Week 12: Ruin

The same lecture as week 1 — the same slides, the same two numbers — given to a room that now has a ledger.

24 May 2027rationedhanded to you, never assessedOpen the slides

This is the week 1 lecture. The slides are the same file. Nothing has been added to them and nothing has been taken out, and there is no new mechanism today — the eleven are the eleven.

The bet paid 4.5% a go and ruined 87 people in 100. In February that was an argument about a bet nobody had offered you. Today it is read in a room where the class ledger is on the screen and roughly a third of you finished the semester with more than you started with.

The people at the top of the ledger#

They are the lecture. They did not cheat and they are not fools. They took the same twelve weeks and came out ahead, and if you ask them why, they will tell you something specific and true about what they chose.

Week 1’s arithmetic says what that explanation is worth. Of a hundred people running the same exposure, about thirteen finish holding

1.10^40  =  45.26

times their stake — and every one of those thirteen holds the identical amount, because a survivor is simply someone the loss did not happen to. There is no spread among them. No skill gradient exists in that bet at all, and yet each of the thirteen has a sincere account of their own judgement, and each account is consistent with the facts.

The account is not evidence. Neither is the money.

The harder half#

The symmetrical point costs more to accept, and it is the one this session is actually for.

Your loss is not evidence either. You ran one path. A single path through a process tells you almost nothing about the process — that is not a consolation, it is the same sentence as week 1 with the sign flipped, and it applies to the bottom of the ledger exactly as it applies to the top.

So “I lost, therefore I judged badly” is the same error as “I gained, therefore I judged well”, made by someone in a worse mood. Both take one sample and read a process off it.

What the ten weeks in between bought you is the ability to say something stronger than either.

What a decomposition is#

A narrative says: the market turned against me and I got out too late.

A decomposition says: this much went to transacting, and here are the four round trips and the side of the quote on each. This much is the gap between the position I held and the one I wrote down on the day I chose against it. This much is the recovery my worst holding would have needed, and the eight weeks I spent waiting for it. And this much is left over — which was the market, or was something I cannot name, and it goes in the account as a residue rather than being absorbed into the story.

The residue is the part that makes it honest. A decomposition that explains 100% of your loss is a narrative wearing a table. Most real accounts attribute somewhere between half and two thirds and report the rest as unexplained, and that is a stronger document than a tidy one.

This is the only thing in the course that survives contact with somebody else’s twelve weeks. Two students with opposite outcomes can compare decompositions; they cannot compare stories.

Re-read what you wrote in February#

Take out the opening record you filed in the first session and have not been allowed to revise. Read all four answers in order.

The percentage you chose without being warned, which week 3 turned into a multiplication. The one sentence beginning “I am buying it because”, which week 7 asked you to price. What your holdings had in common, which week 8 costed to the dollar. And the exit condition, which a lot of you left blank, and which week 10 read back to you as a date eight weeks in the past.

You wrote all four before you had any of the vocabulary. That is why the sheet was sealed, and it is why the final project ends by quoting it rather than summarising it.

There is no coda#

The course has one claim. It made it in February, it is making it now, and the reason it is made twice is that the first time it was information and the second time it is a description of people in the room.

Close your positions. Put the closing rows in the log. Then write the account: three mechanisms, traced to positions you actually held, a residue you do not explain away, and a last section that reads February back to yourself.