Market Fragility Dashboard / NDX

3-layer risk stack: fragility · crowding · amplifiers · auto-updated every hour

Last updated: 2026-09-21 23:58 UTC
Fragility · LPI + Regime
57.8 Falling
Cushion thick, stable — full risk budget
Crowding · Basis-Trade Proxy
90
Basis trade crowded — issuance shocks amplified
Amplifiers · Signals Confirmed
3/5
Threshold reached: mechanical selling likely exhausting
关键结论 Key Takeaway rule-based · sizing & hedging only
generated 2026-09-21 23:58 UTC · deterministic (no forecast)
Fragility: LPI 58 (Low · Falling), ΔLPI 13w -5.1, breadth 0/4 factors elevated-and-rising. In band 40-60, 8-week P(>10% drawdown) was 10.6% vs 8.5% baseline.
Crowding: basis-trade proxy 90 (basis trade crowded — issuance shocks amplified); most-extreme equity positioning E-mini Russell 2000 at 9th pctile (z -1.5).
Amplifiers: dealer gamma positive / dampening via SPY GEX (homebrew, flip -0.2% vs spot); VIX term structure in backwardation; crypto funding positive; 3/5 de-lever signals confirmed.
Stance: Cushion is thick today, but crowded basis positioning makes issuance events the tail to watch — keep calendar hedges around auction / refunding windows.保持仓位—关注基差
第一层 · 脆弱度 Fragility Liquidity Pressure Indexweeks–months
Fast funding & liquidity public-data confirmation modules
Reserves / GDPWARNING
9.28%
15th pctile · raw reserves $3.01T (context only) · two-week persistence required
current observation 2026-09-16 · fetched 2026-09-21 23:56 UTC · weekly (GDP quarterly, carried forward)
About Reserves / GDP
Definition: Fed reserve balances ÷ nominal GDP (GDP carried forward weekly); rolling 156-week percentile.
Why it matters: Scales the reserve buffer to the economy rather than relying on fixed dollar levels.
How to read: Below the 20th percentile for two weekly readings = warning; below the 10th percentile for two = elevated. Raw dollars are context only.
Source: official public data
Cadence: weekly; GDP quarterly · Observation: 2026-09-16 · Fetched: 2026-09-21 23:56 UTC
Limitations: GDP is quarterly and carried forward; it is not a real-time estimate.
NY Fed repo accepted · 7dWATCH
$0.4B
today $0.0B · 5 non-zero days / 7 · 88th pctile of available history
current observation 2026-09-21 · fetched 2026-09-21 23:56 UTC · daily
About NY Fed repo accepted
Definition: Sum of accepted amounts in NY Fed repo operation results; headline is trailing seven calendar-day sum.
Why it matters: Use of the central-bank backstop can corroborate secured-funding pressure.
How to read: Read persistence and the operation's own trailing distribution, not a fixed dollar trigger; the daily amount remains visible for context.
Source: official public data
Cadence: daily · Observation: 2026-09-21 · Fetched: 2026-09-21 23:56 UTC
Limitations: Includes reported repo operations; operational use is confirmation, not a market-top signal by itself.
SOFR p99 − medianNORMAL
8.0 bp
Tail-rate pressure; use with persistent spread and repo use.
current observation 2026-09-18 · fetched 2026-09-21 23:56 UTC · daily business day
About SOFR p99 − median
Definition: NY Fed SOFR 99th-percentile rate minus the published volume-weighted median SOFR.
Why it matters: Tail funding pressure can widen before the median rate moves.
How to read: A widening delta is corroboration when it coincides with persistent SOFR–IORB pressure and repo use.
Source: official public data
Cadence: daily · Observation: 2026-09-18 · Fetched: 2026-09-21 23:56 UTC
Limitations: A distributional funding-stress confirmation signal, not a standalone prediction.
Dollar Funding Pressure 美元资金压力 · SOFR−IORB + bank reserves + TGA
0/ 3
No resonance无共振
0/3 conditions active
None of the three strategy conditions is active; dollar funding looks orderly.
