ORB Trading Examples
Individual sessions narrated as sequences of decisions rather than outcomes. A trade that worked, a trade that lost without a mistake in it, and a day that never resolved into a clear answer in either direction.
A Description of a Session Is Not a Recommendation
Worked examples carry an obvious hazard. A single session told after the fact reads as evidence for whatever the narrator set out to demonstrate, and the outcome colours every decision described along the way to it. The accounts here are written to limit that as far as possible, by stating the information available at each point before saying what happened next. The purpose is to show the sequence a decision runs through, not to suggest that a session shaped like this one will resolve the same way when you meet it.
The Sequence of Decisions
A session contains more decisions than most write ups admit to. Whether to measure the range at all given what is on the calendar, where the range actually sits once the period closes, whether its shape and width permit a trade, which side to work, how the entry is placed, what the stop distance implies for size, and then a series of choices about holding, trimming and exiting. Each one is separable and each one can be right while another is wrong, which is the main reason a single outcome tells you so little about the process that produced it.
Why the Loser Is the More Useful Example
Most people learn more from a losing session described honestly than from a winning one. A winning trade hides its mistakes, because the outcome absolves them, and it teaches habits that happened to survive rather than habits that work. A loss taken strictly by the book demonstrates the thing that is genuinely difficult to internalise, which is that a correctly executed trade is allowed to lose and that nothing whatsoever needs changing when it does.
Sessions That Refuse to Resolve
Between the two clean cases sits the ordinary majority. The range is neither obviously too wide nor comfortably normal, the break is not obviously false but does not run either, and the correct action is nowhere written on the chart. These are the sessions where discretionary damage gets done, because the absence of a clear answer is an invitation to improvise. Describing one of them in detail is more useful than describing ten textbook setups, since the textbook ones were never the problem.
Three Sessions in Detail
Three sessions narrated as sequences of decisions: one that worked, one that did not and was still traded correctly, and one that stayed ambiguous from the first minute to the last. Prices and instruments are deliberately left unspecified, because the numbers from any particular session are not transferable while the order of the questions is. Each account ends where the position ends, without a verdict about what the session supposedly proved.
Latest Guides
A Clean Loser That Was Still Correctly Traded
This session lost money and contains no mistakes. That combination is uncomfortable enough that most people, reviewing it later, will invent a mistake in order to explain the loss. The invention is the actual error, because it produces a rule change that the evidence does not support and that will then apply to every session afterwards.
The Setup Passed Every Filter
The calendar was clear. The range formed at an ordinary height, neither compressed nor stretched, and its shape was balanced with both edges tested more than once. Every question the plan asks before a trade returned an acceptable answer, in the same order and with the same checks as on a session that works out.
This is worth dwelling on because it is the entire point. The filters are not there to select winners, which is not something filters can do. They are there to exclude the situations where the arithmetic is against you before the trade begins. A session passing all of them can still lose, and if it could not, the filter would be a prediction rather than a filter.
The Entry Filled and Went Nowhere
The resting order triggered on a break of the upper edge, filled slightly beyond the trigger, and price moved a short way in the intended direction. Then it stalled.
There was no dramatic reversal and no obvious moment at which the trade was clearly wrong. Price simply stopped extending, drifted sideways above the level, then a little below it, then back inside the range. The absence of a dramatic signal is characteristic. Most losing breakouts do not announce themselves, they just fail to continue, and the failure is only visible in retrospect.
The Decision Not Taken
The interesting part of this session is a decision that was available and was not made. Once price returned inside the range, the position was clearly not doing what it had been entered to do, and closing it there would have produced a smaller loss than the stop eventually did.
The plan did not contain that action, so it was not taken. That is defensible and it is not obviously optimal. An exit rule that closes a position when price re enters the range is a perfectly legitimate design, and had it been in the plan it would have been correct to use it here. What is not legitimate is inventing it mid trade, because the version invented mid trade is the version that fires whenever a position becomes uncomfortable, which is most of the time and includes a great many trades that go on to work.
The correct home for that idea is the review. Test it against the whole record, adopt or reject it as a rule, and then apply it to every session rather than only to the ones where hindsight shows it would have helped.
The Stop Executed
Price continued through the range and reached the opposite edge, where the stop was waiting. It filled a little worse than the level, for the same structural reason entries do, and the realised loss came out slightly larger than the planned figure.
That gap between planned and realised loss is worth recording rather than shrugging off. It is small on any single trade and it is systematic, which means it belongs in the expectancy calculation. A strategy evaluated on planned risk rather than realised risk is being evaluated at a discount it does not receive in practice, and the discount is applied to every trade in the record.
Grading the Trade Instead of the Outcome
Reviewed against the specification, every step here was correct. The session qualified, the range qualified, the entry was placed as written, the size followed from the stop distance, and the exit was the one that had been planned. The grade is clean and the money is gone, and those two facts coexist without contradicting each other.
The value of grading this way is that it makes a losing run diagnosable later. A month of clean losses is a sample from a distribution and calls for patience. A month of losses on trades that were early, oversized or exited by hand is not information about the strategy at all and calls for something else entirely. Without the grading, both months look identical on the equity curve, and the response to them is usually the same undirected panic.
A Clean Winner Narrated Decision by Decision
This is a session that went the way the rules intended, described in the order the information actually arrived. It is included not because winners are instructive in themselves but because it establishes the sequence that the other two accounts depart from.
Before the Open
The first decision was made before any price existed. The economic calendar showed nothing significant scheduled inside the first hour, which meant the range about to form would be the product of ordinary participation rather than a reaction to an event. Had a release been due, this session would have been skipped or the measurement window shifted, and neither of those is a judgement call once the rule has been written down.
