Free Optimization Guide

Enhanced ORB Strategy Optimization Guide

Exact Start / Step / End values for the NinjaTrader 8 Strategy Analyzer (Optimize tab), the order to optimize them in, and how to avoid curve-fitting.

HOW THE SYSTEM WORKS

Builds an Opening Range from OrStartTime to OrEndTime, marking the high and low of that window.
Trades a separate, independently configurable window from TradeStartTime to TradeEndTime. No trade can occur until the OR period has closed AND the trade window is open.
Entries fire on a closed-bar cross: long on CrossAbove(Close, OR High), short on CrossBelow(Close, OR Low). The signal is evaluated against the bar's close, not an intrabar wick.
Two optional entry filters (minimum breakout size, volume vs. 20-period average) sit in front of every signal.
Two independent stop-loss modes (fixed-tick, OR-level) plus a profit target, breakeven, and trailing stop handle exits.

This strategy is fully deterministic: pure price/volume/time logic, no machine learning, no trained model, and no randomness anywhere in the code. The same settings on the same data produce the identical result every single run. A single backtest is trustworthy as-is - there is no reproducibility check to perform and no non-determinism warning needed anywhere in this guide.

UNITS: TICKS

All *Ticks inputs (StopLossTicks, TargetTicks, etc.) are in ticks, the instrument's own minimum price increment (e.g. 0.25 on ES = 1 tick). NinjaTrader has no 4-digit/5-digit broker ambiguity - a tick is a tick for the instrument you're testing. Just confirm your chart's instrument and TickSize before typing the ranges below; scale them up for wider-ranging instruments (e.g. NQ, CL) and down for tighter ones.

STRATEGY ANALYZER SETUP

Use real historical tick or minute data for the instrument/session you intend to trade. Garbage data = garbage results.
Period: 1-2+ years of the instrument's regular session. Hold back the most recent ~25-30% for out-of-sample validation, or use the Strategy Analyzer's Walk-Forward Optimization feature to automate the split.
Optimizer: NinjaTrader 8 Strategy Analyzer, Optimize tab. Use the Genetic optimization method once a table's grid gets large; use Exhaustive/Brute Force for small grids (e.g. a single bool).
Fitness/sort: Profit Factor or a custom criterion that weighs drawdown - don't sort by Net Profit alone (see READING RESULTS).
Bools optimize as Start 0, Step 1, End 1 (tests both false/off and true/on).

TIME PARAMETERS (set/test manually, do NOT numeric grid-sweep)

OrStartTime / OrEndTime : the Opening Range window TradeStartTime / TradeEndTime : the trade window (must start at or after OrEndTime)

These are DateTime/time-of-day properties, not plain integers - the Strategy Analyzer's Optimize tab does not sweep DateTime inputs in a useful way. Instead, test a handful of sensible combinations by hand, e.g. a 15/30/45-minute OR window measured from the session open, with the trade window starting exactly at OrEndTime and running to your session's usual close. Compare a few full runs and lock the winner before touching anything else.

OPTIMIZATION TABLES (Start / Step / End)

GROUP A: CORE - OR WINDOW + TRADE WINDOW (set first, manually)

Input	Default	Note
OrStartTime	09:30	session open; test manually, not a sweep
OrEndTime	10:00	try OR windows of 15 / 30 / 45 min from open
TradeStartTime	10:00	keep equal to OrEndTime unless testing a delay
TradeEndTime	15:45	usual pre-close cutoff; test a couple of values

(No Start/Step/End table here - hand-pick a few combinations, run each as a full backtest, and lock the best before moving to Group B.)

GROUP B: STOP LOSS

Input	Start	Step	End	Note
UseStopLoss	0	1	1	bool, default true: enables the fixed-tick stop (Pass 1 assumes this is true; see below)
StopLossTicks	20	20	200	fixed-tick stop (default 80)
UseORLevelStops	0	1	1	bool: fixed-tick vs OR-level stop
ORStopBufferTicks	0	2	20	only if UseORLevelStops = true

Run this as TWO SEPARATE optimization passes, not one combined grid: Pass 1 - UseORLevelStops forced FALSE, UseStopLoss forced TRUE: sweep StopLossTicks alone. (Leaving UseStopLoss false here would mean the position has NO stop loss at all, since UseORLevelStops is also false in this pass - don't test that combination unless you deliberately want a no-stop run.) Pass 2 - UseORLevelStops forced TRUE: sweep ORStopBufferTicks alone (StopLossTicks and UseStopLoss are irrelevant to this pass).

The strategy's own entry logic picks the OR-level stop first whenever UseORLevelStops is true (see the code's "IMPORTANT: UseORLevelStops works independently of UseStopLoss" comment) - StopLossTicks is not even read for stop placement in that case, so optimizing both stop inputs in the same grid wastes iterations on a combination that never resolves the way the grid implies. Pick a winner from each pass, then carry forward whichever stop MODE (fixed-tick or OR-level) produced the better, more stable result into Group C.