Observations: SOFR 2026-09-18 · reserves 2026-09-16 · TGA 2026-09-16 · fetched 2026-09-21 23:56 UTC
These absolute levels are a user-defined strategy overlay, not a universal empirical law: +3bp, $2.90T/$2.80T and $0.90T/$1.00T are the operator’s own trigger levels, chosen as roughly the 95th percentile of their own framework.
A. Funding price资金价格○ AT OR BELOW IORB
-5.0 bp
SOFR − IORB, basis points
1-day change+0.0 bp
Filtered 5-obs median-3.0 bp
Latest observationnon-calendar · eligible
filtered median -3.0bp vs >= +3bp (not met) · last 3 non-calendar readings not all >= +3bp (not met)
current observation 2026-09-18 · fetched 2026-09-21 23:56 UTC · daily business day
About SOFR − IORB
Definition: Daily SOFR minus the Fed's Interest on Reserve Balances, in basis points. Emphasized line is a 5-observation median of non-calendar readings.
Why it matters: Positive readings mean secured overnight funding trades above the administered reserve rate, which can indicate scarcity or balance-sheet pressure.
How to read: Zero/negative is normal/easier. Calendar observations do not create alerts. The strategy overlay leg turns on when the filtered median and the last three non-calendar readings are all ≥ +3bp; the stricter +2bp / +5bp bands are kept as analytical reference only.
Source: official public data
Cadence: daily · Observation: 2026-09-18 · Fetched: 2026-09-21 23:56 UTC
Limitations: Month/quarter ends, major corporate tax dates, and large Treasury settlement dates are marked as calendar noise; it is a confirmation signal, not a market-top predictor.
B. Liquidity buffer银行准备金○ AMPLE
$3.014T
Reserve balances, $ trillions
1-week change▲ +22B
4-week change▲ +79B
Trend speed▲ +20B/wk
$3.014T vs <= $2.90T (not met) · 4w +79B vs <= -50B (not met)
current observation 2026-09-16 · fetched 2026-09-21 23:56 UTC · weekly (H.4.1, week ended Wednesday)
About Bank reserve balances
Definition: Reserve balances held by depository institutions at the Federal Reserve, weekly average of daily figures for the week ended Wednesday (H.4.1 Table 1; FRED WRESBAL is the same series).
Why it matters: Reserves are the banking system's settlement buffer. A thinner buffer makes secured overnight funding more likely to trade above the administered reserve rate.
How to read: Strategy overlay: leg active when the level is at or below $2.90T and the 4-week change is at or below −$50B; elevated at ≤ $2.85T or ≤ −$100B. One condition alone shows as approaching and does not count toward resonance.
Source: official public data
Cadence: weekly (Thursday release, week ended Wednesday) · Observation: 2026-09-16 · Fetched: 2026-09-21 23:56 UTC
Limitations: Weekly averages lag intraday conditions; no missing week is interpolated or forward-filled. Reserves/GDP remains the separate normalized structural metric.
C. Fiscal drain财政抽水○ NEUTRAL
$0.877T
Treasury General Account, $ trillions
1-week change▼ -6B
4-week change▼ -77B
Rebuild speed▼ -19B/wk
$0.877T vs >= $0.90T (not met) · 4w -77B vs >= +50B (not met)
current observation 2026-09-16 · fetched 2026-09-21 23:56 UTC · weekly (H.4.1, week ended Wednesday)
About Treasury General Account
Definition: The U.S. Treasury's operating cash balance at the Federal Reserve, weekly average of daily figures for the week ended Wednesday (H.4.1 Table 1; FRED WTREGEN is the same series). The Daily Treasury Statement closing balance is used only as an official fallback tail.
Why it matters: Rebuilding the TGA moves cash from the banking system to the Treasury's account at the Fed, which drains reserves one-for-one, all else equal.
How to read: Strategy overlay: leg active when the level is at or above $0.90T and the 4-week change is at or above +$50B; elevated at ≥ $1.00T or ≥ +$100B. One condition alone shows as approaching.
Source: official public data
Cadence: weekly (Thursday release, week ended Wednesday); DTS fallback is daily · Observation: 2026-09-16 · Fetched: 2026-09-21 23:56 UTC
Limitations: Bill-settlement timing makes the balance lumpy; a weekly average smooths but also lags. Missing weeks stay missing rather than being carried forward.
