Prior-Day High/Low Breakout: Stop Clustering and Liquidity Cascade
Abstract
This strategy enters long at a breakout above the prior trading day's high, or short at a breakout below the prior day's low, based on the observation that traders cluster stop-loss orders at obvious technical levels. The rationale is that concentrated stops create liquidity cascades when price pierces these coordinated levels, leading to sharper breakouts than would occur at arbitrary recent extremes lacking stop concentration.
Why this might work
Technical traders typically place stop-loss orders at round numbers and recent swing extremes as a coordination mechanism for risk management. [1] The prior trading day's high and low are salient reference points embedded in widely available charting software and news commentary, making them natural candidates for stop placement. [2]
When price approaches a level with clustered stops from the near side, triggering those stops removes liquidity precisely where new traders are attempting entry, potentially amplifying price movement beyond what information arrival alone would predict. [3] This differs qualitatively from arbitrary recent extremes, which lack the psychological coordination signal and may carry lighter concentrations of resting orders. Practitioner convention in FX and futures markets attributes the violence of clean breakouts through obvious support and resistance to stop cascades. [4]
Importantly, this is observational and partially speculative. Empirical evidence that prior-day levels specifically accumulate materially more stops than competing levels (round numbers, moving averages, prior-week extremes, Fibonacci levels) is limited. [5] The cascade mechanism is testable, however, and a strategy can measure its consequences if they persist.
The rules
Instrument and timeframe: Liquid futures (ES, NQ, CL, GC, 6E) or forex (EURUSD, GBPUSD, NZDUSD) with tight bid-ask spreads and high intraday volume. Trade on a daily timeframe: one entry signal per calendar day per instrument.
Entry trigger (long): Enter a long position if price closes above the prior trading day's high on volume greater than the 20-period simple moving average of volume.
Entry trigger (short): Enter a short position if price closes below the prior trading day's low on volume greater than the 20-period SMA.
Entry price: Place a limit order 2 ticks beyond the level (2 ticks above the prior high for long entries; 2 ticks below the prior low for short entries). This offset reduces exact-level slippage and models realistic order placement.
Stop-loss: For long entries, set the stop 1.5 times the 14-period Average True Range below the entry price. For short entries, set the stop 1.5 times ATR above the entry price. This scales stops with volatility and reduces whipsaw on thin-range days.
Exit: Exit at the stop-loss trigger or at end-of-session close, whichever comes first. Intraday execution only.
Position sizing: Risk 2% of account equity per trade.
Session filter: Trade only during the primary session of the instrument (9:30-16:00 ET for US equity index futures; 22:00-21:00 ET for currency pairs). Avoid the final 15 minutes before session close to reduce gap risk overnight.
Expected trade frequency: 150-250 signals per year on a liquid instrument, generating a statistically meaningful sample.
Code
//@version=6
strategy("Prior-Day High/Low Breakout: Stop Cascade", overlay=true, default_qty_type=strategy.percent_of_equity, default_qty_value=2)
// Inputs
atr_period = input.int(14, title="ATR Period", minval=1)
stop_atr_mult = input.float(1.5, title="Stop Loss ATR Multiplier")
vol_sma_period = input.int(20, title="Volume SMA Period")
vol_threshold_mult = input.float(1.0, title="Volume Threshold Multiplier")
entry_offset_ticks = input.int(2, title="Entry Limit Offset (ticks)")
qty_pct = input.float(2.0, title="Position Size (% of equity)")
// Indicators
atr_val = ta.atr(atr_period)
vol_avg = ta.sma(volume, vol_sma_period)
// Prior day high and low (from daily timeframe)
prior_high = request.security(syminfo.tickerid, "D", high[1])
prior_low = request.security(syminfo.tickerid, "D", low[1])
// Track entry state per day
var bool entered_today = false
if ta.change(time("D")) != 0
entered_today := false
// Entry price levels
tick_size = syminfo.mintick
long_entry_lvl = prior_high + entry_offset_ticks * tick_size
short_entry_lvl = prior_low - entry_offset_ticks * tick_size
// Stop loss levels
long_stop_lvl = long_entry_lvl - stop_atr_mult * atr_val
short_stop_lvl = short_entry_lvl + stop_atr_mult * atr_val
// Entry conditions
long_breakout = close > prior_high and volume > vol_avg * vol_threshold_mult
short_breakout = close < prior_low and volume > vol_avg * vol_threshold_mult
