How to build the PRZ from confluence

A PRZ is not one ratio. It is the band where independent measurements for D agree: the XA retracement or extension that defines the pattern, a projection of the BC leg, and an AB=CD completion. Each one alone is a guess; when three of them land close together, you have a zone worth watching.

Take a hypothetical bullish Bat: X at $40.00, A at $50.00, B at $45.00 (a 0.50 retracement of XA) and C at $48.10 (a 0.62 retracement of AB). The XA leg is $10.00, AB is $5.00 and BC is $3.10. The measurements for D are listed below and drawn as a ladder in the figure. X at $40.00 is the line a Bat must not break.

  • The defining ratio: 0.886 of XA = $50.00 − $8.86 = $41.14. For a Bat this is the anchor; without it, the pattern is not a Bat.
  • The BC projection: 2.24 × $3.10 = $6.94, so $48.10 − $6.94 = $41.16. The 2.0 projection gives $41.90 and the 1.618 projection $43.08.
  • The AB=CD completion: the alternate 1.27 AB=CD gives $48.10 − $6.35 = $41.75; the plain AB=CD gives $43.10.
  • The PRZ: $41.14 to $41.75, three measurements within $0.61. The $43.08–$43.10 pair is a decoy: it completes a simple AB=CD, but price has not yet reached the 0.886 retracement that defines a Bat.

Turn your phone sideways to read the image. Tap to close.

A small Bat zigzag on the left with a box around D, zoomed into a price ladder on the right: lines at 43.08–43.10 (AB=CD and 1.618 BC, grey, labeled too early for a Bat), 41.90 (2.0 BC), 41.75 (1.27 AB=CD), 41.16 (2.24 BC), 41.14 (0.886 XA, in bold) and 40.00 (X). The band from 41.14 to 41.75 is shaded as the PRZ and bracketed $0.61, with the notes 3 measurements within $0.61 and a tight cluster = a usable PRZ.

Why the PRZ is a zone, and when it is too wide

The measurements never agree to the cent, and price rarely turns at an exact number: it overshoots, stops short or wicks through. Treating the PRZ as a band lets you plan for that instead of being surprised by it. The zone runs from the measurement closest to the current price to the farthest one in the cluster, and the pattern’s defining XA ratio usually sits at or near one edge.

Width matters. A simple test is to compare the PRZ width with the distance from the zone to the first target. In the example the zone is $0.61 wide and the first target sits about $3.30 above the zone’s low, so the zone is small next to the reward. If the measurements spread over a range close to the distance to the first target, the pattern is too loose to trade: the entry, the stop and the targets all become arbitrary.

Then check what else sits in the zone. A PRZ that overlaps an earlier support or resistance level, a supply and demand zone or a round number has a reason to hold beyond the ratios. A PRZ in empty space, against a strong trend on the higher timeframe, has fewer reasons. The Fibonacci retracement tool helps you check each ratio by hand.

Confirmation or an aggressive limit order?

Reaching the PRZ is the start of the setup, not the entry. There are two ways in, and each has a cost.

Confirmation means waiting for evidence that buyers (in a bullish pattern) are actually showing up. The three common forms are drawn below and listed after this paragraph. You pay for confirmation with a worse price: by the time a candle closes back above the zone, price has already moved away from D, so the stop is farther away and the reward smaller.

An aggressive limit order rests inside the zone before price arrives, for example a buy limit at the PRZ top, $41.75, with the stop beyond X. You get a better price, and you catch reversals that turn sharply without a second chance. You also get filled on every pattern that fails, because a limit order cannot tell a reversal from a collapse. Traders who use limits usually accept fewer winning trades in exchange for larger winners; whether that trade-off suits you is something to test, not assume.

  • A reversal candle that closes back out of the zone: in a bullish PRZ, a candle that closes above the zone top after trading into it, such as a hammer or a bullish engulfing candle from candlestick patterns. A wick alone proves nothing until the candle closes.
  • RSI divergence: price makes a lower low inside the PRZ while the RSI makes a higher low, a sign that selling momentum is fading. See RSI divergence; divergence can last for several swings, so it supports a reversal rather than triggering one.
  • A lower-timeframe structure break: on a chart three to five times faster, price breaks above the last lower high formed inside the zone, the change of character described in smart money concepts.

Turn your phone sideways to read the image. Tap to close.

Three panels with the same shaded PRZ band. Left: falling candles enter the band and a green candle closes back above it, labeled a candle closes back above the zone. Middle: price makes a lower low inside the band while the RSI line underneath makes a higher low, labeled lower low in price, higher low in RSI. Right: a 5-minute zigzag inside the band breaks above a dashed line labeled last lower high, captioned lower timeframe: break of the last lower high.

