Delta

← opportunities · hypothesis b75281b0 · 09-09 14:00Z · MOCK data

WHY GBP/USD SHORT?SHORTrejectedfx.mean_reversion4H-2DCRYPTO DELEVERAGINGMOCK

Lifecyclecandidate → qualified → approved → executed, or rejected / expired

No lifecycle transitions recorded: this opportunity was decided before lifecycle tracking began.

Thesisrules-explainer-1.0.0 · from recorded evidence

fx.mean_reversion SHORT GBP/USD: mean reversion signal, horizon 4H-2D
volatility not expanding
price 3.26 sd above its 50-bar mean
efficiency 0.10 < 0.25: ranging market
RSI 83: stretched
Against: usd weakening regime (0.35) hurts a short GBPUSD (assumed sensitivity -1.0)

DecisionREJECTED

rejected: confidence_minimum; consecutive_loss_pause

tradeconfidence minimum42.3 vs 55uncalibrated score; event policy adds 0
portfolioconsecutive loss pause0 vs 4paused until 2026-09-10T00:00:00+00:00

Supporting evidence4

▲regime0.47volatility not expanding
▲technical3.26price 3.26 sd above its 50-bar mean
▲technical0.10efficiency 0.10 < 0.25: ranging market
▲technical83RSI 83: stretched

Contradicting evidence3

▼macro-0.22usd weakening regime (0.35) hurts a short GBPUSD (assumed sensitivity -1.0)
▼cross asset0.88EURUSD momentum +0.9 sd with correlation +0.71 contradicts the trade
▼cross asset1.07AUDUSD momentum +1.1 sd with correlation +0.60 contradicts the trade

Invalidation conditionschecked every strategy bar

stop at 1.39713 (3.0 x ATR(14))
move extends beyond 3.7 sd: range broken
horizon exceeded: 48 bars held

Confirmationuncalibrated scores

cross asset	CONTRADICTS	−0.34	2 ev.
macro	MIXED	−0.19	2 ev.
technical	CONFIRMS	+0.76	3 ev.

Event riskscore 0.83

next event: none in window
blocks entry: no
size multiplier: 1.00×
confidence add: +0

Liquidity

state at decision: high
score: +1.00

Risk layers0.24% at stop

eligibility	1 ✓	0 ✗
trade	9 ✓	1 ✗
portfolio	10 ✓	1 ✗
system	2 ✓	0 ✗

Three-layer riskcomputed on read from stored checks (no system report then)BLOCK

TRADE RISKBLOCK

blocked by confidence_minimum

confidence minimum42.3 / 55130%
strategy stage eligibleBACKTEST (sandbox) / PAPER|SHADOW|LIMITED_CAPITAL|PRODUCTION
market liquidity allows entryhigh / not closed/maintenance/rollover/reopen
quote fresh0 / 150%
spread sane1.06 / 2.542%
event policyclear / clear
risk reward minimum1.768 / 1.585%
stop distance sane3 / 0.517%
no conflicting position0 open, 0 pending / none
size above minimum22,000 / 1,0005%
risk per trade0.244 / 124%

PORTFOLIO RISKBLOCK

blocked by consecutive_loss_pause

consecutive loss pause0 / 40%
daily loss limit-0.015 / -21%
drawdown limit0.411 / 104%
gross exposure30.59 / 5006%
leverage0.306 / 56%
currency exposure-30.59 / 30010%
usd factor exposure28.13 / 25011%
correlated cluster risk0.244 / 1.516%
crypto exposure0 / 500%
portfolio volatility3.254 / 2016%
max open positions1 / 813%

SYSTEM RISKPASS

2 checks passed

kill switch inactiveinactive / inactive
trading mode validpaper / paper|backtest|live

BLOCK = a hard check failed; WARN = passed but at >= 80% of a maximum, within 10% of a minimum, or a monitor above its level. Right column: how close a passed check is to its limit. risk_layers-1.0.0+63bba1f7

ConfidenceMODERATE · conf-1.0.0

Not a probability. Heuristic weighted score. Not a probability of profit. Calibrate on >=200 paper trades (reliability diagram) before treating as one.

