Kalshi Economics — Mispricing Audit

Report date: 2026-06-17 · Category: Economics · Horizon: markets closing ≤ 45 days · Capital: $1,000

Bottom line. Economics is the most efficiently-priced category on Kalshi right now, and this cycle is no exception. The dominant macro story — a steep, steady decline in retail gasoline — has already done its work: over the past 72 hours the CPI and gas ladders repriced hard to reflect the resulting June headline-CPI deflation, eating most of the obvious edge. After screening 1,035 in-horizon markets down to 31 liquid candidates and diligencing every cluster, I found one defensible residual edge: the market appears to have over-extrapolated the linear AAA price decline into late June, when seasonal summer demand typically flattens it. I deploy ~54% of capital across two expressions of that single thesis and hold the rest in reserve. Everything else is rejected, loudly, below.

Concentration warning up front: both final picks share one underlying driver (the path of AAA retail gasoline). They are positively correlated — if gas keeps falling at its established pace, both lose. Size accordingly; this is not a diversified book.

1 · How this was researched

Universe. Pulled every active market whose event category = "Economics" closing between now and 2026-08-01 by joining trading_markets → trading_events on the read-only Kalshi mirror: 1,035 markets across 110 series.
Liquidity screen. Took the latest market_snapshots row per ticker and cut anything with 24h volume < 1,500, bid–ask spread > 5¢, or already fully priced (YES bid ≥ 95¢ / ask ≤ 5¢). That left 31 candidates, clustered in CPI, gasoline, payrolls, jobless claims, the Fed decision/dot-plot/dissent set, and a few singletons (Musk net worth, 10Y yield, SOFR).
Structural checks. Computed leg-sum arbitrage on every mutually-exclusive event and monotonicity on every "> X" ladder. No free arbs — the only bid-sum > 100 was a 2¢ blip on a same-day-resolving dissent market (fees eat it). The two parallel CPI-YoY series (KXCPIYOY cumulative vs KXECONSTATCPIYOY exact-bucket) reconcile to within 1–2¢. Ladders are internally efficient.
Diligence per candidate. Read the resolution rules verbatim (several titles are looser than the rules), pulled 14-day intraday history to spot 48h repricings, hit the live Kalshi orderbook for depth, and grounded priors in primary sources: AAA national gas average, the Cleveland Fed Inflation Nowcast methodology, the June FOMC, and the Bloomberg Billionaires Index.
Rejection bar. Drop anything with ambiguous rules, a thesis already moved in the last 48h, a price I can't establish from public info, or > 3¢ slippage to size.

Prices below are live as of the 2026-06-17 read. "Implied" = mid of YES bid/ask. The mirror has no live book, so all order-book figures are from the Kalshi public API at audit time.

2 · Markets reviewed

The liquid shortlist and the call on each. Two kept; the rest rejected (detail in §5).

Ticker	Market	YES bid/ask	Vol 24h	Verdict
KXAAAGASM-26JUN30-3.80	Gas > $3.80 on Jun 30	55 / 57	2,522	BUY YES — Pick 1
KXAAAGASW-26JUN22-3.940	Gas > $3.94 on Jun 22	32 / 35	9,170	BUY YES — Pick 2
KXAAAGASW-26JUN22-3.900	Gas > $3.90 on Jun 22	82 / 90	2,005	WATCH — 8¢ spread
KXAAAGASM-26JUN30-3.85	Gas > $3.85 on Jun 30	27 / 31	1,490	WATCH — thin book
KXCPI-26JUN-T0.0	June CPI MoM > 0.0%	10 / 13	4,533	PASS — efficient
KXCPI-26JUN-T-0.1	June CPI MoM > -0.1%	31 / 34	14,597	PASS — moved 48h
KXCPIYOY-26JUN-T3.9	June CPI YoY > 3.9%	11 / 13	5,119	PASS — efficient
KXECONSTATCPIYOY-26JUN-T3.9	June CPI YoY exactly 3.9%	17 / 18	12,551	PASS — consistent
KXPAYROLLS-26JUN-T125000	June payrolls > 125k	37 / 38	4,729	PASS — no edge
KXJOBLESSCLAIMS-26JUN18-225000	Initial claims ≥ 225k (wk Jun 13)	53 / 54	2,176	PASS — noise
KXFEDDECISION-26JUL-H0	Fed holds in July	88 / 90	103,228	PASS — reprices today
KXFEDDECISION-26JUL-C25	Fed cuts 25bps in July	5 / 7	123,079	PASS — event risk
KXFOMCDISSENTCOUNT-26JUN-0	0 dissents at June FOMC	65 / 66	8,754	PASS — resolves today
KXDOTPLOT-26JUN-3.5	June median dot > 3.5%	86 / 90	1,617	PASS — resolves today
KXMUSKNW-26JUN30-T1400	Musk net worth > $1.4T on Jun 30	34 / 35	6,378	PASS — just spiked
KXUST10-26JUN30-T4.65	10Y yield > 4.65% before month-end	14 / 15	1,626	PASS — ambiguous rule
KXCREDITC-26JUL01-T3.75	Q2 quarter-end SOFR > 3.75%	16 / 17	2,949	PASS — data-dependent