Where to put the stop: beyond X or beyond the PRZ, plus an ATR buffer

The stop goes where the pattern is proven wrong, then a little farther, because price often probes just past obvious levels. The buffer is easiest to size with the average true range (ATR) of the timeframe you trade: a fraction of one ATR, often a quarter to a half, keeps ordinary noise from stopping you out. A buffer that is too small leaves the stop exposed to that noise; one that is too large wastes risk.

Choose the stop first, then check the targets. If the first target does not pay enough for the stop you need, skip the trade; moving the stop closer to make the numbers work only makes it easier to hit.

  • Gartley and Bat: D completes inside XA, so a move beyond X breaks the structure. The classic stop is beyond X plus the buffer. In the Bat example, X is $40.00 and the 1-hour ATR is $0.60, so half an ATR puts the stop at $39.70.
  • Tighter alternative: beyond the actual extreme of the reaction in the PRZ, plus the buffer. The low at D is $41.05, so the stop goes at $40.75. It risks less per share but gets hit more often, because it sits inside space the pattern is still allowed to use.
  • Butterfly and Crab: D completes beyond X, so X is broken by design. The stop goes beyond the PRZ extreme or beyond the next ratio out, such as the 1.618 extension for a Butterfly, plus the buffer.
  • Cypher: D sits at 0.786 of the XC leg and stays inside X, so the stop usually goes beyond X, as in the Gartley.
  • AB=CD: there is no X, so the stop goes beyond D’s extreme or beyond the next projection, plus the buffer.

Targets, scaling out and moving to breakeven

The usual harmonic targets are measured on the AD leg, from A to the actual extreme at D, not to the theoretical PRZ: the 0.382 retracement of AD as the first target, the 0.618 retracement as the second, and A itself as a final, less likely target. In the example AD is $50.00 − $41.05 = $8.95, so T1 = $41.05 + $3.42 = $44.47, T2 = $41.05 + $5.53 = $46.58, and T3 is A at $50.00.

Scaling out turns those targets into a plan: sell part of the position at each one, as described in scaling in and out. A common version sells half at T1 and moves the stop on the rest to the entry price (breakeven), sells another quarter at T2 and moves the stop to T1, and leaves the last quarter for A or a trailing stop. The figure shows one possible path; price can just as easily reverse before T1, and a breakeven exit still costs commissions and can slip.

Moving to breakeven is a trade-off. It removes the risk of a winning trade turning into a full loss, but a normal pullback to the entry will take you out of trades that later work. Some traders wait for T1 to fill, others for a structure point such as the first higher low. Choose one rule and keep it.

Turn your phone sideways to read the image. Tap to close.

One possible path after a long entry at 42.00: price climbs to T1 at 44.47 (step 1), pulls back, reaches T2 at 46.58 (step 2) and keeps rising toward A at 50.00. A red dashed stop starts at 40.75, moves up to 42.00 after T1 and to 44.47 after T2. Labels on the right give R multiples with 1R = 1.25: T1 about 2.0R, T2 about 3.7R, A 6.4R. Bars along the bottom show shares held: 160, then 80 after selling 80 at T1, then 40 after selling 40 at T2.

Position size from the stop, in R

R is the amount you lose if the stop is hit. Every outcome then becomes a multiple of R, which lets you compare trades with different stops and prices. Decide R before the trade, usually a small fixed percentage of the account, then let the stop distance set the share count, as in the position size calculator: shares = R ÷ (entry − stop).

Hypothetical: a $20,000 account risking 1% per trade has R = $200. With the entry at $42.00 and the tight stop at $40.75, the risk per share is $1.25, so the position is $200 ÷ $1.25 = 160 shares. With the stop beyond X at $39.70, the risk per share is $2.30 and the position is 86 shares ($200 ÷ $2.30 = 86.9, rounded down). Same dollar risk, different size.

The targets then read in R. With the tight stop, T1 is $2.47 away (about 2.0R), T2 $4.58 (about 3.7R) and A $8.00 (6.4R). With the wide stop, T1 is only about 1.1R, a weak risk/reward ratio. Which stop is better is not obvious: the tight one pays more per win and loses more often, and only your own records, read through trading expectancy, can tell. Gaps and fast markets can fill a stop worse than planned, so a real loss can exceed 1R, and fees reduce every result.

