The Assay Desk tests market stories against price, base rates, and time.
Simple trend systems.
Prediction-market misbelief.
Public notes, not advice.
If a claim cannot survive a ledger, I try not to keep it.
— M. Vale
Research lane, funding RV — week ending 29 Jun. Paper, pre-paper: no signal acted on, no trade in this lane.
Universe of 6. Weekly tape: 0 positive episodes, 1 negative (3 hours). A quiet week — funding paid almost nothing in either direction, and nothing crossed the extreme threshold (0.006%/hr).
Longs paid most on HYPE (+0.00129%/hr avg) and PUMP (+0.00109%/hr). Shorts paid most on SOL (-0.00069%/hr).
OI tape: 39.6 days accrued, 176 coins on the 24h cut. Ready for the OI overlay; still nothing the rule can act on.
The thesis only goes paper-live when extreme funding co-occurs with rising OI often enough to support a hedged pair. This week, the data does not authorize anything.
Kalshi: CPI June 2026 month-over-month > 0.0%, YES side. The desk wanted in. The risk book said no.
Ladder market, BLS anchor, 22 days to resolution. Estimated edge looked large on paper. The live sleeve was already full when the decision came through, so no order placed.
Spread was wide too — even if the sleeve had room, the cost would have eaten into it. The journal gets this one, not the book.
Research lane, funding RV — week ending 22 Jun. Paper, pre-paper: no signal acted on, no trade in this lane.
Universe of 8. Weekly tape: 6 positive episodes (27 hours), 0 negative. One-sided week — longs paid shorts across the board, nobody paid longs.
Longs paid most by: WLD (0.00169%/h), HYPE (0.00159%/h), ZEC (0.00133%/h). ETH and BTC also positive but quieter.
OI overlay tape: 32.6 days accrued, 178 coins on the 24h cut. Close to the window where the lane can ask whether extreme funding co-occurs with rising OI often enough to mean anything. Not there yet.
The data does not yet authorize anything.
Kalshi: MLB, Mets vs Phillies (Jun 20), NO New York leg. The desk flagged it and kept it on watch — no order placed.
Kalshi YES on the Mets: 50¢. Sportsbook consensus across 11 books: Phillies 61.8%, Mets 38.2%. Gap of about 11.8¢ on the NO side.
Interesting enough to mark in the book. Not interesting enough to chase past the desk's daily cap on singleton games.
Kalshi: major Eastern Pacific hurricanes this year, the 6-or-more rung, NO side. The desk wanted in. The risk book said no.
Ladder market, estimated edge ~20%, EV ~62%. Decision cleared the gates; the live sleeve was already full, so the order didn't go.
Tracked without placing.
Kalshi: August 2026 CPI month-over-month > 0.3%, YES side. The desk flagged it and kept it on watch — no order placed.
BLS as the anchor. Per-rung blend put fair around 63%; the book had YES at 52¢. Interesting gap, but the card never reached execution — observed, not traded.
Fourteen months to resolution. Plenty of prints between here and there.
Kalshi result: CPI YoY for May 2026, T4.2 rung, NO side. Resolved YES. Loss.
Entry 60¢. Held ~7.4 days. -100% on the contract.
NO at 60¢ implied the rung wouldn't print — the rule took that side, the print landed in the rung, the contract paid zero. A ladder market with a BLS anchor, and the anchor went the other way.
The journal gets this one.
Kalshi result: CPI YoY for year ending May 2026 > 4.2%, NO side. Resolved NO. Win.
Entry 29¢. Held 18 days. +245% on the contract.
NO at 29¢ implied the market thought ~29% chance YoY CPI prints above 4.2%. BLS settled it under the rung. The anchor held.
Kalshi result: CPI May 2026 > 0.4%, NO side. Resolved YES. Loss.
Entry 16¢. Held ~19 days. -100% on the contract.
NO at 16¢ implied the rule was paying for an 84% chance that May CPI would land at or below 0.4%. It didn't. The anchor was clean (BLS print), the horizon fit, the number went the other way.
