Drawdowns Are Information If You Keep a Journal

Every investor meets drawdowns. Few extract clean information from them. Memory edits the story toward heroism or victimhood. A journal resists that edit. Drawdowns become information when you record what you believed, what you sized, what you changed, and what you refused to change while the path was ugly.

What to write while it hurts

Write the thesis in one paragraph. Write the invalidation. Write the size rationale. Write the emotional temperature in one honest sentence. Write any process exceptions. Do it during the episode, not only after recovery. After-the-fact journals are speeches. Contemporaneous journals are data.

I am not asking for literary diaries. Short, dated notes beat eloquent reconstructions.

Separate market pain from process pain

Some losses are the cost of a risk you accepted. Some losses are the cost of a risk you took by accident—unclear mandate, drifted size, skipped checklist, crowded factor you pretended was unique. Journals help separate tuition you agreed to pay from tuition you should refuse next time.

Recovery can teach the wrong lesson

If a broken process is rescued by a market rebound, teams learn that process is optional. Journals make that false lesson harder to sustain because the breakage remains on record even if P&L healed.

Systematic tools still need human notes

Automated systems produce logs. Humans still need interpretive notes: why an override happened, whether risk envelopes felt wrong, whether data quality degraded. AI decision support can summarise patterns later. It cannot replace the honesty of a note written on a bad Tuesday.

For allocators, drawdown journals across sleeves reveal whether “diversifiers” actually diversified. That question is hard to answer from memory alone.

A habit worth keeping boring

Templates that reduce friction

Use a fixed template: date, market context in two sentences, positions involved, thesis status, size changes, emotions in one line, process adherence score, next review date. Friction kills journaling. Templates keep it alive on bad days when free writing feels impossible.

Store journals where they can be searched later. Paper is fine if you actually keep it. Digital is fine if you do not hide notes in scattered chats. The medium matters less than retrieval.

Team journals versus personal journals

Desks should keep a shared decision log for mandate-level choices. Individuals may keep private notes for behavioural patterns. Do not force performative vulnerability in public channels; do require factual process logging in shared ones.

After recovery, schedule a reading of the ugly pages. Ask what you would repeat. Ask what you romanticised. Convert one insight into a checklist or limit change. Then move on. Journals are for learning, not for permanent self-prosecution.

Information wants a record. Drawdowns will provide the events. You provide the pen.

Keep the journal format stable so you can compare episodes. Review quarterly. Extract one process change maximum per review. Too many changes after pain is another form of panic.

Privacy and honesty trade-offs

Some behavioural notes should stay private so honesty survives. Some process facts must be shared so organisations learn. Draw the line explicitly. Ambiguity creates either performative journals or secret ones that never improve the desk.

Over a career, journals become a personal research archive on your own decision quality. That archive is more valuable than most tip streams. Keep writing.

Using journals in research and AI loops

Structured journals can later train better decision-support prompts and postmortems—if you consent and anonymise appropriately. The point is not surveillance. The point is turning expensive experience into reusable pattern libraries: sizing errors, override clusters, and repeated invalidation blindness.

Do not wait for a perfect software stack. A consistent text template beats a fancy system you abandon. Start now; migrate later. Information compounds only if it exists.

Drawdowns will visit. Meet them with ink, not only with hope.

Using journals in research and AI loops

Structured journals can later train better decision-support prompts and postmortems—if you consent and anonymise appropriately. The point is not surveillance. The point is turning expensive experience into reusable pattern libraries: sizing errors, override clusters, and repeated invalidation blindness.

Do not wait for a perfect software stack. A consistent text template beats a fancy system you abandon. Start now; migrate later. Information compounds only if it exists.

Drawdowns will visit. Meet them with ink, not only with hope.

Drawdowns will still hurt. With a journal, they can also teach. Without one, they mostly haunt.