A full worked example: the hypothetical Bat, step by step

Here is the whole trade in order, with the same hypothetical numbers. Read it as a template for the decisions, not as a typical result.

  1. Before price arrives: the 1-hour chart shows X $40.00, A $50.00, B $45.00 and C $48.10. You compute the PRZ, $41.14 to $41.75, note that it overlaps a prior support area on the daily chart, and write the plan: confirmation entry, stop at the reaction low minus half an ATR, targets T1, T2 and A.
  2. Price trades into the zone and prints a low at $41.05, $0.09 below the PRZ. Nothing happens yet: the PRZ is a place to watch, and a small overshoot is normal.
  3. Confirmation: the next 1-hour candle closes at $42.00, back above the zone top at $41.75, while the RSI makes a higher low than at the previous swing low. You buy 160 shares at $42.00.
  4. Stop: $41.05 − $0.30 (half of the $0.60 ATR) = $40.75. Risk: 160 × $1.25 = $200, which is 1R.
  5. Targets from AD = $8.95: T1 $44.47, T2 $46.58 and A $50.00.
  6. At T1, sell 80 shares (+$197.60, about 1R on the whole position) and move the stop on the other 80 to $42.00. If price then falls back to $42.00, the trade ends at about +1R before fees.
  7. At T2, sell 40 more (+$183.20) and move the stop on the last 40 to $44.47. Exit those at A or at the trailing stop, whichever comes first.
  8. The failure branch: had price closed back below $41.05 and hit $40.75 before T1, the loss would have been about $200 plus fees, more on a gap. That outcome is part of the plan, not an exception to it.

Invalidation and choosing the timeframe

Write down what cancels the setup before price reaches the zone, so you are not deciding in the moment. The common invalidations are listed below.

Timeframe changes everything. On a 5-minute chart small zigzags fit some harmonic pattern all day, and the ATR buffer is so small that noise decides most trades. Daily and 4-hour patterns are rarer, their PRZs line up more often with levels other traders watch, and their stops are wider, which means smaller positions. Many traders find patterns on the 1-hour to daily charts and use a faster chart only for confirmation.

No timeframe is right for every market. Results differ between stocks, futures, forex and crypto, and between timeframes on the same symbol, so test your exact rules before trusting them: replay past patterns with TradingView bar replay, log every one, and follow how to backtest a trading strategy. A pattern measured after the fact always looks good, which is why backtest overfitting is a real danger with discretionary patterns.

  • Price closes beyond X in a Gartley, Bat or Cypher, or beyond the outer ratio in a Butterfly or Crab: the pattern is gone.
  • Price trades into the PRZ and keeps going without any confirmation within a set number of bars: cancel the setup.
  • The first target pays less than your minimum for the stop you need, which for many traders is about 1.5R to 2R.
  • Price reaches the first target before it reaches the PRZ, or C runs past A: the measurements no longer describe the move.
  • A major event, such as earnings or a central bank decision, falls inside the expected holding period and you have no plan for a gap.

Pre-trade checklist and journal

Run the checklist below before every harmonic trade; if any answer is no, there is no trade. After the trade, record in your trading journal the pattern and timeframe, a screenshot with X, A, B, C, D and the PRZ marked before the entry, the entry type (confirmation or limit), the stop and the ATR you used, the result in R, and whether the PRZ held, overshot or failed. Over dozens of trades, those notes show which patterns, timeframes and entry types work for you.

UpTickFlow, our invite-only TradingView indicator, shows rare A+ setups at key levels, with an entry, a stop and a target drawn before the trade, on 5-minute to 1-hour charts. If a PRZ lines up with one of its levels, treat that as one more reason to pay attention, not as a reason to skip your checklist.

  1. Pattern: the defining ratio is met (0.886 XA for a Bat, 0.786 for a Gartley, and so on), measured on swings you could point to without zooming in.
  2. PRZ: at least three measurements cluster, and the zone is narrow compared with the distance to T1.
  3. Context: the zone sits on a level from a higher timeframe, and the higher-timeframe trend is not strongly against the trade. See multi-timeframe analysis.
  4. Entry rule: confirmation or a limit order, chosen before price arrives.
  5. Stop: beyond X or beyond the PRZ extreme, plus an ATR buffer, written down.
  6. Targets: T1 and T2 at 0.382 and 0.618 of AD, and T1 pays at least your minimum R.
  7. Size: shares = R ÷ (entry − stop), with R a small fixed share of the account.
  8. Invalidation and events: you know what cancels the trade, and nothing scheduled can gap through the stop without a plan.