The journal gets this one.
Kalshi: Bank of Canada cuts exactly 25bp at the October 2026 meeting, YES side. Priced 50/50. The desk observed and moved on.
Resolution is sixteen months out. No current BoC statement, no Governing Council minutes, no overnight index swap reading the rule could anchor to a meeting that hasn't been scheduled into a forecast cycle yet.
Interesting market. Nothing to mark it against today.
Research lane, funding RV — week ending 8 Jun. Paper, pre-paper: no signal acted on, no trade in this lane.
Universe of 10. Weekly tape: 9 positive episodes (26 hours), 7 negative (55 hours). Longs paid most by HYPE at +0.234 bp/hr across 5 episodes; shorts paid most by ZEC at -0.217 bp/hr across 3.
OI overlay still not ready — 18.6 days accrued, the rule wants more before it calls anything crowded.
The thesis needs extreme funding co-occurring with rising OI often enough to support a hedged pair. Neither leg of that is in evidence yet. The data does not yet say anything.
Kalshi: NL Central division winner, Cubs leg. The desk passed.
Division-winner contracts resolve at season end — months of baseball still to play, and no clean anchor the rule could mark against a singleton like this. Sportsbook futures exist, but the shape doesn't match: a season-long division price is a different animal than a game-line.
Spread was wide, resolution far out, and the rule's read on the leg came in below the book by a wide margin. Nothing here to underwrite.
Journal gets this one.
Kalshi: CPI month-over-month for May 2026, the 0.6% rung, YES side. Small live position, filled.
Ladder market with a BLS anchor. The book had YES at 19¢; the desk's per-rung blend put the rung closer to 25%. ~6¢ of gap on a clean settlement source.
Medium confidence, high-quality source. Filled small.
Kalshi: which name will the first Atlantic hurricane carry this year — single-name leg. The desk passed.
179 days to the window opening, no current storm to anchor to, and no NOAA/NHC reading the rule can cite for a 2026 season distribution. A 26-name list priced as a guessing game is general knowledge, not a settlement anchor.
Spread was wide and depth thin on the leg besides. Nothing to do here.
Kalshi: nominal U.S. GDP growth for Q2 2026 > 1.5%. Priced 50/50. The desk observed and moved on.
BEA won't post the advance estimate for Q2 2026 until late July 2026 — over a year out. No anchor producer to lean on between now and then, and the 50/50 mark is the market admitting the same. Interesting to watch, not to place.
Kalshi result: MLB, Tigers vs Rays, NO Detroit. Resolved YES. Loss.
Entry 59¢. Held ~12 hours. -100% on the contract.
NO at 59¢ implied Detroit around 41% to win — the rule took the underdog side and the underdog lost. Three MLB singletons logged this week: two wins, one loss. The sample is what it is.
Kalshi: core CPI for May 2026 > 0.2%. The desk observed and moved on.
May data isn't out until 10 Jun — BLS only has April on the page (+0.4% core). Every forward number on offer is a forecaster's estimate, not a settlement-grade reading. Nothing to anchor to yet, and the spread was wide on top of that.
Worth a second look once the release prints.
Kalshi: number of tornadoes in Jun 2026, NO side, ladder market. The desk wanted in. The risk book said no.
Estimated edge ~40%, EV ~9%. Decision cleared the gates; the live sleeve was already full, so no order placed.
Noted for the journal, not the book.
Kalshi: MLB, Padres vs Phillies (Jun 2), NO San Diego. Small live position, filled.
Kalshi YES on San Diego: 50¢. Sportsbook consensus across 11 books: San Diego 44.7%.
Gap of 5.3¢ on a singleton game-winner. Anchor is the vig-removed line; resolution is the game itself.
Kalshi result: MLB, Marlins vs Nationals, NO Miami. Resolved NO. Win.
Entry 44¢. Held ~17 hours. +127% on the contract.
NO at 44¢ implied Miami around 56% to win — the rule took the underdog side of a singleton game and the book agreed by the end.