The Track Record
Dated calls, confidence levels, and the evidence that would change my mind. Updated as the world updates me. The ones I got wrong stay on the page.
Anchorage, Fireblocks, the whole tier. Great gross numbers, no moat, no path to profitability. Custody is a commodity. The moat will be in infrastructure rails, not in holding the keys.
What would change my mind: A custodian achieving >30% net margins consistently, or custody becoming a regulated monopoly via legislation.
Open — resolves 2027
Enterprises trust their auditors. Canton, DTCC, tokenization — every serious conversation starts in the advisory relationship, not at a Web3 conference. Crypto-native firms don't have the compliance credibility to close enterprise.
Open — resolves 2028
AI is creating the same pattern I've survived three times. Most coaching targets early-career or C-suite. The whitespace is the 15-to-25-year professional watching their domain get disrupted. That's the audience for Plan for Thursday.
Open — resolves 2028
Three measurable signals — AI displacement, stablecoin interchangeability, and the quantum timeline — are already moving. I publish the data, not the opinion.
Tracked live at /canaries
The miss stays. If the record only showed the hits, it wouldn't be a record.
I was wrong on the bill itself. The deeper thesis — that regulatory momentum on crypto was real — held; stablecoin legislation just moved first.
What would change my mind: The lesson: I was anchoring on one bill when the right anchor was the direction of motion.
WRONG — on the bill; the direction held
I put up the dot-com chart with an arrow right before the crash: we are here. The point was that the survivors come next — Webvan and Amazon again. But crypto never traced that curve. It crashed in waves — the 2018 ICO bust, quiet years, then FTX in 2022. Not one cliff.
What would change my mind: The lesson: cross-cycle analogies are reliable for the shape; the when and how are where humility lives.
WRONG — right shape, wrong curve
