🎯 In 30 seconds: The same stock can be trending up on a weekly chart and down on a daily chart. Top-down analysis deals with this by giving each time frame one job: the longer chart (weekly) sets the direction, the middle chart (daily) finds a setup like a pullback to support, and a shorter chart (daily or hourly) times the entry. Trading in the direction of the longer trend tends to put the odds on your side. When the time frames disagree, trade smaller or skip. Combining signals helps only if they measure different things.
Why one chart is not enough
A time frame is how much time each bar or candle on a chart represents: one week, one day, one hour, five minutes. See Chart basics.
Each time frame tells a different story. A daily chart might show a sharp three-week drop that looks like a downtrend. Zoom out to a weekly chart and the same drop might be a normal pullback inside a year-long uptrend. Zoom in to an hourly chart and the last two days might already be rising.
None of these views is wrong. The problem comes when you pick the one that agrees with what you want to do. Top-down analysis is a fixed routine that stops you from doing that.
The three-chart routine
Pick three time frames, each roughly four to six times longer than the next. A common set for swing traders, who hold for days to weeks:
| Chart | Job | Questions to answer |
|---|---|---|
| Weekly (long) | Direction | Higher highs and higher lows, or lower? Price above or below a rising 40-week (or 10-week) average? Any major level close by? |
| Daily (middle) | Setup | Is price pulling back toward support, a moving average or a broken level, in the direction of the weekly trend? |
| Hourly or daily (short) | Trigger | Has price stopped falling and started to turn? A break of a short downtrend line, a higher low, a close above the prior bar's high. |
Day traders shift everything down: daily for direction, 15-minute or hourly for setup, 1 to 5-minute for trigger. Longer-term traders shift up: monthly, weekly, daily.
The rule that makes it work: the longer time frame wins. You only take setups on the middle chart that agree with the longer chart's direction.
Worked example: one trade, three charts
Hypothetical numbers following the figure. Account $15,000, risking 1% ($150) per trade.
- Weekly: higher highs and higher lows for a year, price above a rising 10-week average. Direction: long only.
- Daily: price has pulled back from about $90.50 to a support zone of $81.00 to $82.60, made of earlier swing lows. That is a pullback in an uptrend, the setup we want.
- Trigger: after a low of $81.34, the next day closes at $82.55, above the prior day's high. That is the first sign buyers are back.
| Entry | $82.55 |
| Stop: below the pullback low ($81.34) with a buffer | $80.90 |
| Risk per share: $82.55 minus $80.90 | $1.65 |
| Shares: $150 / $1.65, rounded down | 90 shares ($7,430) |
| First target: the prior high | $90.50 |
| Reward per share: $90.50 minus $82.55 | $7.95 |
| Reward to risk: $7.95 / $1.65 | about 4.8 to 1 |
Notice how each chart contributed something different. The weekly chart said which side to be on. The daily chart said where to look. The trigger gave a precise entry with a nearby stop, which is what made the reward-to-risk ratio good.
Also notice the position is nearly half the account. That is fine for risk on normal days (the stop limits the loss to about $150), but a large overnight gap could cost much more. See Breakouts and gaps and check the earnings calendar before entering.
When the time frames disagree
They often will. A few ways to handle it:
- Weekly up, daily down: this is a pullback. It is the setup you want, but wait for the trigger. Buying while the daily is still falling is catching a falling knife.
- Weekly down, daily up: this is a bounce in a downtrend. A long trade here fights the bigger trend. Many traders skip it, or treat it as a short setup once the bounce stalls.
- Weekly sideways: there is no direction to follow. Trade the edges of the range, or wait for a breakout. See Support and resistance.
- A major weekly level right above your entry: your target is capped. The reward-to-risk may not be worth it.
Putting several signals together
When several independent pieces of evidence point the same way at the same price, traders call it confluence. For example: an uptrend on the weekly chart, a pullback to a prior breakout level on the daily, the 50-day average at the same price, and a bullish candle on above-average volume.
Confluence helps only if the signals measure different things. Good combinations mix categories:
- Trend: swing structure or a moving average (Trends, Moving averages).
- Location: a support or resistance level.
- Momentum: one oscillator, such as RSI (Momentum indicators).
- Participation: volume.
- Risk: ATR for the stop distance (Volatility tools).
Bad combinations stack similar tools. RSI, stochastic and MACD all oversold is one fact, counted three times.
A second trap is adding conditions until the past looks perfect. Every extra rule removes some losing trades from history, but also reduces the number of trades, until the "system" is describing a handful of lucky examples. Keep the checklist to four or five items you can explain. Does technical analysis work? explains why.
What the evidence says
Two research findings support the logic of "longer trend first, shorter time frame for timing":
- Medium-term trends tend to persist. A study of 58 futures markets found that returns over the past 1 to 12 months tended to continue, then partly reversed over longer periods. That supports favoring the direction of a multi-month trend.
- Very short-term moves tend to reverse. An older study of individual U.S. stocks found that one month's return tended to be followed by a move in the opposite direction the next month. That fits the idea that short pullbacks inside a longer trend are often temporary.
Neither study tested the three-chart routine itself, and both describe averages across many assets and many years, not what will happen to one stock. Also be careful with the shortest time frames: a large study of day traders in Taiwan found that the vast majority lost money, and only about 15% earned abnormal returns (returns beyond what market moves would explain) after fees in an average year. Shorter charts mean more trades, more costs and more noise. See Day trading.
Try it
- Pick five stocks from your watchlist. For each, write one line per time frame: weekly direction, daily setup, trigger present or not.
- Only one or two of the five should qualify on a typical day. If all five look like buys, your rules are too loose.
- For a qualifying stock, write the full plan like the worked example: entry, stop, shares, target, reward to risk.
- Place it in the paper portfolio. Repeat for 20 trades before using real money, and compare the trades where all three charts agreed with any you took when they did not.
- In the three-chart routine for a swing trader, what is the job of each chart? Answer: Weekly sets the direction, daily finds the setup, and a shorter chart (hourly or daily) times the trigger.
- When the time frames disagree, which one wins? Answer: The longer time frame. Only take setups on the middle chart that agree with the longer chart’s direction.
- The weekly chart is up but the daily chart is falling. What is this, and what should you do? Answer: It is a pullback, the setup you want. Wait for the trigger. Buying while the daily is still falling is catching a falling knife.
- Your account is $15,000 and you risk 1%. Entry is $82.55 and the stop is $80.90. How many shares? Answer: 90 shares ($150 / $1.65, rounded down).
Sources
- Moskowitz, Ooi and Pedersen: Time series momentum (Journal of Financial Economics, 2012), author PDF
- RePEc: Jegadeesh, Evidence of Predictable Behavior of Security Returns (Journal of Finance, 1990)
- Barber, Lee, Liu and Odean: The Cross-Section of Speculator Skill: Evidence from Day Trading (author PDF, 2011)