The ruin account

Twelve weeks, three mechanisms, and a last section that re-reads what you wrote in February without letting you revise it.

Due: 28 May 2027 · Weight: 60%

The brief#

Account for your semester as a decomposition rather than a narrative: three named mechanisms, the positions they took money from, and what is left over.

“What is left over” is the interesting phrase. Total your losses, attribute what you can to named mechanisms, and the residue — the part that was the market, or the part you cannot explain — gets reported as a residue rather than absorbed into a story.

Three distinct mechanisms#

Three, from the eleven the course named, each traced to a position your log can evidence.

Distinct means distinct. Concentration and correlation collapse are both about what happens when one thing fails, and a report that treats them as two labels for the same event has supplied one mechanism, not two. If you cannot say what would have had to be different for mechanism B to have not applied while A still did, they are the same mechanism.

Six of the eleven can be committed deliberately and are therefore always available: cost drag, concentration, return chasing, story, overfitting, loss aversion. The other five have to be handed to you. If the semester handed you one, use it and say so; if it did not, say that too. A report that claims a rationed mechanism it cannot evidence is the one failure the marking does not forgive, because the entire premise is that the numbers happened to the person who wrote them.

The closing section#

The last part of the account re-reads your opening record — all four answers, quoted verbatim, from the sheet you filed in the first session and were not permitted to revise.

For each answer, two sentences: what you knew when you wrote it, and which week of the course took it apart. Then one paragraph on the week 1 lecture, which you have now heard twice, read against your own ledger position and the class’s.

The question that paragraph answers is not “was the course right”. It is narrower: what would have to be true about your semester for you to be able to tell the difference between having judged well and having been one of the thirteen? Most accounts conclude that nothing in twelve weeks could establish it. That is an acceptable conclusion and the arithmetic supports it. What is not acceptable is not noticing that the question applies to you.

What you submit#

A single PDF, 2000 to 3000 words, plus the complete position log for the semester and your opening record as an appendix. If you photographed the record in week 1, the photograph is fine.

How it is read#

Not on the outcome. A student who finished up 40% and can account for three mechanisms out of their own log does well; a student who lost everything and cannot evidence why does not. The class ledger is read out in the last session and then plays no part in the marking, which is consistent with what the course spent twelve weeks arguing: the number at the end is one sample, and the decomposition is the only part that transfers.