GROUP C: PROFIT TARGET

Input	Start	Step	End	Note
TargetTicks	40	20	300	profit target (default 120)

Optimize Group B and Group C TOGETHER (in whichever stop-mode pass you're running) - stop distance and target distance interact directly through the risk:reward ratio and win rate, so sweeping TargetTicks against a single fixed StopLossTicks (or ORStopBufferTicks) value will miss the real edge. Run the full StopLossTicks x TargetTicks grid (or ORStopBufferTicks x TargetTicks grid) for each stop-mode pass.

GROUP D: BREAKEVEN (default ON)

Input	Start	Step	End	Note
BreakEvenTriggerTicks	20	10	100	default 40
BreakEvenOffsetTicks	0	5	30	default 10

(UseBreakEven defaults to true in this strategy. See GROUP E below for why you should not enable Group E at the same time.)

GROUP E: TRAILING STOP (optional, OFF by default)

Input	Start	Step	End	Note
TrailingStopTriggerTicks	20	10	150	default 60
TrailingStopOffsetTicks	10	10	100	default 30

CRITICAL: Never enable UseBreakEven and UseTrailingStop together, and never optimize Group D and Group E in the same pass. With the defaults, breakeven's trigger (40 ticks) is LOWER than trailing's trigger (60 ticks), so breakeven fires first and moves the stop via SetStopLoss() while the trailing-stop logic is independently placing its own initial protective stop order every bar until ITS trigger is reached - two stop-management paths acting on the same position before trailing ever activates. The strategy's own code has an explicit comment warning about this conflict. Also note: whenever UseTrailingStop is true, the entry-time stop selection (fixed-tick or OR-level, Group B) is skipped entirely - the trailing-stop system manages its own initial stop using StopLossTicks regardless of UseORLevelStops. If you want the OR-level stop mode from Group B, keep UseTrailingStop off. Pick ONE of Group D or Group E, never both.

GROUP F: TRADE MANAGEMENT

Input	Start	Step	End	Note
MaxTradesPerDay	1	1	3	default 1
RequireClosedCandleBreakout	0	1	1	bool, default true

GROUP G: ENTRY FILTERS (one at a time)

Input	Start	Step	End	Note
UseMinimumBreakoutSize	0	1	1	bool, default true
MinimumBreakoutTicks	2	2	20	only if filter on (default 5)
UseVolumeFilter	0	1	1	bool, default true
VolumeMultiplier	0.8	0.2	2.0	only if filter on (default 1.2)

SET, DON'T OPTIMIZE

ContractQty : your own position-size choice, not an edge parameter. Set it to what you'll actually trade live; optimizing it only rescales P&L, it doesn't change the edge.

DO NOT TOUCH (purely cosmetic, no effect on trading logic)

ShowORBox, ORBoxOpacity, ShowORLines, ORLineWidth, ORBoxColorSerialized / ORBoxColorSerializable, ORLineColorSerialized / ORLineColorSerializable.

PHASES (order; lock each before the next)

Pick the OR window and trade window (Group A) - manual, a few sensible combinations (e.g. 15/30/45-minute OR windows from the session open). Lock it before proceeding.
Optimize Group B + C TOGETHER, run as two separate stop-mode passes (fixed-tick StopLossTicks x TargetTicks, then OR-level ORStopBufferTicks x TargetTicks). Compare the two passes and lock whichever stop mode + target combination wins.
Group F (trade management / frequency) - MaxTradesPerDay, RequireClosedCandleBreakout.
Group G (entry filters) - one filter at a time: first UseMinimumBreakoutSize/MinimumBreakoutTicks alone, then UseVolumeFilter/VolumeMultiplier alone, keeping the other at its locked/default state while you test each.
Choose ONE of Group D (breakeven) or Group E (trailing stop) as a final optional overlay on top of the already-locked settings from steps 1-4. Never enable both. Optimize it last, against the locked OR window, trade window, stop mode, target, trade management, and filters.

READING RESULTS / VALIDATION

Don't sort by Net Profit alone. In the Strategy Analyzer's optimization results grid, weigh:
- Profit Factor > 1.2 (>1.5 is strong)
- Max Drawdown - can you stomach it live?
- Sharpe Ratio - rewards smoother, more consistent equity
- Total trades - 100+ over the test period is ideal; under 30 is noise, don't trust it
CLUSTER TEST: the best row should be surrounded by other good rows in the grid. If its neighbors are losers, it's a curve-fit - discard it and pick from a broad, stable plateau of good results instead of an isolated spike.
OUT-OF-SAMPLE VALIDATION: use the Strategy Analyzer's Walk-Forward Optimization feature, or do it manually - optimize on an older slice of data, then run the locked winner UNCHANGED on a held-out, more recent slice. Keep the setting only if it holds up with similar character on the unseen data; if it collapses, it was curve-fit - go back and pick a broader, coarser setting from the cluster.
Repeat per instrument and per session, since OR window, filters, and stop/target behavior are all session- and instrument-dependent.