Pressure path 传导路径 · read left to right
TGA rebuild财政抽水
○ NEUTRAL
Treasury cash balance rises
drains
Reserves drain准备金下降
○ AMPLE
bank settlement buffer thins
prices up
SOFR above IORB资金价格上升
○ AT OR BELOW IORB
secured funding prices above the Fed's floor
Direction. A rebuilding TGA moves cash from banks to the Treasury’s account at the Fed, which drains bank reserves one-for-one; a thinner reserve buffer makes secured overnight funding price above IORB. All three moving together is what the strategy overlay treats as resonance.
Three aligned panels share one x-axis and one range control.
A. Funding price SOFR − IORB, basis points · raw muted, filtered non-calendar 5-observation median emphasized, hollow markers = calendar noise
Show daily change bars (Δ bp/day)
B. Liquidity buffer bank reserve balances, $ trillions · $2.90T / $2.80T reference band, 4-week slope cue
C. Fiscal drain Treasury General Account, $ trillions · $0.90T watch / $1.00T threshold, 4-week slope cue
Read guide. 0/3 normal, no resonance. 1/3 watch — a single leg is common and is not a fragility signal. 2/3 yellow warning — funding pressure building. 3/3 structural-top risk signal — fragility confirmation, explicitly not a deterministic market-top prediction. A stale or unavailable leg never counts as triggered; the header shows an incomplete-data state instead. The stricter +2bp / +5bp bands are kept as analytical reference lines: they are this dashboard’s persistence-based statistical bands, while the emphasized +3bp line is the operator’s strategy threshold.
About the dollar funding pressure trio · definitions, formulas, sources, limitations
What this section is. Three official funding indicators in one place: the price of secured overnight funding, the size of the banking system’s reserve buffer, and the Treasury’s cash balance that drains it. These absolute levels are a user-defined strategy overlay, not a universal empirical law: +3bp, $2.90T/$2.80T and $0.90T/$1.00T are the operator’s own trigger levels, chosen as roughly the 95th percentile of their own framework.

A. Funding price — SOFR − IORB. Formula: (SOFR − IORB) × 100 in basis points, daily. The emphasized line is the median of the last five non-calendar observations; month-end, quarter-end, major corporate tax dates and large Treasury settlement dates are flagged as calendar noise and excluded from the filter and from the rule. Why it matters: a positive spread means secured funding trades above the Fed’s administered floor, which points to collateral or balance-sheet pressure. Strategy overlay leg: active when the filtered median ≥ +3.0bp and the last 3 eligible (non-calendar) observations are all ≥ +3.0bp, so one calendar spike cannot trigger it. The +2bp / +5bp lines remain as this dashboard’s stricter analytical reference bands.

B. Liquidity buffer — bank reserve balances. Reserve balances of depository institutions at Federal Reserve Banks, weekly average of daily figures for the week ended Wednesday (H.4.1 Table 1; identical to FRED WRESBAL). Changes are computed as current − as-of(t−7d) and current − as-of(t−28d) in $ billions against real observation dates; trend speed is the 4-week change ÷ 4. Strategy overlay leg: active when level ≤ $2.90T AND 4-week change ≤ −$50B; elevated when level ≤ $2.85T or 4-week change ≤ −$100B. If only one of the two conditions holds, the card reads approaching and the leg does not count. Reserves/GDP with its rolling percentile is kept separately above as normalized structural context.

C. Fiscal drain — Treasury General Account. The Treasury’s operating cash balance at the Fed, same weekly basis (H.4.1 Table 1; identical to FRED WTREGEN); the Daily Treasury Statement closing balance is an official fallback tail when the weekly path is stale. Strategy overlay leg: active when level ≥ $0.90T AND 4-week change ≥ +$50B; elevated when level ≥ $1.00T or 4-week change ≥ +$100B.

How to read the three-leg resonance. 0/3 normal · 1/3 watch (the named leg only) · 2/3 yellow warning, funding pressure building · 3/3 structural-top risk signal. 3/3 is a fragility confirmation under this overlay — it is not a standalone or deterministic market-top forecast, and it carries no timing claim.

