exactly how our agents are wired: Petro-Lwa strategies define the setup and stop, Lasirèn sizes and executes off that stop distance, Brigitte holds the thesis and its invalidation, and Gede enforces the discipline of actually cutting the loss. Confluence finds the lopsided bet; fixed-fractional sizing guarantees no single bad flip can break you.
1. Price makes a higher high but RSI makes a lower high. This is:
2. An EMA differs from an SMA because it:
3. A hammer candle is most meaningful when it prints: