Building a 3-Option Decision Tree
Execution Precision
8 min read
Create a simple hold/reduce/exit decision tree that removes ambiguity from trade management under pressure.
Loading
8 min read
Create a simple hold/reduce/exit decision tree that removes ambiguity from trade management under pressure.
Most traders think in binary: hold or close. Professionals think in three options — hold, reduce, or exit — and they know which one to apply at every decision point.
When a trade is under pressure, the untrained mind defaults to two choices: stay in or get out. This binary frame creates a psychological trap. Holding feels like stubbornness. Exiting feels like failure. Neither option satisfies, and the trader oscillates between them, often choosing the wrong one at the worst time.
The solution is to expand your decision space. With three options — hold, reduce, exit — you introduce a middle path that allows you to de-risk without fully abandoning a position, to acknowledge uncertainty without paralysis, and to manage capital with precision rather than panic.
At every management decision point during a live trade, you have exactly three choices:
Hold — Maintain current position size. The thesis is intact, conditions support the trade, and no adjustment is warranted.
Reduce — Cut position by a pre-defined percentage (typically 30-50%). The thesis is weakening or uncertainty has increased, but outright invalidation has not occurred.
Exit — Close the entire remaining position. The thesis is invalidated, the context has broken, or your bail triggers have been hit.
The power of this framework is that it eliminates the need for certainty. You do not need to know whether the trade will work to make a sound management decision. You only need to assess where conditions fall on the spectrum from confirming to invalidating.
Hold is the correct choice when:
The most important aspect of holding is doing nothing when nothing needs to be done. Holding is an active decision, not a passive one. You are choosing to maintain exposure because the evidence supports it.
Reduce is the correct choice when:
Reduce when: Thesis is wounded but alive Reduce by: 30-50% of remaining position Effect: Cuts risk by half while preserving upside if the thesis recovers
The reduce option is what separates reactive traders from strategic ones. It lets you take chips off the table without conceding the entire position. If the thesis recovers, you still participate in the move. If it fails, your loss is materially smaller.
Exit is the correct choice when:
Exit means the trade is over. Not paused, not reduced — closed. The decision to exit should be followed by stepping back, logging the trade, and waiting for the next setup. Do not re-enter the same trade on impulse after exiting.
You entered a long on BTC/USDT based on a higher-low continuation pattern on the 5m chart.
| Decision Point | Condition | Action |
|---|---|---|
| 10 min after entry | Higher low holding, delta positive, price grinding up | Hold |
| 20 min after entry | Price stalls near minor resistance, delta flattens | Hold — stall at resistance is expected |
| 30 min after entry | Price retraces 50% of the move, still above higher low | Hold — normal pullback in trend |
| 30 min after entry | Price retraces 50%, delta turns negative, volume increasing on decline | Reduce 50% — thesis weakening |
| 40 min after entry | Higher low broken on 5m close | Exit — structure invalidated |
| 40 min after entry | Price bounces from retracement, delta turns positive again | Hold remaining — thesis recovering |
Used 3-option tree: held through pullback, reduced at delta divergence, exited remainder at structure break.
BTC/USDT long from 5m higher low. After 25 minutes, price reached $63,750 (+$350) then pulled back to $63,500. Delta diverged — price held but buyers were weakening. Applied the Reduce decision: closed 50% at $63,500 for +$100. Remaining 50% stayed with stop at $63,000. Price continued to drop, breaking the higher low at $63,350. Applied the Exit decision on the remainder at $63,300 for -$100. Net result: roughly breakeven instead of a -$400 full stop loss.