Official sources. NY Fed SOFR reference rates · Federal Reserve PRATES (IORB) · Federal Reserve H.4.1 · FRED WRESBAL · FRED WTREGEN · Treasury Daily Treasury Statement

Cadence, observation dates and fetch times.
SOFR − IORB: observation 2026-09-18 · fetched 2026-09-21 23:56 UTC · daily business day · current · source: official
Reserve balances: observation 2026-09-16 · fetched 2026-09-21 23:56 UTC · weekly (H.4.1, week ended Wednesday) · current · source: FRED WRESBAL (FRED cached official CSV) + Federal Reserve H.4.1 current release
Treasury General Account: observation 2026-09-16 · fetched 2026-09-21 23:56 UTC · weekly (H.4.1, week ended Wednesday) · current · source: FRED WTREGEN (FRED cached official CSV) + Federal Reserve H.4.1 current release

Limitations. Reserves and the TGA are weekly averages released with a lag, so they cannot confirm intraday funding stress. No missing official observation is fabricated, interpolated or forward-filled: a gap stays a gap, a stale leg is labelled stale and excluded from the resonance count, and each leg fails independently so one unavailable series never blanks the other two. The absolute dollar and basis-point levels are regime-dependent — they are the operator’s chosen strategy thresholds, not an estimated statistical law.
57.8 / 100 · 4/4 factors Neutral / Watch for inflection 中性 Low · Falling
0 Resilient406080100 Extreme
Regime · 状态
Cushion thick, stable — full risk budget
ΔLPI 13w -5.1 · ΔLPI 4w +5.8 · breadth 0/4 factors >70th and rising
Quadrant occupancy · trailing 52w
Danger zone 危险区18w
Cushion thick 缓冲厚 ●17w
Inflection watch 拐点观察10w
Decompress 减压7w
Regime Clock · LPI level × Δ13w
Short Rate 资金价格61
3.65% SOFR overnight
wow ▲ +1 · 13w ▲ +4
SOFR 3.65%, steady — neutral funding backdrop. 资金中性
Duration Supply 久期供给77
$241B coupon supply, 4w sum
wow ▬ +0 · 13w ▼ -1
Heavy issuance ($241B 4w) — refunding waves dominate; keep calendar hedges. 久期供给高企
Net Liquidity 银行水位 (inv)43
$5.84T WALCL − TGA − RRP
wow ▬ +0 · 13w ▲ +3
Net liquidity steady near $5.84T — reserves stable. 流动性平稳
Vol Amplifier 波动放大器34
14.9 VIX spot
wow ▼ -22 · 13w ▼ -29
Calm vol regime (VIX 14.9) — gamma dampening, shocks absorbed. 波动低企
Conditional Tail Table — forward S&P 500 by LPI band
Current LPI 58 → band 40-60: historically mean 4-week return +0.7% (vs +0.8% baseline) but P(>10% drawdown) 3.1% vs 2.3% baseline — tail risk ~1.3× normal.
LPI bandHorizonNMean retMedianFwd vol (ann.)P(<-5%)P(<-10%)Avg MaxDD
All weeks (baseline)4w1067+0.8%+1.3%13.6%15.9%2.3%-2.6%
All weeks (baseline)8w1063+1.5%+2.2%14.4%31.6%8.5%-4.4%
0-404w347+0.5%+1.0%9.8%10.7%0.6%-1.9%
0-408w347+1.0%+1.6%11.1%23.1%3.5%-3.4%
40-604w482+0.7%+1.5%14.0%14.5%3.1%-2.7%
40-608w479+1.1%+2.1%14.9%31.9%10.6%-4.8%
60-804w238+1.2%+1.8%18.4%26.5%3.4%-3.5%
60-808w237+3.1%+3.9%18.0%43.5%11.4%-5.3%
80-1004winsufficient
80-1008winsufficient
Drawdown probability by LPI band — 90% block-bootstrap CI
Forward windows overlap weekly, which inflates effective N (observations are not independent). Max drawdown uses weekly closes and therefore understates intraweek troughs. Percentiles feeding each week’s LPI are expanding and strictly causal (no look-ahead); forward returns are realized outcomes. Chart CIs are 90% (5th–95th pct) from a circular block bootstrap of the weekly forward-return series within each band (block length = horizon in weeks to respect overlapping-window autocorrelation, 1500 resamples); bands with N<30 are merged upward for the chart and flagged ⚠ in the table.