// Execute entries
if long_breakout and not entered_today and strategy.position_size == 0
strategy.entry("long", strategy.long, qty=strategy.calc_percent_of_equity(qty_pct / 100), limit=long_entry_lvl)
entered_today := true
if short_breakout and not entered_today and strategy.position_size == 0
strategy.entry("short", strategy.short, qty=strategy.calc_percent_of_equity(qty_pct / 100), limit=short_entry_lvl)
entered_today := true
// Exit on stop loss
if strategy.position_size > 0
strategy.exit("long_exit", from_entry="long", stop=long_stop_lvl)
if strategy.position_size < 0
strategy.exit("short_exit", from_entry="short", stop=short_stop_lvl)
// Visualize prior-day levels
plot(prior_high, color=color.green, linewidth=2, title="Prior High")
plot(prior_low, color=color.red, linewidth=2, title="Prior Low")
How the code works
Lines 16-17 calculate the 14-period ATR and the 20-period simple moving average of volume, both standard indicators of volatility and participation. Lines 20-21 retrieve the prior trading day's high and low using request.security(..., "D", high[1](#ref-1)) and request.security(..., "D", low[1](#ref-1)). This cross-timeframe data fetch ensures that even on intraday charts (1-minute, 5-minute, or any finer resolution), the code reads the daily OHLC from the exchange feed, with [1](#ref-1) referring to the prior daily bar.
Lines 24-26 implement a state flag entered_today that resets each time the calendar day changes (detected by ta.change(time("D"))). This ensures the strategy allows only one entry per session and prevents multiple fills on the same prior-day level.
Lines 29-32 calculate entry price levels 2 ticks beyond the prior-day high or low, modeling realistic limit order placement and reducing slippage at exactly the round level. Lines 35-36 set stop-loss levels 1.5 ATR away from entry, automatically scaling with market volatility to reduce whipsaws in calm and choppy regimes alike.
Lines 39-40 define entry signals: price must close beyond the prior-day level AND volume must exceed the 20-bar average. Both conditions together filter weak or low-volume breakout attempts. Lines 43-49 execute the entry if the signal fires, no position is currently open, and no entry has already occurred today. Lines 50-56 submit stop-loss orders that persist until triggered. Lines 59-60 plot the prior-day reference levels on the chart for visual verification that the strategy is reading the correct levels.
Testing it honestly
Backtest on at least 3-5 years of historical data, splitting the period: 70% in-sample (for parameter tuning) and 30% out-of-sample (for validation against fresh data). Before accepting any result, verify that the strategy generates at least 150-200 trades per year. Fewer than that, and random variation dominates any edge: a 55% win rate on 50 trades per year is noise. A 52% win rate on 200 trades is statistically meaningful.
Include realistic costs: futures commissions of $4-10 per round-turn, and slippage of 1-5 ticks at breakout entry (breakouts are inherently high-slippage events). If the strategy's gross profit before costs is less than 20-30% above total costs, suspect severe overfitting or a marginal edge that will not survive live execution.
Test separately on high-volatility and low-volatility periods. If stop-clustering effects exist, they may evaporate during dislocations (when all levels break indiscriminately) or during range-bound markets (when few breakouts form). Perform walk-forward optimization: retrain parameters monthly on the latest data, then test the refined parameters on the following period. If parameters drift sharply (ATR period from 12 to 30, stop multiplier from 1.5 to 3.0), overfitting is occurring.
A strategy passing these tests still does not guarantee future profit. Past performance is not predictive, market structure changes, and regimes shift unpredictably.
Limitations
Stop clustering is observational, not empirically proven: The premise that traders specifically concentrate stops at prior-day highs and lows lacks rigorous verification. No full public dataset exists documenting where retail or institutional traders actually place stops. [6] The strategy assumes prior-day levels accumulate materially more stops than competing levels such as round numbers, moving averages, or Fibonacci ratios, but this assumption is unverified and may be false.
Entry and exit are inherently costly: Breakouts of support and resistance are notorious for high slippage. Bid-ask spreads widen at round levels, and a limit order 2 ticks beyond the level risks missing the fill entirely if price moves fast. On slower instruments or during news events, slippage can reach 10-50 ticks, easily eliminating a thin edge.
