Every hypothesis gets tested until it breaks or proves itself. Both results get published.
PAPER · SSRN
Six Ways Our Backtest Lied to Us — and None of Them Showed in the Results
A new paper with Zivana Zerjal: six implementation defects caught by replaying live days through the backtest — a thin data feed, adjusted prices, a runaway sizing step, a UTC clock, missing inputs, dead config. None showed in the results. Each now has a test.
Latest writing
INFRASTRUCTURE · OBSIDIAN + AGENTS
Obsidian + Claude Code: I Gave an AI My Notes and It Argued for an Idea I'd Killed
Research note
It read my vault and made a well-sourced case for a thesis I'd killed two months earlier. I nearly acted on it. The fix was a handful of documents — 2% of the words — that tell any reader how to read the rest.
MACRO · JAPAN
The Yen Carry Trade Explained: Why One Small Rate Hike in Tokyo Can Crash US Stocks
Research note
In 2024 a crowded yen trade turned a small Tokyo hike and one weak US jobs report into the Nikkei's worst day since 1987. It unwound in two weeks, then got rebuilt. How the trade works, why the exit is violent, and what the September 2026 Fed and BoJ meetings did to it.
PAPER · SSRN
New Paper: The Protocol That Cuts Backtest False Positives from 37% to 0.4%
Research note
The validation machinery behind every note on this site, written up properly: a failure-mode taxonomy, a Monte Carlo of both error types, and a new pre-registrable criterion. Free on SSRN, simulation code included.
THESIS · COMPLIANCE AS CODE
Regulation Is Code. You're Running It on Humans.
Research note
A compliance sign-off tells you the system passed on the day it was checked. Nobody measures what it does the rest of the year.
REJECTED · STOP-LOSSES
Monthly Stop-Losses: Every Threshold We Tested Lost
Research note
The worst months and the best months are usually the same month. Every threshold we tested took the loss and then sat out the recovery.
INFRASTRUCTURE · DATA STACK
The Menu Has Twenty Tools. We Run Five.
Research note
A one-person lab doesn't need twenty tools. Here are the five we kept, and why the license file mattered as much as the benchmarks.
Subscribe
New notes, straight to your inbox.
New notes as they publish — the failed experiments included. Delivered via Substack, no spam, unsubscribe anytime.
Or read everything free at alexvidovich.substack.com · Questions? support@alexvidovich.com