A rules-based crypto allocation system — and its record, kept in the open.
Public · append-only · pushed the day each call is loggedCrypto, without the 80% wipeout.
It holds the strongest coins while the market is healthy, and steps aside into gold or cash when Bitcoin turns. You are not trying to be a genius. You are trying to still be here in two years.
The system ranks a pool of major crypto assets by relative strength, sizes positions by inverse volatility so one violent alt can't dominate the book, and goes fully defensive when Bitcoin breaks its long-term trend. The rules are the product. The coin pool is disclosed, and fixed in the indicator.
2018: typically +19% (−16% to +40%) while Bitcoin crashed −73% · 2022: typically −4% (−7% to +4%) while Bitcoin crashed −64% · 2025: typically +56% (+32% to +80%) while Bitcoin chopped to −6%
Each range covers every start day of the rebalancing cycle; the indicator on the chart shows one of them.
Vero means true. So here is the whole truth, including the parts that are not flattering.
2021 alone produced about half of everything the backtest made. Any multi-year crypto number deserves the same question: which single year did it come from?
This is the most important sentence on the page. The twelve coins were picked knowing which ones survived. Testing the same rules on a survivorship-free universe — coins as they actually ranked at the time, including the ones that later went to zero — returns materially less. Plan around the rules and the protection, not around the historical multiple. The live record below is the honest test, and it is running now.
Its worst drop was 26–37% depending on the day its rebalancing cycle started — typically 31% — peak to trough, from January 2018 to September 2026, measured on a survivorship-free universe — one that keeps the coins which later went to zero. Bitcoin fell 81.5% over that same window. On the chart, which starts in 2017 on the hand-picked pool, the panel shows the worst drop up to the latest day, recomputed every day. When the core protective rule was run on the far more brutal 2011–2017 era, it still took roughly 70% while Bitcoin took 85%. Real protection, in a genuinely violent market. Not a shield.
About 92% of days sit below the best balance the system has ever reached. The longest flat stretch was typically about 15 months — between 12 and 24 depending on the start day — and typically 1 in 10 one-year starting points ended below where they began (between 1 in 18 and 1 in 7). This is normal for a system like this, and it is exactly where most people quit.
One line in the sand. If a drop of nearly 40%, or two and a half years without a new high, would make you quit, this is not for you. I would rather say that now than lose you in month 14.
Written down in advance and published in this repository before the results it is judged against: what to expect, and what would prove this system isn't working.
Every figure above is computed from the log
and public market prices by scripts/build_summary.py — none of it is written by hand, and
the Bitcoin comparison is shown whether it flatters this record or not. Recompute the whole thing
yourself with python scripts/verify.py.
Credibility doesn't come from a screenshot. It comes from a record that accumulates continuously, in public, in real time, on infrastructure I own — witnessed forming, not posted after the fact. Every call is pushed here the day it is logged, where anyone can see and copy it. Nothing here is back-dated or edited, and a wrong call stays in the record permanently.
| # | Logged | State | Allocation | Result since prior | Made by |
|---|---|---|---|---|---|
| Loading… | |||||
Full history: see the log and this repository's commit history.
Add the indicator to a daily (1D) chart of INDEX:BTCUSD, the only chart it works on, and create one alert on it, set to "Any alert() function call" — its headline says EXIT or ENTRY.
Getting out on time is the edge. Getting in is not urgent. That asymmetry is the single most important thing to know about running this.
Backtest vs live: historical performance shown by the indicator is a backtest and is independently verifiable there; this log is the live forward record from its inception date. The pool is a disclosed input, not a claimed edge — a backtest over a hand-picked list of coins that survived is survivorship-biased and is never presented here as a performance headline. Nothing here is back-dated or edited; a wrong call stays in the record. Your own figures may differ slightly by exchange feed.
Not financial advice. Vero Compass is a research tool, not a promise of returns, and I never touch your money. Crypto is volatile; only risk what you can afford to lose, and do your own diligence.