Stop effects evaporate during dislocations: Cascading stops are relevant only in orderly markets where support and resistance matter. During flash crashes, gap openings across economic data, or extreme volatility, stops trigger indiscriminately regardless of level, and the differentiation between prior-day levels and other levels collapses.
Multiple clustering candidates compete: Traders place stops at prior-day extremes, and at round numbers (100, 5000, 4500), prior-week levels, moving averages (200-day MA), Fibonacci retracements, and other technical reference points. If stops are distributed across many levels, the prior-day level may not accumulate enough clustering to matter.
Overfitting risk from many tunable parameters: The strategy contains multiple inputs (ATR period, stop multiplier, volume threshold, tick offset, position size). Optimizing these parameters on historical data to improve backtest results is nearly inevitable and almost always produces curve-fitted results that fail on live data.
Key evidence is missing: Does the prior-day level accumulate materially more stops than arbitrary levels? Do cascade effects translate to favorable fills compared to non-cascade breakouts? Does the effect persist across all instruments, sessions, and volatility regimes? The strategy ships untested: these questions remain unanswered.
Key definitions
Support and resistance: Price levels where historical buying or selling has congregated, or where traders expect it to occur, creating mechanical barriers to further movement.
Stop-loss order: An instruction to automatically exit a position if price moves against the trade to a specified level, used to limit maximum loss per trade.
Prior-day high and low: The highest and lowest prices achieved during the most recently completed trading session, widely accessible in charting software and news.
Volume threshold: A condition requiring entry-bar volume to exceed a simple moving average of recent volume, filtering weak breakouts that lack participation.
ATR (Average True Range): A volatility measure computed as the average of true ranges over a specified period, used to scale position sizing, stops, and targets with market conditions.
Limit order: An order to buy or sell only at a specified price or better, offering control over entry cost but risking a miss if price moves past the limit level.
Tick size: The minimum price movement defined by the exchange (e.g., $0.01 for equities, $0.0001 for forex, $0.25 for some futures contracts).
References
-
Investopedia, "Support and Resistance", https://www.investopedia.com (accessed 2026). Describes technical analysis convention for stop placement at swing extremes and round numbers.
-
Pring, M. J., "Technical Analysis Explained", McGraw-Hill (1991). Classical textbook reference on support and resistance as risk management reference points and trading levels.
-
Kirilenko, A. A., Kyle, A. S., Samadi, M., and Tuzun, T., "The Flash Crash: High-Frequency Trading in an Electronic Market", Journal of Finance, Vol. 72, No. 3 (2017): 967-998, https://doi.org/10.1111/jofi.12498. Documents liquidity cascades at technical price levels during market stress.
-
CME Group, "Foundations of Trading", https://www.cmegroup.com/education (accessed 2026). Practitioner and educational material reflecting convention that stop cascades amplify breakouts through obvious technical levels.
-
No peer-reviewed study quantifying whether prior-day levels specifically accumulate materially more stops than other technical levels is publicly available.
-
Stop-placement data across institutional and retail market participants remains proprietary and largely inaccessible for independent research.
Educational research on historical data only. Not investment advice, not a signal, and never a performance promise. Past results do not predict future performance. Every reference is link-verified before publication and every paper is re-audited weekly against the library's editorial standard.
Last reviewed by the PropLedger research pipeline: 2026-11-08. Educational research on historical data, not financial advice.
Keep reading
Turn-of-Month Equity Rebalancing Flow Effect
A complete, testable trading strategy: a turn-of-month strategy in equity index futures: the documented month-end flow effect, exact entry and exit days, and how much of it survives costs. Exact rules, full Pine Script code, and an honest reading of the evidence.
Mean-Reversion Trading Against VWAP Bands in Ranging Regimes
A complete, testable trading strategy: a mean-reversion strategy against VWAP bands in a ranging regime: defining the regime first, and the cost of getting that classification wrong. Exact rules, full Pine Script code, and an honest reading of the evidence.
Inside-Bar Breakout Strategy for Futures
A complete, testable trading strategy: an inside-bar breakout strategy on futures: rules, filters, and why most published versions do not survive realistic costs. Exact rules, full Pine Script code, and an honest reading of the evidence.