LPI — Trailing 52 Weeks
Composite percentile pressure · thresholds at 60 / 80
LPI — Full Reconstruction
1071 weeks from 2006-02-08 · expanding no-look-ahead percentiles · use range selector
2026 H2 Stress Calendar · 压力日历
severity-sized markers · today cursor
2026 H2 压力日历 (projected stress calendar)
Sep (~79th pct): FOMC Sep 15-16 · corp estimated tax Sep 15 · quarter-end TGA refill Sep 30 — 4 simultaneous drains
Nov (2nd highest): quarterly refunding issuance + FOMC late Oct
Aug (3rd): refunding settlement peak mid-August
Jul: lowest pressure month of H2
第二层 · 拥挤度 Crowding CFTC COT positioningweeks
S&P 500 concentrationINSUFFICIENT-HISTORY
38.63% / 28.05%
Top-10 / Top-5 · 31 locally archived daily snapshot(s)
current observation 2026-09-18 · fetched 2026-09-21 23:56 UTC · daily business day
About S&P 500 concentration
Definition: Sum of the ten and five largest weights in the official SPY daily holdings file.
Why it matters: Concentration increases the index's dependence on a small group of securities.
How to read: Top-10 / Top-5 are observations. Percentile alerts stay disabled until the locally archived series has sufficient history.
Source: official public data
Cadence: daily business day · Observation: 2026-09-18 · Fetched: 2026-09-21 23:56 UTC
Limitations: SPY is a practical S&P 500 proxy and the issuer only supplies its current holdings file; history begins with this pipeline.
COT as of 2026-09-15 — 6d old
Equity index · Leveraged Funds net
E-mini S&P 500
-293,143
lev net contracts · 4w -11,741
pctile 51z +0.7
E-MINI S&P 500
Nasdaq-100
-13,052
lev net contracts · 4w +54,657
pctile 19z +1.1
NASDAQ-100 Consolidated
E-mini Russell 2000
-97,203
lev net contracts · 4w +2,583
pctile 9z -1.5
RUSSELL E-MINI
Treasury notes · Leveraged Funds net (basis-trade legs)
2-Year UST
-1,294,575
lev net contracts · 4w -51,571
pctile 17z +1.9
UST 2Y NOTE
5-Year UST
-1,986,928
lev net contracts · 4w +182,886
pctile 13z +1.3
UST 5Y NOTE
10-Year UST
-1,868,126
lev net contracts · 4w +360,887
pctile 8z -0.2
UST 10Y NOTE
Context · Managed Money / lev net
WTI Crude
-10,022
MM net contracts · 4w +674
pctile 26z +2.0
CRUDE OIL, LIGHT SWEET-WTI
Gold
+133,116
MM net contracts · 4w -8,532
pctile 64z +0.1
GOLD
US Dollar Index
-4,909
lev net contracts · 4w -13,021
pctile 41z -0.2
USD INDEX
90 / 100 · basis-trade proxy 基差交易拥挤度 Basis trade crowded — issuance shocks amplified
0 Benign406080 Crowded100
2Y (AM long 90 / LF short 83)86
5Y (AM long 90 / LF short 88)89
10Y (AM long 100 / LF short 92)96
Proxy = mean over 2y/5y/10y of [pctile(Asset-Mgr net long) + pctile(Lev-Fund net short)] / 2, measuring how extreme the AM-long / LF-short configuration is vs its own history. Not added to the 4-factor LPI composite (a 5-factor variant is a future option).
第三层 · 放大器与触发器 Amplifiers & Triggers de-lever bottom signals + dealer gammahours–days
Credit transmission & AI leadership fast variables · fixed basket
HY OASSTALE
Last observation 2026-08-06; not rendered as current.
