The tilt cascade and revenge trading

A single loss rarely destroys an account. What destroys accounts is what happens in the forty minutes after a loss. This lesson walks through that sequence step by step, because a cascade you can see coming is one you can interrupt.

The cascade

Here is the sequence in its most common form. The details vary; the shape does not.

Step 1: the trigger loss. Not usually the biggest loss of the week. Often it is a loss that feels unfair: a stop hit by one tick before the move goes your way, a fill that slipped, a trade taken slightly against the plan that "should have worked." The unfairness is the important part. A clean, planned loss produces less arousal than a loss with a story attached.

Step 2: the reference point shifts. From the previous module: you were at zero, now you are at −1R, and −1R has become the new anchor. Every subsequent decision is measured against getting back to zero rather than against the plan.

Step 3: the physiology arrives. Heart rate up, breathing shallow, attention narrowed onto the one instrument that hurt you. The planning circuits are down-ranked. You are now physically a different decision-maker than the one who wrote the plan.

Step 4: the "getting it back" frame. The trade you are about to look for is not a setup. It is a repair. The thought is some version of "I just need one good trade to fix this." Notice that the criteria for "good" have silently changed from "matches my plan" to "makes enough to cover the loss."

Step 5: the setup gets looser. Because you are now looking for a trade rather than waiting for one, you find one. It is a B or C setup, or it is the same market re-entered because "the idea was right."

Step 6: the size goes up. Getting back −1R with a normal-size trade at your normal target takes a full winning trade. Getting it back in one shot requires more size. The math of the repair frame demands oversizing; it is not a separate mistake.

Step 7: the second loss. An oversized trade on a loose setup, taken while physiologically impaired, loses more often than your system's average and loses more when it does. You are now at −3R or −4R.

Step 8: the loop. Return to step 2 with a bigger number. Each iteration has a worse setup, larger size and a more impaired trader. The loop ends when the daily limit is hit (if there is one and it is enforced), when the account cannot support the next size, or when exhaustion sets in.

Key idea: Revenge trading is not one bad decision. It is a loop with a predictable structure: unfair loss, shifted reference point, physiological arousal, repair frame, looser setup, bigger size, bigger loss, repeat. Every step is a place to break it.

The two most dangerous words

"Get back." As in "get it back," "get back to even," "get back to green." The phrase deserves its own section because it marks the exact moment a trader stops trading and starts repairing.

The arithmetic of getting back is worth stating plainly. Your system, if it has an edge, produces something like +0.3R per trade on average. A −1R loss is therefore about three average trades of work. A −4R hole is about thirteen. There is no single trade in your plan that recovers it, because your plan never had a trade designed to recover anything. The moment you want a trade to do a job the plan did not give it, you are outside the plan.

The correct response to "I need to get this back" is: "No, I need to take the next planned trade at planned size, and the equity curve will take care of itself over the next thirteen trades or so." That sentence is boring. Boring is the point.

Variants of the cascade

The stop-hunt narrative. "They ran my stop and reversed." Sometimes true in the sense that stops cluster and get swept. Never true in the sense that it justifies re-entering at a larger size. The narrative gives the loss a villain, and a villain makes the repair frame feel like justice.

Averaging down. Instead of a new trade, the losing position is added to. Same cascade, with the added feature that the position's average price improves, so the loss looks smaller on the screen and the reference point trick is done by the platform itself.

The correlated revenge. Stopped out long on one index future, immediately long a different index future "because it's a cleaner chart." It is the same trade. The Risk Management course covers why correlated positions are one bet; here the point is that the cascade can hide behind a different ticker.

The slow revenge. No dramatic oversized trade. Instead, six or seven slightly-larger, slightly-looser trades over the afternoon, each individually defensible. The total damage is the same as one revenge trade and it is much harder to see in the log.

Breaking the loop

The cascade has eight steps and you cannot reliably act at step 4 or 5 because those are the steps where the planning circuits are impaired. The reliable break points are:

• Before step 1, by having a rule that a loss triggers a mandatory pause (Module 3, time-outs).
• At step 3, by using a physical reset the moment you notice the physiology (pulse, breath, standing up), which you have practised in calm conditions so it is automatic.
• At step 6, by making size a fixed number on the platform that requires deliberate effort to change (Module 3, friction design).
• At step 8, by a hard daily limit that ends the session mechanically.

None of these depend on you making a good decision while tilted. That is what makes them work.

Try it: Find the worst day in your log from the last three months. Reconstruct it trade by trade, with times. Label each trade with the cascade step it corresponds to. Then mark the earliest point at which one of the four break points above would have ended the day, and compute what the day's result would have been if it had. That number, the difference, is what one rule is worth to you.

Recap

• The tilt cascade: an unfair-feeling loss shifts the reference point, arousal impairs planning, the goal becomes repair, setups loosen, size rises, the loss grows, and the loop repeats.
• "Get back" marks the moment a trader stops executing a plan and starts trying to make a trade do a job the plan never gave it.
• Recovering a loss takes several average trades; no single planned trade is designed to recover anything.
• Variants include the stop-hunt narrative, averaging down, correlated revenge and slow revenge, all with the same structure.
• Break the loop at points that do not require good judgement while tilted: a mandatory pause after a loss, a practised physical reset, fixed size with friction, and a hard daily limit.