Worked example: 3-option management on a BTC/USDT long
| Stage | Price | Action | Slice P&L |
|---|---|---|---|
| Entry | $63,400 | Open long | — |
| +25 min | $63,750 | Peak (+$350 unrealized) | — |
| Pullback plus delta divergence | $63,500 | Reduce 50 percent | +$100 on closed half |
| Higher low broken | $63,300 | Exit remainder | -$100 on remaining half |
| Net result | — | Breakeven | vs -$400 full stop |
You shorted BTC/USDT at the top of a 4H range, expecting a move back to the midpoint.
| Decision Point | Condition | Action |
|---|---|---|
| Price drops 40% toward target | Moving as expected, no issues | Hold |
| Price stalls at midrange | Volume declining, range narrowing | Reduce 30% — momentum fading before target |
| Price reverses from midrange back toward entry | Buyers stepping in at midrange support | Exit remainder — thesis failed at midrange |
| Price breaks below midrange with volume | Fresh momentum, sellers in control | Hold remaining (default); add-back only if your playbook explicitly permits and total-position risk cap is not breached — most playbooks should disallow this for simplicity |
You are long BTC/USDT on an intraday setup when an unscheduled macro headline drops.
| Decision Point | Condition | Action |
|---|---|---|
| Headline is ambiguous, market whipsaws | No clear direction, volatility spikes | Reduce 50% — protect capital during uncertainty |
| After 2 minutes, price stabilizes above entry | Headline absorbed, trend resuming | Hold remaining |
| After 2 minutes, price breaks below entry with aggressive selling | Headline is bearish, flow confirms | Exit — environment has changed |
| Headline is clearly bullish for your position | Confirmation, potential acceleration | Hold full (or hold remaining) |
Decide before any trade what your standard reduction size will be. Most traders use 50% because it is simple and meaningful. Some prefer 30% for the first reduction and another 30% for the second, leaving 40% as the runner. The exact number matters less than having it defined in advance so you do not deliberate during the trade.
Follow this process to construct a decision tree for any setup in your playbook:
Step 1: Identify your decision points. These are the moments during a trade where you will evaluate conditions. They can be time-based (every 10 minutes), price-based (at each structural level), or event-based (on flow shifts or news).
Step 2: For each decision point, define the conditions for Hold, Reduce, and Exit. Be specific. "Things look bad" is not a condition. "5m higher low broken with delta making new lows" is a condition.
Step 3: Define your reduction size. Typically 30-50% of the remaining position. Write it down.
Step 4: Define what happens after a reduction. Does the stop on the remaining position change? Does the target change? What would cause you to exit the rest?
Step 5: Test the tree on historical trades. Go through your last 20 trades and apply the decision tree retroactively. Compare the outcomes to what actually happened. Adjust the tree based on the data.
At each decision point, evaluate:
The reduce option is the hardest to execute consistently because it requires accepting uncertainty. Holding feels like conviction. Exiting feels like decisiveness. Reducing feels like indecision — but it is not. Reducing is the most sophisticated response to ambiguity.
Over a sample of 100 trades, consistent use of the reduce option at appropriate decision points typically:
Impact of disciplined reduce-option use over 100 trades
Midpoints of cited ranges: loss size -27.5 percent, win size -12.5 percent
Decision trees do not eliminate variance — they bound it. A perfectly calibrated tree applied to a low-edge setup still loses money. The tree improves consistency of execution, not the underlying expectancy.
Reducing a position is not admitting defeat. It is allocating capital proportionally to your confidence level. If your confidence is at 50%, your position should reflect that. Full positions should be reserved for full conviction.
At every management decision point during a live trade, you have exactly three choices: Hold (maintain current size when the thesis is intact), Reduce (cut position by a pre-defined 30-50% when the thesis is wounded but alive), and Exit (close the entire remaining position when the thesis is invalidated or a bail trigger fires).
Reduce when one or two health dimensions (structure, flow, time, volume) have weakened but the core thesis is not yet invalidated. Exit only when structure breaks, multiple dimensions fail simultaneously, or a pre-defined bail trigger has been activated.
Use a pre-defined percentage — typically 30-50% of the remaining position. The exact number matters less than having it decided in advance so you do not deliberate during the trade.
No. Reducing is allocating capital proportionally to your confidence level. If your confidence is at 50%, your position should reflect that. Full positions should be reserved for full conviction.