STALE / unavailable observation 2026-08-06 · fetched 2026-09-21 23:56 UTC
About HY OAS
Definition: ICE BofA option-adjusted spread from the listed FRED series.
Why it matters: Credit spread widening can confirm risk transmission beyond equities.
How to read: Read the current spread with the 60-day change; no hard-coded current value is used.
Source: official public data
Cadence: daily business day · Observation: 2026-08-06 · Fetched: 2026-09-21 23:56 UTC
Limitations: Index-level OAS is not an AI-issuer credit spread.
IG OASSTALE
Last observation 2026-08-06; not rendered as current.
STALE / unavailable observation 2026-08-06 · fetched 2026-09-21 23:56 UTC
About IG OAS
Definition: ICE BofA option-adjusted spread from the listed FRED series.
Why it matters: Credit spread widening can confirm risk transmission beyond equities.
How to read: Read the current spread with the 60-day change; no hard-coded current value is used.
Source: official public data
Cadence: daily business day · Observation: 2026-08-06 · Fetched: 2026-09-21 23:56 UTC
Limitations: Index-level OAS is not an AI-issuer credit spread.
BBB OASSTALE
Last observation 2026-08-06; not rendered as current.
STALE / unavailable observation 2026-08-06 · fetched 2026-09-21 23:56 UTC
About BBB OAS
Definition: ICE BofA option-adjusted spread from the listed FRED series.
Why it matters: Credit spread widening can confirm risk transmission beyond equities.
How to read: Read the current spread with the 60-day change; no hard-coded current value is used.
Source: official public data
Cadence: daily business day · Observation: 2026-08-06 · Fetched: 2026-09-21 23:56 UTC
Limitations: Index-level OAS is not an AI-issuer credit spread.
AI basket breadthNORMAL
68% / 79%
above 50d / 200d MAs · 19/19 fixed constituents available
current observation 2026-09-21 · fetched 2026-09-21 23:56 UTC · daily business day
About AI basket breadth
Definition: Percent of the fixed documented basket above its 50-day and 200-day moving average.
Why it matters: Breadth shows whether leadership is broadening or narrowing inside the selected AI supply-chain basket.
How to read: Read 50d / 200d together with semiconductor relative strength; the basket is fixed to limit drift.
Source: official public data
Cadence: daily business day · Observation: 2026-09-21 · Fetched: 2026-09-21 23:56 UTC
Limitations: The basket is subjective and has survivorship bias; it is not an index or recommendation. SMH is used only as an ETF benchmark.
SMH / SPYNORMAL
0.7706
60d -11.4% · vs 50d 0.7444 / 200d 0.6805
current observation 2026-09-21 · fetched 2026-09-21 23:56 UTC · daily business day
About SMH / SPY relative strength
Definition: Adjusted-close ratio of SMH to SPY, with 50-day and 200-day moving averages.
Why it matters: Shows whether semiconductor leadership is outperforming the broad market.
How to read: A ratio below a declining 200-day average is a transmission/leadership warning, not a directional call.
Source: official public data
Cadence: daily business day · Observation: 2026-09-21 · Fetched: 2026-09-21 23:56 UTC
Limitations: ETF composition changes can create structural breaks; the ETF is a benchmark, not a basket constituent.
SOXX / SPYNORMAL
0.7231
60d -15.2% · vs 50d 0.6923 / 200d 0.6157
current observation 2026-09-21 · fetched 2026-09-21 23:56 UTC · daily business day
About SOXX / SPY relative strength
Definition: Adjusted-close ratio of SOXX to SPY, with 50-day and 200-day moving averages.
Why it matters: Shows whether semiconductor leadership is outperforming the broad market.
How to read: A ratio below a declining 200-day average is a transmission/leadership warning, not a directional call.
Source: official public data
Cadence: daily business day · Observation: 2026-09-21 · Fetched: 2026-09-21 23:56 UTC
Limitations: ETF composition changes can create structural breaks; the ETF is a benchmark, not a basket constituent.
① VIX Term Structure · Cboe official
N/A pts
VIX9D N/A → VIX N/A → VIX3M N/A
STALE / DATE MISMATCH
current observation 2026-09-18 · fetched 2026-09-21 23:56 UTC · daily business day
About VIX term structure
Definition: Cboe official VIX9D, VIX, and VIX3M closes. The curve is used only when all three latest observations share the same date.
