Trading Signals

Empirically validated, mechanism-grounded trading signals from the Observatory.

0 ACTIVE Updated 2026-04-16 Not investment advice

Kitchin Composite

BULLISH

+5.6% Expected 6m S&P return

77% Historical hit rate

2.33 Sharpe ratio

~6951 Implied S&P 6m target

ISRATIO (Inventory/Sales)

1.35 FALLING

Score: +1

EBP (Excess Bond Premium)

-0.296 EASING

Score: +1

2 /2

Maximum bullish score. Backtest n=66 score+2 months (1993–2026).

ISRATIO 12m momentum + EBP 3m momentum. Backtest n=66 score+2 months, 1993-2026.

Kitchin Phase Clock

CONTRACTION

73% Through contraction

1mo To trough est.

2026-05-07 Estimated trough date

Portfolio tilt: DEFENSIVE — Inventory liquidation, production cuts. Favor bonds, quality, low-vol.

Next transition: TROUGH est. 2026-05-07

OOS validation: p=0.0003 — expansion returns statistically higher than contraction returns.

VIX Term Structure

NEUTRAL

0.896 VIX/VIX3M ratio

19.12 VIX (spot)

21.34 VIX3M (3-month)

1.1 Trigger threshold

When ratio ≥ 1.1 (severe backwardation): +9.69% mean 60d return, 81% hit rate, Cohen's d=1.005

Ratio >= 1.10 = severe backwardation = CONFIRMED 60d forward edge (Cohen d=1.005, 81% hit rate). DA: crisis clustering PASSED (65% non-crisis events, hit rate 84.4%). Current: CONTANGO

COT Positioning

MONITORING

5 Markets tracked

0 Active signals

COT positioning z-scores vs 52-week history. Note: All directional COT trading signals in this dashboard were retired by the April 2026 five-test battery. Corn, wheat, and crude oil failed all pass/fail criteria (0/4). EUR/USD was directionally inverted. Soybeans failed out-of-sample replication. Z-scores are shown for monitoring only — no active trading edge.

✕ Corn (CBOT): z=-2.62 ⚑ at extreme — SIGNAL KILLED: 0/4 battery — pure noise

✕ Crude Oil WTI (NYMEX): z=1.28 — SIGNAL KILLED: 0/4 battery — specificity 1.11×

◎ Gold (COMEX): z=0.94 — SUGGESTIVE_NOT_CONFIRMED

◎ E-mini S&P 500 (CME) — Asset Manager: z=-1.76 — MONITORING

✕ Wheat SRW (CBOT): z=2.09 — SIGNAL KILLED: 0/4 battery — z=−0.29 (directionally wrong)

Gold/Silver Ratio

LONG SILVER / SHORT GOLD

83.1 Current GSR

90th Historical percentile

+15.4pp Median silver edge 24m

p=1.4e-8 Statistical significance

Cheap gold
GSR <40

Neutral
40–80

Signal Zone
GSR >80

42 Mean 68 Threshold 80 Max 126

$4451 Gold / oz

$53.6 Silver / oz

1.8 5yr z-score

GSR ≥80 (historical 94th pct at entry) = contrarian LONG SILVER vs SHORT GOLD. CONFIRMED p=1.4e-8, n=780 months. Battery 3/4 — specificity 11.29×, held-out z=4.60.

⚑ Signal entered 2026-04-09 at GSR=86.07 (94th pct). GSR has declined as silver outperformed gold (+5.3% vs +1.6% since entry). Signal remains active above the 80 threshold.

ENSO Corn-Wheat Spread

FLAT

-0.16 ONI (current)

6m Signal horizon

La Niña
−4.8% spread

Neutral
No edge

El Niño
−3.0% spread

Neutral ENSO (ONI=−0.16). No directional edge until ONI crosses ±0.5.

Validated: La Niña n=72mo mean spread −4.8% (p=0.015), El Niño n=60mo mean spread −3.0% (p=0.050). Signal activates when ONI crosses ±0.5.

Structural Valuation Backdrop

STRUCTURAL WARNING

~ Shiller CAPE (est. 2026)

th pct CAPE percentile

% Implied 10yr real return

Shiller CAPE th pct

Est. ~ vs historical mean . Post-1980 R²=0.667, p=. Implied 10yr real return: % annualised.

Real House Price Index th pct

+% above 30yr trend. . Post-1990 confirmed p= at 5yr horizon.

Earnings Yield Gap th pct

% gap (equity earnings yield minus 10Y Treasury). Bonds now yield more than equities on earnings yield basis.

⚑ This is not a short-term timing signal. CAPE at 30 in 1996 would have missed the 1997–1999 blowoff. Structural signals operate at 5-10 year horizons. All three metrics can remain elevated for years.