The Pre-Mortem: How to Plan a Trade by Imagining It Already Failed

A post-mortem analyzes why a project failed after it failed. A pre-mortem does the same thing before the project starts, you imagine it's already failed and work backward to identify why. Applied to trading, the pre-mortem is one of the highest- leverage rituals you can adopt. It pre-emptively activates the same skeptical thinking that confirmation bias would suppress once the trade is on.

The exercise

Before placing a trade, write a short answer to two questions:

"It is one month from now. The trade has failed completely, full -1R loss, possibly worse. What happened?" List the specific reasons the trade could fail: market regime shift, news catalyst, structural breakdown, liquidation cascade, my own behavioral mistake. Be concrete.
"Given that list, what's my response if any of these start to materialize?" For each failure mode, define the early warning signal and the specific action you'll take.

The whole exercise takes 2-3 minutes. You can do it in a journal app, a notes file, or on paper. The act of writing it down is what makes it work, mental rehearsal alone isn't enough because the brain glosses over details that the page forces you to specify.

Why this works

Several mechanisms:

It activates contradictory thinking before the bias takes over. Once you have a position, confirmation bias filters out failure modes. The pre-mortem forces you to identify them while you're still neutral.

It surfaces low-probability tail risks. "What could catastrophically go wrong?" is the question the pre-mortem asks. Answers often include things you wouldn't have explicitly listed in the bullish thesis but obviously matter.

It pre-commits responses to specific failure signals. When the failure signal arrives (price breaks the support you were trading off, on-chain flow flips against you, news breaks unexpectedly), you've already decided what to do. No real-time decision needed under pressure.

It improves trade selection. The act of writing out "how this could fail" surfaces trades that, on inspection, have too many possible failure modes. Many trades you would have taken on impulse get filtered out at this stage, saving R you would have lost.

It creates a record for future learning. When the trade plays out (success or failure), you can compare reality against the pre-mortem. Were the failure modes you identified the actual ones that mattered? Were there failure modes you missed? This is the highest-quality data for improving your process.

A worked example

You're considering a long BTC perp at $66,000, stop $65,000, target $70,000. Pre-mortem:

"It's one month from now. The trade failed. What happened?"

BTC broke down through the $65,000 support I was using as stop reference. The break wasn't a wick, it was a clean body close below.
A macro event (FOMC, regulatory action, ETF outflow) drove a sharp risk-off move that took crypto with it.
I ignored a divergence in exchange flows that was developing. They turned net-inflow-bearish two days before the breakdown.
I sized larger than my normal 1R because "this setup looks great." The loss was actually -2R because of the size-up.
I held past the stop because I was sure the support would hold "this time." It didn't.

"What's my response if any of these start to materialize?"

Clean break below $65,000 → exit at the close, no second- guessing. Bracket order with stop is reduce-only.
Macro event during the trade → size down or close preemptively if event is scheduled and material (FOMC, CPI). For unscheduled news, react to actual price action, not the headline.
Exchange flows flipping bearish → reduce size by 50% if it persists 48 hours.
Sizing: 1% account risk, no exceptions. Position sized from stop distance, not from "conviction."
Stops: pre-committed via bracket order. Not movable manually under any circumstances.

That whole exercise took 5 minutes. The trade is now equipped with structural defenses against the most likely failure modes, and you've already pre-decided your responses.

Common failure modes the pre-mortem typically surfaces

Across many trades, the same categories of failure show up:

Market regime shift. Your setup was based on conditions that changed. The trade was theoretically right for last week's market, wrong for this week's.

News catalyst. A scheduled event (FOMC, CPI, earnings, upgrade) or unscheduled news (regulatory, hack, exchange event) overrides your TA and on-chain context.

Structural break. The level you were trading off (S/R, trendline, MA) failed. The chart structure broke.

Liquidity event. A liquidation cascade or flash crash moved price through your stop on the way to a level it wouldn't have reached in normal trading.

Behavioral failure. You moved your stop, sized larger than planned, or held past invalidation. The market did what it should have done; you didn't follow your own plan.

Time-based. The trade took longer than expected to play out. You ran out of patience or the funding cost ate the profit.

A pre-mortem should hit at least 3-5 of these categories per trade. If you can only think of 1-2 ways the trade could fail, you haven't thought hard enough.

A common mistake: skipping the pre-mortem on "obvious" trades

The trades that feel most obvious are exactly the ones you should pre-mortem most rigorously. "Obvious" usually means "my brain has already filtered out the failure modes." The pre-mortem is most valuable when it surfaces things your brain wanted to dismiss.

Test: if you wouldn't write a pre-mortem for this trade because "it's obviously going to work," that's a flag. Write the pre-mortem anyway.

A common mistake: doing the pre-mortem half-heartedly

A vague pre-mortem ("the trade could fail if the market drops") gives you nothing. The value comes from specifics, specific levels, specific catalysts, specific signals, specific responses. If your pre-mortem could apply to any trade in any market, it's not actually doing the work.

Discipline: every pre-mortem item should be a concrete condition with a concrete response. "Price closes below $65,000" is concrete. "If things look bad" is not.

A common mistake: treating the pre-mortem as a forecast

The pre-mortem isn't predicting that the trade will fail. It's identifying the conditions under which it would, so you can monitor for them. Most of your trades won't trigger the failure modes you list. The point isn't that they will, the point is that if they do, you've already decided how to respond, and you've spent the upfront time identifying which signals matter.

A common mistake: not reviewing pre-mortems against actual outcomes

You wrote a pre-mortem before the trade. The trade played out. You should now go back and check: which of your predicted failure modes (if any) showed up? Were there modes you missed? Were any of your responses wrong?

This review is where most of the learning happens. Without it, the pre-mortem is a one-shot defensive tool. With it, it's a self-improving feedback loop that makes your future pre- mortems sharper.

Mental model, the pre-mortem as the project lawyer's first meeting

Before any major project, a competent organization runs the plan past their lawyers, who specifically look for what could go wrong: legal exposure, regulatory issues, contractual gaps. The lawyer's job is to be paranoid on the organization's behalf, before the project starts, when fixes are cheap.

Your pre-mortem is hiring yourself as the trade's lawyer for ten minutes before entry. The lawyer's specific brief: find the things that could go wrong, identify the early signals, and pre-write the response. The trade then proceeds with the lawyer's checklist active, ready to react when any signal comes up.

After entry, you're back to being the trader, not the lawyer. But the lawyer's work is done, the responses are pre- specified, the failure modes are catalogued. Confirmation bias has less room to operate because you've already considered the cases against the trade in writing.

Why this matters for trading

The pre-mortem is one of the few rituals that actively neutralizes multiple cognitive biases at once: confirmation bias (you're looking for failure modes), overconfidence (you explicitly imagine the trade failing), sunk cost (you pre- commit to exit conditions), recency bias (you consider regime shifts that haven't happened recently). Hex37's journal page supports per-trade notes, use the entry note specifically for the pre-mortem text. After 50 trades, you'll see the patterns of what really kills your trades.

Takeaway

The pre-mortem imagines the trade has failed and works backward to identify why. It surfaces failure modes confirmation bias would suppress, pre-commits responses to specific signals, improves trade selection, and creates learning data. Specifics matter, concrete conditions with concrete responses, not vague risks with vague reactions. Five minutes per trade, every trade. Reviewed against actual outcomes, the practice compounds into structurally better trade selection and execution.