The second pre session step was to recall where the instrument's recent opening ranges had been sitting in height. Not a calculation, just a rough memory of the last several sessions, so that whatever formed could be described as narrow, normal or wide rather than simply as a pair of numbers with no context attached.
The Range Completes
The period ran and the range came out close to typical height, marginally under it. That passed the width filter immediately, and passing it mattered more than it might appear. A normal width range means the stop at the far edge is a normal distance away, which means position size is the usual size, and it means the distance available beyond the break has not already been spent producing the range.
The shape was closer to balanced than one sided. Price had touched both edges more than once and finished somewhere around the middle. Both extremes had been reached and rejected, so both were levels somebody had defended, and the far edge was therefore a defensible place to hold a stop rather than an arbitrary point price happened to reach once on a single shallow dip.
Placing the Order
With the range acceptable, the entry was a resting stop order placed a small distance beyond the upper edge rather than exactly on it. The offset exists to filter the single tick probe, and it is a fixed part of the rule rather than something judged fresh each morning.
Size followed from the stop, not the other way around. The stop went at the opposite edge, the distance between the entry and that edge converted into money at risk per unit, and the position was whatever number of units held the total at the predetermined figure, rounded down. On this session that produced a slightly smaller position than usual, because the entry offset had lengthened the stop distance a little. That is the arithmetic working correctly, not a problem in need of a fix.
The Break and What Followed
Price approached the upper edge, hesitated below it for a few minutes, then traded through. The order filled a little worse than the trigger, which is the ordinary cost of transacting at that instant and was expected rather than noted with any surprise.
What made this session clean was what happened next. Price did not return to the level. It extended away from the range and held its distance from the edge, which is the behaviour the strategy is paid for and the behaviour that is absent on most days. There was no decision to make during that stretch, and making one would only have been a way of interfering with something that was working.
Managing and Exiting
Part of the position came off at a distance related to the range height, which is a mechanical target rather than a judgement about where the move ought to end. The remainder was held with the stop moved to the entry price, so the trade could no longer produce a loss whatever happened afterwards.
That remainder eventually exited on the trailing arrangement rather than at a chosen price, giving back some of the best level the move reached. Giving something back is a feature of any trailing exit and it is the price of leaving a longer move possible. Complaining about it afterwards amounts to saying the trade should have been closed at the high, which was not knowable at the time and will not be knowable next time either. The session finishes without a verdict, because one winner says almost nothing about a strategy and the reason for writing it down was the order of the steps rather than the number at the end of them.
An Ambiguous Session and How to Handle One
The two previous sessions were legible. Most are not. This one is worth describing because the majority of trading days sit somewhere in the middle, offering neither a clean signal nor a clear reason to stay out, and the middle is where discretionary damage actually gets done.
The Range Was Neither One Thing Nor the Other
The calendar carried a second tier release, the kind that sometimes moves the market and often does not, scheduled shortly after the range period closed. Not obviously a reason to skip, not obviously ignorable either.
The range then came out wider than typical but not dramatically so. It landed in the region where a rule expressed as a multiple of normal height gives an answer only if the multiple was written down precisely. Vague versions of the rule, something along the lines of avoid unusually wide ranges, provide no answer at all here, which is the first thing this session teaches: an imprecise filter is not a filter, it is a place for a preference to hide.
The shape was mixed as well. One edge had been tested repeatedly, the other had been reached once on a single quick move and never revisited. Half balanced, half one sided, and the plan had nothing to say about that combination.
A Break That Did Not Commit
Price broke the well tested edge, extended a short distance, and came straight back to the level. Then it hovered, trading either side of the line for several minutes without settling on a direction.
This is the condition that produces the most improvisation. Holding a resting order meant being filled on the first break automatically. Watching meant being genuinely uncertain whether a break had occurred at all. Either way the position, if taken, was neither working nor clearly failing, and no new information was arriving that would resolve it one way or the other.
Both Edges Went
Later in the session price left the range on the other side, at which point both edges had been broken and neither break had led anywhere. A trader taking every signal had two losses. A trader waiting for confirmation had most likely taken the second break, which is the one arriving after a failed first attempt, at the moment when conviction is lowest and the setup is least distinguishable from noise.
The instrument had spent the morning demonstrating that neither level meant anything, and the strategy, which assumes the levels mean something, had no premise left to work from. That is the honest description of the day, and it is not a description that suggests trying harder would have helped.
The Options at That Point
Three responses are defensible, and they should be chosen between beforehand rather than discovered under pressure. The first is a one trade rule: the first signal is taken, and once it fails the session is over regardless of what happens later. This caps the damage on precisely this kind of day and forfeits the occasional second break that runs.
The second is to stand down whenever the range shape is mixed, on the reasoning that a range with one untested edge is only half a range. That skips some workable sessions in exchange for avoiding most of the confusing ones.
The third is to take both signals at reduced size, accepting that ambiguous days will produce small losses and that the occasional day where the second break runs will pay for some of them. All three are coherent positions. What is not coherent is choosing among them at the moment the second break appears, because by then the first loss is already influencing the choice and the decision is no longer being made on the merits.
Choosing Beforehand What Ambiguity Means
The general point is that ambiguity is not an unusual event to be handled with judgement in the moment. It is the ordinary condition of most sessions, and it deserves a written response in exactly the same way a wide range or a scheduled release does.
A plan specifying what to do when the setup is clean, and leaving everything else to be settled on the day, has specified the easy cases and abandoned the hard ones. The rules that matter most are the ones covering the sessions where you would rather not have a rule, because those are the sessions where you will otherwise do something you cannot describe afterwards, and a decision you cannot describe is one you cannot learn anything from.