3 · Top picks

Ordered by conviction (defensibility of the edge), highest first. Conviction is judged off source quality + rule clarity + book depth — not raw EV.

Pick 1 — KXAAAGASM-26JUN30-3.80 · BUY YES @ 0.57 · Medium conviction

"US average regular gas price strictly > $3.80 on Jun 30, 2026 per AAA" · closes 2026-06-30 · YES bid/ask 55/57 (last 57, prev 63) · OI 5,765

Mispricing thesis. The AAA national average is $4.025 today and falling at a steady ~1.8¢/day. Extrapolated linearly, that lands at $3.79 on June 30 — fractionally below the strike. But gasoline declines almost never stay linear for a month: they flatten as they approach a floor, and late June sits inside the Memorial-Day-to-July-4th driving-demand peak that seasonally supports pump prices. You only need the decline to slow by ~1–2¢ over 13 days for this to resolve YES. The market is already at 57% (it is not pricing the naïve $3.79 linear path, which would imply ~45%), and the strike bounced intraday from a 51¢ low back to 56–57¢ — a tell that the prior session over-extrapolated. I put fair value at ~63–67% (central 64%).

Evidence.

AAA Fuel Prices (fetched 2026-06-17): national regular average $4.025 today, $4.151 a week ago, $4.513 a month ago — a clean, decelerating-in-absolute-terms but steady ~13¢/week glide.
Rules are unambiguous: a single-day AAA national average reading on the resolution date, "strictly greater than $3.80." No averaging, no proxy, no discretion.
14-day mirror history: this strike traded 79–87¢ through June 13, then sold off to a 51¢ low on Jun 16 as gas kept dropping — and recovered to 56–57¢ on Jun 17. The deflation move is in the price; the recovery is the residual edge.
The companion weekly market (gas on Jun 22) prices ~$3.93–3.94, fully consistent with the AAA glide path — i.e. the curve is coherent and I'm only fading the tail of the extrapolation, not the trend.

Tail risk (the single cleanest way this loses). The decline simply continues at its established linear pace — it has shown no deceleration across four weeks — and gas prints ~$3.79 on June 30, resolving NO. The most likely trigger is a further leg lower in crude (e.g. a Middle-East de-escalation pulling the risk premium out of oil) dragging the pump down faster than seasonal demand can offset.

Entry / limit0.57

Implied prob56%

Fair prob~64%

Edge+7¢

EV / contract+7¢

EV %~12%

Liquidity / book: live API shows ~846 contracts offered at 57¢ and ~6,586 at 58¢ — a 660-contract fill costs ≤ 1¢ slippage. 2¢ touch spread. Comfortable for this book.

Pick 2 — KXAAAGASW-26JUN22-3.940 · BUY YES @ 0.35 · Low conviction

"US average regular gas price strictly > $3.94 on Jun 22, 2026 per AAA" · closes 2026-06-22 · YES bid/ask 32/35 (last 38) · OI 10,947

Mispricing thesis. Same driver, shorter fuse. From $4.025 today, even the full linear decline only reaches ~$3.935 by June 22 — sitting right on the $3.94 strike. Over a 6-day window there's little room for the trend to carry it far below, and any seasonal deceleration tips the single-day reading above $3.94. The market prices this at just 33–35%, i.e. it expects gas a touch below the linear path (~$3.92). I put fair value at ~46%. This is a deliberately higher-variance, smaller expression of the same view: a knife-edge strike where I'm paid 35¢ for what I think is closer to a coin flip.

