Rejected before testingRejected · long · n/a — no backtest
The idea
One mechanism that passed this lab's entry test is a US rule: a clearing firm that fails to deliver shares must close the position by a fixed deadline. Europe has an apparent twin. Under the EU settlement regulation, CSDR, a participant with a settlement fail on a European central securities depository can be bought in. If that works like the US rule, EU-listed names would get the same forced, price-insensitive buying.
The question also corrected a habit. Every mechanism examined until then had been American, not by a scope decision but by default.
How it was tested
No backtest. The entry test asked two things: whether the buy-in obligation is actually in force, and whether any per-instrument settlement-fails data exists to rebuild history from. Both answers came from one search, without a deep read of a primary document.
Where it died
Rejected on 2026-09-30, on two independent grounds. The obligation is not live: mandatory buy-ins were suspended pending the CSDR Refit and remain suspended. Even the revised text treats them as a last resort, after cash penalties and other measures, which builds discretion into when they trigger. The US rule has a fixed calendar; this one does not.
And there is nothing to test on. Per-instrument settlement-fails disclosure is not scheduled to begin until December 2026, so no history exists to rebuild. Either ground alone would have ended it.
The record
Thesis
The EU analogue of the US Reg SHO close-out: a participant with a settlement fail on an EU central securities depository is bought in, creating forced, price-insensitive buying in the failed security.
Gates
Entry test — obligationfail
Is someone actually obliged to trade badly here, and unable to stop?
Why it died
Two independent grounds. First, the obligation is not in force: mandatory buy-ins were suspended pending the CSDR Refit and remain suspended, and even the revised framework frames them as a last resort, applied "only where other measures ... have not resulted in a long-term sustainable reduction of settlement fails" — discretion built into when it triggers. Second, there is no data: per-instrument settlement-fails disclosure is not scheduled to start until December 2026, so there is nothing to reconstruct history from.
What it taught
Checking for a non-US analogue corrected a default bias — every mechanism examined so far had been US-centric by habit, not by a scope decision. Both findings took one search, no primary-document deep read. Worth reopening only if the buy-in regime is reactivated and the per-instrument fails data actually lands.
Verdict history
2026-09-30Rejected at the entry test — buy-in suspended, no per-instrument fails data Rejected before testing