Early Signs a Trend Is About to Shift: What Institutions Watch Before the Chart Turns

The early signs a trend is about to reverse show up days before the reversal candle. A trend does not end when the moving averages cross. By then the move is half over and the people who caused it are already on the other side. Trends end the way large positions end: someone with size stops adding, starts taking the other side of the flow, and price stops doing what it did for weeks. That leaves fingerprints on the chart days before the obvious reversal candle, and every one of them is visible with nothing more than a clean chart, a Fibonacci tool and a clock.

This is the institutional checklist — what changes first, how far a trend can stretch before it snaps back, which levels and times matter, and how to tell a pause from a turn. Nothing here predicts. It measures.

1. How far the move has travelled (displacement from the mean)

The first thing a desk looks at is not a pattern. It is distance. Every trend has an anchor — the level where most of the recent business was done — and price rubber-bands back to it. Two measurements:

Distance from the 20-period mean on the daily and weekly. Use the 20 EMA (or a 20-period VWAP if you have it). A healthy daily trend spends most of its time within about one to two Average True Ranges (ATR) of the 20 EMA. When NQ closes three or more daily ATRs from the 20 EMA, or the weekly candle sits far outside the 20-week range, the trend is extended: the move has outrun the participation behind it. Extended is not reversed — strong trends stay extended for a while — but it means the next pullback will be deeper and faster than the last ones.

Consecutive closes in one direction. Six or more daily closes in a row, or three weekly closes that each open at the prior close and run, is the kind of streak that ends in a sharp mean-reversion day. Institutions track this because their own risk models force them to trim after runs like that.

The practical read: the further and faster price is from the daily/weekly mean, the smaller the position and the sooner the exit on the trend side.

2. The Fibonacci levels that actually matter

Fibonacci is not magic; it is a shared reference point. (See Investopedia on Fibonacci retracement levels for the standard ratios.) Enough large participants use the same ratios that the levels become self-fulfilling reaction zones. Draw the tool on the last impulsive leg of the trend — swing low to swing high in an uptrend — and watch three things:

• The 38.2% – 50% zone. Shallow pullbacks that hold here say the trend is intact; institutions are still adding. If price cannot get back to these levels between pushes, the trend is accelerating (and getting more extended).
• The 61.8% – 78.6% zone. A pullback that reaches deep into this zone and stalls is the first structural warning. The trend can still resume, but the buyers needed a much bigger discount than before.
• Extension targets: 1.272 and 1.618 of the prior leg. Trends frequently terminate near an extension of the previous swing. Price pushing into 1.618 with a shrinking body and a long wick is where many institutional trend-following programmes are scaling out.

The signal is not the level; it is the change in how price treats the level. The trend that used to bounce at 38.2% and now needs 78.6% has told you something.

3. Structure: the first lower high (or higher low)

This is the cleanest early sign and the most under-rated. In an uptrend, price makes higher highs and higher lows. The first sign the trend is tiring is not a lower low — it is a failure to make a new high after a pullback (the push stalls under the prior high, often at an extension level), followed by a pullback that goes deeper than the last one. The order matters:

1. Higher high, higher low, higher high — trend.
2. Push fails under the last high — first warning.
3. Pullback breaks the last higher low — the "change of character" (CHoCH) many traders talk about.
4. A lower high after that — trend has shifted on that timeframe.

On the daily and 4-hour charts, step 2 is the early sign; step 3 is confirmation; step 4 is what everyone sees. Trade planning happens between 2 and 3.

4. Displacement and the candles that made the trend

Institutions leave a specific mark when they move a market: a large-bodied candle (or several) that breaks structure with little overlap — displacement. The last opposing candle before that displacement is the order block: where the position was built. Two things to watch:

• Displacement is fading. Impulse candles getting smaller, with more overlap and longer wicks, means each push is finding sellers sooner. Compare the body size of the last three daily impulse candles to the first three of the trend.
• Price returns to an order block and does not bounce. Order blocks from the trend's early legs should act as support in an uptrend. When price comes back to one and closes through it on the daily, the participant who built the position is no longer defending it.

5. Momentum divergence and volume

Indicators lag price, but two of them lead the visible reversal because they measure what is happening inside the candles:

• RSI (14) divergence on the daily/4H. Price makes a higher high; RSI makes a lower high. On its own it is weak — trends produce divergence for weeks. Combined with an extension level (section 2) and a failed push (section 3), it is strong.
• Volume (or delta, if your feed shows it). Rising prices on falling volume is distribution: the trend is running on fewer participants. A climax day — the biggest volume in weeks on a candle that closes off its highs — is often the last push before a shift.
• The 20 and 50 EMA relationship. Not the cross (too late) but the slope: when the 20 EMA flattens while price is still making highs, momentum has already left.

6. Time: when shifts actually happen

Trend shifts are not evenly distributed through the day or the week. On NQ:

• Daily and weekly candle closes. The most reliable shift signals print on daily closes (6:00 pm ET) and especially on the weekly close (Friday). A weekly candle that closes back inside the prior week's range after pushing out of it is a classic shift signal.
• The 10:00 am ET 4-hour close. Intraday, the end of the opening drive is where a lot of daily reversals begin, and where a failed push under the prior high first becomes visible.
• Session opens. London (2–3 am ET) and New York (9:30 am ET) are where liquidity gets taken — the sweep of the prior high or low that fails and reverses is the intraday version of section 3.
• Data and Fed days. CPI, payrolls, FOMC at 2:00 pm ET. The first move is frequently the trap; the shift is the second move, after the 4-hour close.
• Monthly and quarterly ends. Rebalancing flows can push an extended trend one last time before it turns. Be more sceptical of a breakout in the last two sessions of a quarter.

The checklist (score it, don't feel it)

Give each one a point on the daily chart. We score it this way ourselves because a number is harder to argue with than a feeling. Three or more, and you plan for a shift; five or more, and you stop adding to the trend.

1. Price is 3+ daily ATRs from the 20 EMA, or 6+ consecutive closes in one direction.
2. The last pullback reached the 61.8–78.6% zone (previous ones stopped at 38.2–50%).
3. The latest push stalled at a 1.272 or 1.618 extension.
4. A failed push under the prior high (or above the prior low).
5. Impulse candles shrinking; wicks growing; a daily close through a prior order block.
6. RSI divergence on the daily or 4H.
7. Volume climax or rising price on falling volume.
8. The signal printed on a daily/weekly close or at the 10 am ET 4H close, not mid-candle.

What you do with a high score is not "short the top." It is: stop adding, tighten the trail, take partials at extensions, and wait for step 3 (the break of the last higher low) before positioning the other way. Institutions do not pick tops; they stop feeding them.

Where MyTradingBuddy fits

This checklist is mechanical, and mechanical is what an assistant is for. MyTradingBuddy reads the chart you have open on TradingView, maps the structure — the higher highs and lows, the order blocks, the levels — and explains where the current candle sits relative to them and what it would wait for. It will not call the top. It will tell you when the trend has stopped making higher highs, which is the only part that matters.

Start the 3-day trial for $14.07 and run the checklist on the chart you already have open: https://mytradingbuddy.ai/pricing

Related reading

• Nasdaq futures weekly patterns: last week to this week
• When to use MyTradingBuddy: best times, timeframes, trend flips
• Order block filled or just tapped?
• Reading NQ by the clock: when the big candles close and why it matters
• The 80/20 of reading a chart
• Why traders fail evaluations on execution, not analysis

Nothing here is financial advice. A high checklist score is a reason to stop adding, not a prediction. Trading involves risk.