Evidence.

Same AAA primary data and identical, clean single-day resolution rule as Pick 1.
Linear projection math: $4.025 − (1.8¢/day × 5 days) = $3.935, which rounds against the strike but is one slow session away from clearing it.
Very liquid (24h volume ~9,170; OI ~10,947; 3¢ spread) — easy to size and easy to exit if gas gaps down before the 22nd.

Tail risk. A linear-or-faster decline puts the June 22 reading at $3.93 or below and it resolves NO on a single day's print — there's no time for a recovery. Because the strike is essentially at-the-money on the central path, this is genuinely two-sided; hence low conviction and a smaller allocation.

Entry / limit0.35

Implied prob34%

Fair prob~46%

Edge+11¢

EV / contract+11¢

EV %~31%

Liquidity / book: deepest market in the gas complex; a 480-contract fill is trivial. Rest the order at 33–34¢ for price improvement before lifting the 35¢ offer.

4 · Recommended $1,000 portfolio

Conservative deployment given the single-thesis concentration and modest conviction: ~54% at risk, ~46% held in reserve.

Pick	Side	Limit	Contracts	Cost	Max payout	Fair P	EV $	EV %
KXAAAGASM-26JUN30-3.80	YES	0.57	660	$376.20	$660	64%	+$46.20	12%
KXAAAGASW-26JUN22-3.940	YES	0.35	480	$168.00	$480	46%	+$52.80	31%
Deployed				$544.20	$1,140	—	+$99.00	~18%
Cash reserve (opportunistic adds)				$455.80	—	—	—	—

Risk profile

Worst case: gas continues its linear decline; both picks resolve NO. Loss = full $544.20 deployed (54.4% of capital). Reserve is untouched.
Best case: gas decelerates / firms into July 4; both resolve YES. Payout $1,140 on $544.20 → +$595.80 (+109% on deployed, +59.6% on the full book).
Most likely outcome: a split — Pick 1 (the $3.80 strike, with a real safety margin) resolves YES while the at-the-money Pick 2 is closer to a coin flip. A YES/NO split returns $660 on $544.20 (+$115.80); both-YES and both-NO are the tails.
Concentration: 100% of deployed capital rides one driver — the path of AAA retail gasoline over the next 13 days. These are not independent bets. The reserve and the deliberately light deployment are the hedge.

Execution notes

Limit discipline. Pick 1: lift the 57¢ offer (deep book) or rest at 56¢. Pick 2: rest at 33–34¢ first, lift 35¢ only if gas firms. Never pay above 58¢ / 36¢ respectively — past that the edge is gone.
Watchlist triggers. If AAA prints a flat or up day before Jun 22, add the $3.85-on-Jun-30 strike (KXAAAGASM-26JUN30-3.85, ~28¢) from the reserve — that's the highest-convexity expression of deceleration. If the weekly $3.90 strike's spread tightens below 5¢, it becomes a high-probability anchor add.
What invalidates the thesis. Two consecutive AAA days of ≥ 2¢ declines (acceleration, not deceleration), or a sharp crude sell-off on Middle-East de-escalation headlines. Either one → cut both positions; do not average down.
Hedge. The natural hedge is the NO side of the same complex, but that just flattens the thesis. The honest hedge here is position size — keep the reserve in cash, not in a correlated offset.

5 · What I rejected and why

The screen earns its keep here. Most of the category is correctly priced; a few candidates look juicy until you read the rules or the tape.