signal agreement	0.12	0.0
macro confirmation	0.18	−1.7
regime alignment	0.12	−0.5
cross asset	0.12	−2.0
volatility	0.08	−4.0
liquidity	0.08	+4.0
event risk	0.10	−4.2
historical performance	0.12	0.0
risk reward	0.08	+0.7

Trade plan

Entry zone: 1.38515 – 1.38700
Reference: 1.38608
Stop: 1.39713 · 3.0 ATR
Target 1: 1.36653 · 50-bar mean
Reward / risk: 1.77
Expected horizon: 4H-2D · max 48 bars
Expected return: unavailable
OOS win rate (uncal.): unavailable

unavailable: no out-of-sample record for this strategy/instrument

Size and portfolio impactrisk-1.0.0

Size0.000not approved

Risk0.24%$243 · notional $30,494

Portfolio impact

Not evaluated: the opportunity was rejected before the portfolio projection.

Every risk check

Layer	Check		Value	Limit	Detail
system	kill switch inactive	✓	inactive	inactive
system	trading mode valid	✓	paper	paper\|backtest\|live
eligibility	strategy stage eligible	✓	BACKTEST (sandbox)	PAPER\|SHADOW\|LIMITED_CAPITAL\|PRODUCTION	MOCK sandbox: strategy has not earned PAPER; simulated on synthetic data only
trade	market liquidity allows entry	✓	high	not closed/maintenance/rollover/reopen	session overlap: london + new_york
trade	quote fresh	✓	0	15	minutes since last quote
trade	spread sane	✓	1.06	2.5	spread / typical spread
trade	event policy	✓	clear	clear
trade	confidence minimum	✗	42.3	55	uncalibrated score; event policy adds 0
trade	risk reward minimum	✓	1.768	1.5
trade	stop distance sane	✓	3	0.5	stop distance in ATRs
trade	no conflicting position	✓	0 open, 0 pending	none
portfolio	consecutive loss pause	✗	0	4	paused until 2026-09-10T00:00:00+00:00
portfolio	daily loss limit	✓	-0.0146	-2	% since trading-day start
portfolio	drawdown limit	✓	0.4113	10	% below peak equity
trade	size above minimum	✓	22,000	1,000	units (GBP)
trade	risk per trade	✓	0.2441	1	% of equity at the stop
portfolio	gross exposure	✓	30.59	500	% of equity
portfolio	leverage	✓	0.306	5
portfolio	currency exposure	✓	-30.59	300	largest net currency exposure: GBP
portfolio	usd factor exposure	✓	28.13	250	beta-weighted USD-index-equivalent exposure, % of equity
portfolio	correlated cluster risk	✓	0.244	1.5	no correlated open positions
portfolio	crypto exposure	✓	0	50	% of equity
portfolio	portfolio volatility	✓	3.254	20	forecast annualised volatility
portfolio	max open positions	✓	1	8

Sizing tracenever sized from confidence alone

Step	Qty after	×	Detail
fixed_risk	45,069		0.5% of $99,675 = $498 over a stop of 0.0110581 ($0.0111/unit)
confidence_adjusted	22,534	0.50	confidence 42 -> x0.50 (never above x1.00)
portfolio_risk_adjusted	22,534	1.00	correlated open risk $0 of $1,495 cluster budget -> x1.00
event_policy	22,534	1.00	event policy x1.00, liquidity x1.00
caps	22,000		rounded down to step 1000

Execution and outcome

No orders: this hypothesis did not reach the broker.

Versionsevery trade is traceable

data mode: MOCK
timestamp: 2026-09-09T14:00:00Z
data source: mock
risk version: risk-1.0.0+e561f3f2
model version: regime-rules-1.0.0+ba09c5ab
feature version: fs-1.0.0+bbfe6af5
strategy version: fx.mean_reversion@1.0.0
execution version: exec-1.0.0+821dd562
confidence version: conf-1.0.0+18bcb659
event policy version: evpol-1.0.0+391f9e5d