ADX, moving-average slope and multi-timeframe alignment
Lesson 16 · about 10 min
Swing structure is the definition of trend. This lesson adds two indicators that confirm what the structure says, and then the most useful idea in the module: aligning the structure across timeframes so that you trade in the direction the bigger picture agrees with.
Moving-average slope as confirmation
A simple moving average (SMA) of the close over N candles is the average of the last N closes. A 20-period SMA on a daily chart is the mean close of the last 20 days.
| Day | Close | 5-day SMA |
|---|---|---|
| 1 | 40.0 | |
| 2 | 40.6 | |
| 3 | 41.1 | |
| 4 | 41.5 | |
| 5 | 42.0 | 41.04 |
| 6 | 42.4 | 41.52 |
| 7 | 42.1 | 41.82 |
Day 5 SMA = (40.0 + 40.6 + 41.1 + 41.5 + 42.0) ÷ 5 = 41.04. Day 6 drops day 1 and adds day 6: (40.6 + 41.1 + 41.5 + 42.0 + 42.4) ÷ 5 = 41.52.
The slope is simply whether today's SMA is above yesterday's. Rising slope: the average close is rising. The reason this confirms trend is that a sequence of higher highs and higher lows must, over enough candles, pull the average of closes up. The moving average is a smoothed, lagged summary of the same structure.
The lag is the cost. In the table above, if day 8 crashed to 39.0, the swing structure would show a CHoCH immediately while the 5-day SMA would still read 41.4, barely below day 7. The moving average is late to every turn; that is what smoothing means.
Use it this way:
- Structure says uptrend and the 20-SMA slopes up: agreement; full confidence in the label.
- Structure says uptrend but the 20-SMA slopes down: the trend is young (the average has not caught up) or the structure reading is on too small an N. Reduce confidence.
- Structure says CHoCH and the SMA still slopes up: the SMA is lagging. Trust the structure; the SMA will follow if the CHoCH was real.
Never let the moving average override the structure. It is a second opinion from a witness who is always a few days behind.
ADX as a trend-strength gauge
ADX (Average Directional Index) is a single line from 0 to 100 that measures how strongly price is moving in one direction, without saying which direction. It is built from the differences between successive highs and successive lows, smoothed over 14 periods by default. The formula is tedious to do by hand and every platform computes it; what matters is the reading.
| ADX reading | Conventional read |
|---|---|
| Below 20 | No trend; range or chop |
| 20-25 | Trend may be starting |
| 25-40 | Trending |
| Above 40 | Strong trend; often late in the move |
ADX rises when price moves consistently in one direction and falls when it stalls, whichever the direction. That makes it a regime filter, not a signal:
- ADX below 20 and structure says range: agreement; trade the edges, avoid breakouts.
- ADX above 25 and structure says uptrend: agreement; trade pullbacks with the trend.
- ADX below 20 and structure says uptrend: the trend is weak or very young; expect deep pullbacks.
- ADX falling from above 40: the trend is losing momentum; tighten stops, expect a range or CHoCH.
ADX lags like every smoothed indicator. Its real use is answering "is this a market where breakouts have been working?"
Key idea: Moving-average slope and ADX are lagging confirmations of the swing structure. When they agree with the structure, act with confidence; when they disagree, the structure is right about the present and the indicators are describing the recent past.
Multi-timeframe alignment
Because swing points depend on timeframe, every chart has several trends at once: weekly, daily, 1-hour. They frequently disagree, and the disagreement is information to use, not a problem to solve.
The framework:
- Pick the timeframe you trade on. Call it the trading timeframe.
- Look one step up (roughly 4-6× larger) for the context timeframe: it defines the direction you prefer to trade in.
- Look one step down for the entry timeframe: it gives you the precise candle and stop.
| Trading timeframe | Context timeframe | Entry timeframe |
|---|---|---|
| 5-minute | 30-minute | 1-minute |
| 15-minute | 1-hour | 5-minute |
| 1-hour | 4-hour or daily | 15-minute |
| Daily | Weekly | 4-hour |
The rule: trade in the direction of the context timeframe, at the levels of the trading timeframe, using the candles of the entry timeframe.
What alignment looks like
Best case: all three agree. Weekly uptrend, daily pulling back to a higher low, 4-hour printing a bullish engulfing at that low. Every timeframe's trapped group is on the same side; the pullback sellers on the daily are about to be squeezed by the weekly trend.
Common case: context up, trading timeframe in a CHoCH. The daily has closed below its last higher low but the weekly is still HH/HL. This is a deep pullback in the weekly trend. It is not the time to short (against the context) and not yet the time to buy (the daily has no structure). Wait for the daily to print a new higher low and BOS, which is where the two agree again.
Worst case: context down, trading timeframe up. Every rally on the daily is a pullback on the weekly, and the weekly's trapped group (longs from higher prices) is selling into it. Long trades here are fighting the larger structure and should have smaller targets and tighter expectations, or be skipped.
Weekly: HH HL HH HL -> context: UP
\
Daily: LH LL HL HH -> pullback, then re-aligned
\
4-hour: hammer at the daily HL -> entry
Putting the module together
Before any trade, the structure check is:
- Context timeframe: up, down, range or undefined?
- Trading timeframe: same question. Do they agree?
- Where is the most recent swing that would break the trading-timeframe structure? That is your stop reference.
- Do the 20-SMA slope and ADX agree, or is one of them lagging a turn you can already see?
If the answers to 1 and 2 disagree, the default is to wait. Module 5 shows what to do when they agree.
Try it: Pick one instrument. On the weekly, daily and 4-hour charts, label the trend using the swing definition (N = 5 on each). Write the three labels side by side. Then check the 20-SMA slope and ADX on the daily. Record whether they agree with the daily structure. Repeat weekly for a month; you will see the timeframes drift into and out of alignment, which is precisely when trades appear and disappear.
Recap
- A moving average's slope is a smoothed, lagged summary of the swing structure; rising slope confirms an uptrend but is always late to a turn.
- ADX measures trend strength without direction: below 20 is range, above 25 is trending, falling from above 40 is a tiring trend.
- Indicators confirm structure; they never override it.
- Trade in the direction of the context timeframe, at the levels of the trading timeframe, with the candles of the entry timeframe.
- When the context and trading timeframes disagree, the default is to wait for them to re-align.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.