Ticker / cluster	Why it looked interesting	Why I passed
KXCPI-26JUN MoM ladder	Implies a ~75% chance of negative June headline CPI — looks shocking until you check it.	It's correct, not mispriced. June gasoline is running far below May ($4.03 vs ~$4.51), so a negative headline MoM is well-founded. And the ladder repriced violently in the last 48h (the >0.0% strike fell 89→11¢) — the thesis is fully in the price. Efficient.
KXCPIYOY vs KXECONSTATCPIYOY	Two parallel June-CPI-YoY series — classic cross-market arb hunting ground.	They reconcile to within 1–2¢ once you map the cumulative "> X" ladder onto the exact-bucket ladder. No arb. (Watch the rounding conventions if you ever do trade these.)
KXCPICOREYOY-26JUN	Core ex-energy strips out the gas noise; sticky core could beat the dovish read.	Liquidity too thin (24h volume ~200 on the live strikes, wide one-sided quotes). Can't size without paying the spread.
KXPAYROLLS-26JUN	The whole ladder softened (median drifted to ~$95–100k) — looks like a labor-weakness lean.	No primary-source nowcast gives me an edge over a liquid consensus, and the softening is already priced. Pure noise-trading from here.
KXJOBLESSCLAIMS-26JUN18	Near-term (releases Jun 18), 53% at the 225k line.	A single weekly claims print is statistical noise; the market sits right at consensus with no public information that moves it. No edge.
KXFEDDECISION-26JUL (H0 / C25)	July hold at 88–90%; a soft-labor + falling-gas backdrop could lift cut odds above 6%.	Reprices the moment today's June decision + Warsh dot plot land, and the reporting points to a hawkish tilt (erasing the 2026 cut, hike risk discussed). Trading it now is pure event risk. Pass.
KXFOMCDISSENTCOUNT / KXDOTPLOT (June)	The June dot-plot and a possibly-contested vote (one source reports an 8–4 hold) could surprise the 0-dissent line at 65%.	Resolves today. I won't establish an edge on a same-day event off a single unverified secondary source, hours before the announcement. Out of scope by construction.
KXMUSKNW-26JUN30-T1400	Implied prob of Musk net worth > $1.4T jumped 6→34% in three days — momentum.	That spike is just the repricing of SpaceX's June IPO (net worth ~$1.32T per the Bloomberg index). It's equity-driven, just moved, and would require chasing a 3-day pop on a number I can't forecast better than the tape. Pass.
KXUST10-26JUN30-T4.65	Gapped from ~28% to ~14% on a yield drop — looks like a clean repricing to fade.	The rule — "10Y yield before month-end above 4.65%" — is ambiguous on whether it's a touch (any day) or a month-end value. A definitional dispute could flip resolution. Reject on rules, per the screen.
KXCREDITC-26JUL01-T3.75	Quarter-end (Jun 30) SOFR can spike on repo pressure; 16% might underprice a year-end-style jump.	Base SOFR is ~3.58–3.66% (fed funds 3.50–3.75%); clearing 3.75% needs a 10–17bp quarter-end spike — real but uncertain, and the market is only modestly liquid. Too data-dependent to defend a number. Watchlist, not a pick.
KXEFFR / KXEFFTARIFF	Effective-funds and effective-tariff-rate ladders — interesting macro reads.	24h volume = 0 on every leg. Untradeable. (Useful only as a cross-check: they confirm the 3.50–3.75% funds target.)
KXAAAGASW-26JUN22-3.900	High-probability anchor (gas needs a >12.5¢ drop in 5 days to fail) at the same gas thesis.	The book widened to an 8¢ spread (82/90); at the 90¢ ask the edge is gone. Demoted to watchlist — add only if the spread tightens.

None of the 27 contracts in this run's already-published pick list re-appear above as numbered picks. The two gas strikes I feature (the $3.80 monthly and $3.94 weekly) are not on that list — the previously-published gas pick was the $4.40 June-30 NO, a different strike and direction.

6 · Sources

AAA Fuel Prices — national & state gas price averages (current $4.025; week-ago $4.151; month-ago $4.513, fetched 2026-06-17). Primary input for both picks.
Federal Reserve Bank of Cleveland — Inflation Nowcasting (methodology; the live nowcast page returned HTTP 403 at audit time, so the CPI read is grounded in the AAA gasoline trajectory and the market's own repricing rather than a quoted nowcast value).
REX Shares — FOMC June 2026 preview and Seeking Alpha — June Fed meeting, dot plot in focus, Fed holds steady (target range held at 3.50–3.75%, first meeting under Chair Kevin Warsh).
StockTitan — Fed rate decision, June 17, 2026 and TechTimes — June 2026 Fed meeting, dot-plot/hike-risk context.
Bloomberg Billionaires Index — Elon Musk (net worth ~$1.32T as of 2026-06-17, post-SpaceX-IPO) — used to reject KXMUSKNW.
Kalshi market mirror (read-only DB: trading_markets, trading_events, market_snapshots) for the universe, rules, and 14-day price history; Kalshi public API for live orderbook depth.