Why it matters: Backwardation can accompany acute volatility demand.
How to read: Contango is VIX3M above spot; a date mismatch or stale series is unavailable, never ratioed.
Source: Cboe daily index history
Cadence: daily business day · Observation: 2026-09-18 · Fetched: 2026-09-21 23:56 UTC
Limitations: a curve shape is a volatility-market condition, not a directional forecast.
② Crypto Funding (Binance)
BTC +0.7541%
ETH +0.5424% · OKX BTC +0.0100%
BOTH POSITIVE ✅ — Risk appetite returning
③ CTA Proxy (CFTC COT)
-13,052
-3.8% OI · as of 2026-09-15
WoW: +23,832 contracts
SHORT COVERING ✅
④ Cross-Sector Correlation
0.216
2W vs 1M avg: 0.179 → 0.216
ELEVATED ⚠️
Signals confirmed
3/5Threshold reached: mechanical selling likely exhausting
Index GEX (homebrew) 指数伽马 · SqueezeMetrics 约定 · $bn / 1% move
De-lever signal #5 — Dealer Gamma · source SPY GEX (homebrew) (+8.24 $bn/1%): POSITIVE ✅ dealers dampening
SPYlive
+8.24 $bn/1%
net GEX · positive — dealers dampen vol
Flip 771.64 · Spot 773.50 · spot +0.2% above flip
gamma walls: 774 (+1.59), 772 (+1.51), 780 (+0.99)
call wall 774 · put wall 761
front 45d · 10 exp · 2135 contracts
accumulating history (46/60 days)
QQQlive
+4.34 $bn/1%
net GEX · positive — dealers dampen vol
Flip 722.42 · Spot 741.47 · spot +2.6% above flip
gamma walls: 725 (+0.59), 740 (+0.35), 735 (+0.31)
call wall 725 · put wall 705
front 45d · 10 exp · 2174 contracts
accumulating history (46/60 days)
SPY Net GEX by Strike (±10% of spot)
Green = positive gamma (dampening) · red = negative · dashed = spot, solid = gamma flip
GEX Monitor 做市商伽马监控 (Dealer Gamma)
NVDA cached 2026-08-13 19:08 UTC
$668M
net GEX · positive (dealers stabilize)
Flip 219.3 · Spot 226.2 · +3.1% above flip
exp 2026-08-14 · as of 2026-08-13T19:08:05.3497863Z · cached
excluded: tier-restricted
AAPL cached 2026-08-13 15:02 UTC
$176M
net GEX · positive (dealers stabilize)
Flip 301.1 · Spot 304.9 · +1.2% above flip
exp 2026-08-14 · as of 2026-08-13T15:02:52.1756633Z · cached
excluded: tier-restricted
AMD
N/A
excluded: tier-restricted
MU
N/A
excluded: tier-restricted
Phase-1 GEX signal — NVDA proxy (free-tier: SPX unavailable): POSITIVE ✅ dealers dampening (confirmed)
NVDA Net GEX — History
Positive = dealer long gamma (dampening) · negative = short gamma (amplifying)
Requests used today: 0 / 5 · free tier resets 00:00 UTC · excluded: AMD, MU, NVDA, AAPL
VIX 30-Day History
Spot VIX above 20 = elevated fear regime
VIX Term Structure Snapshot
Upward slope = contango = panic receding
NDX Leveraged Money Net (COT)
CTA proxy covering = selling abating
Cross-Sector Correlation (Rolling 21D)
Declining = stocks re-dispersing to fundamentals
LPI: FRED (WALCL/WTREGEN/RRPONTSYD/SOFR/DFF) + FiscalData + Yahoo VIX
Tail table: Yahoo ^GSPC weekly
COT: CFTC Socrata TFF + Disaggregated
Funding: Binance (CoinGecko) + OKX API
Index GEX: homebrew SPY/QQQ (Yahoo option chains, BS gamma)
GEX: FlashAlpha free tier (NVDA/AAPL/AMD/MU, 5 req/day)