HuginaiI
Directional confluence book
Reason: structurally negative; the conviction score measured anti-predictive
Closed
A crypto research desk · paper only
Every strategy this desk has run, including the ones that died.
One person’s research project. Every book trades on paper against live prices, gets measured, and closes when it does not pay. Two of the three books are closed; the wall below keeps them, cohort by cohort.
Live · the one book still running
Hyperliquid funding carry, on paper. Each trace is one coin’s funding, hour by hour, over the last 30 days.
Paper book
—
HYPE · NEAR · LDO
Held
—days
run 2 since 2026-10-17
Earned so far
—
after the fees and slippage paid so far
Reading the desk’s hourly funding record…
hourly funding, bps an hour
weekly funding floor: 7 bps over 7 days, drawn as 7 bps ÷ 168 hours
zero
Source: the desk’s carry tracker, one record an hour, read live from the desk. Gaps are hours with no record. A coin whose last 7 days pay under 7 bps is closed by the tracker. Paper positions on a single operator’s own account, never real capital.
Waiting for the desk
The memory wall
One line per book, with the number that closed it.
Odin keeps two ravens: Hugin for thought, Munin for memory. The sagas have him fear more for Munin. This is the memory.
2books closed
1book running · paper
A book is one strategy traded on its own paper account. The directional book ran as 3 cohorts before it closed; the same count is on the desk’s memory wall.
I·1
Cohort I · local run
Reason: no measurable edge
Closed
I·2
Cohort II · production run
Reason: the +1.2% headline was a daemon outage
Closed
−$33.60net once the artifact was removed
I·3
Cohort III · auto-main-v3.1
Reason: structurally negative payoff. 40% win rate against the 57.4% it needed.
Closed
−2.31%of NAV at close
II
Binance funding-arb
Reason: the spread does not clear the fees, structurally
Closed
III
Hyperliquid funding-floor carry
Run 2 on paper since 2026-10-17, hourly on the desk.
Running · paper
3.9–5.6%net APR over 60 days
The Munin margin
What the desk believed at open, against what happened.
The book stays closed absent new structural evidence. A parameter set that backtests better does not count as evidence — that is the mistake the +1.2% headline came from.
Number
Believed at open
What happened
01
Believed at openFusing social, Telegram and on-chain signals finds real setups
What happenedNo positive slice at n≥5 — by side, source, ticker or hour of entry
02
Believed at openA 1–10 conviction score separates good entries from bad
What happenedConviction measured anti-predictive in all three cohorts
03
Believed at openTight ROI targets bank winners early and protect the book
What happenedThe ladder cut winners at +0.7–1% while stops released losers at 3–6.4%
04
Believed at openThe losses are fees and slippage; net them out and the edge appears
What happenedA zero-fee, zero-slippage replay still lost about $150
05
Believed at openOne cohort printed +1.2%, so the approach works and needs tuning
What happenedThat +1.2% was two positions held 131.6 hours through an outage, into an SOL rally
Still running
One book survived, and it needs no directional view.
Short the Hyperliquid perpetual, hold the spot leg, collect the funding difference. Delta-neutral, so the part that failed — predicting direction — is not involved.
HYPE5.6%
NEAR5.2%
LDO3.9%
Window 2026-05-10 → 2026-07-08 · re-costed 2026-08-13rendered worn because that is how old the measurement is
Cost basis: taker fees plus 5 basis points of assumed slippage on every leg, every side, amortized over a 30-day hold. The one order-book walk in our data measured 0.5–2.9bps at this clip size on a calm day, so 5bps sits above what we have seen — and the Binance spot legs have never been measured at all. These figures were 6.4–7.7% before slippage entered the model; that earlier version was an upper bound and is what this page showed until 2026-08-13. LDO at 3.9% now sits below the 4.5% net APR the desk requires before a position is considered for capital, so on its own expectation it no longer qualifies. Paper positions on a single operator’s own account, never real capital.
How the desk works
Build it, measure it, close it when it does not pay.
The pipeline runs and the ingestion works. What it produced is what closed the book.
01
Ingest
runsPublic Crypto Twitter, Telegram channels and on-chain flows land in one stream. This part still runs.
02
Cluster
gates nothingRelated mentions are grouped and de-duplicated, sources weighted, each cluster scored with its sources kept. The score measured anti-predictive, so it gates nothing now.
03
Paper-test
paper onlyEverything trades on paper against live prices, with taker fees and slippage modelled. No real capital, ever.
04
Judge
3 closedA cohort clears its breakeven win rate or it closes. Three did not. A fourth result that looked good turned out to be an outage, and was thrown out too.
Access
There is nothing to buy.
One person’s research project. No plans, no accounts, no checkout — the pricing page that used to sit here described a purchase path that was never built.
The dashboard
The desk interface. The carry paper test reads live from the tracker: run 2, started 2026-10-17, one record an hour, paper positions only. The pages of the closed directional book are an archive on labelled sample data. No account, and nothing here places an order.
Open the deskThe write-ups
How the directional book was measured and why it closed, and how the funding-carry forward-test is run. Including the results that did not work.
Read the write-upsNo capital is managed for anyone and no real capital has ever been deployed. Every strategy described here was traded on paper. Nothing on this site is financial advice.