Volatility is most unusual in Nasdaq-100, at the 96th 2010.
How violently are markets moving?
Prices are Yahoo Finance data, shown with their attribution. Cboe VIX, OVX and ICE MOVE are not used here yet, so no implied volatility is shown.
In brief
How far each market moves day to day against its own history, and whether the moves are spreading- Against its own history since 2010, volatility is most unusual in Nasdaq-100 (96th percentile); the calmest is USD/JPY.→ Exhibit 1
- S&P 500 is 23.2% on the 20-day window, the 89th 2010.→ Exhibit 2
- The 10-year yield moves 105bp a year on a 20-day basis and the 2-year 119bp.→ Exhibit 2
- Equity and 10-year yield volatilities have moved together with a correlation of +0.39 over two years of weekly changes (73rd percentile).→ Exhibit 6
- At every horizon from two weeks to a year, S&P 500 volatility sits between the 43rd and 95th percentiles since 2010.→ Exhibit 4
| What would change this view | Trigger | Today | Decided by |
|---|---|---|---|
| Equity volatility calms | S&P 500 20-day volatility < 12.1% (2015–19 mean) | 23.2% (16 Nov 2026)not crossed | each session's close (next 18 Nov 2026) |
| The next decision week departs from the study | UVXY, 8 Dec 2026 to 11 Dec 2026 close, outside −11.4% to +15.2% (10th–90th of 92 weeks) | −3.9% (the 16 Sep 2026 week)pending | FOMC decision, 9 Dec 2026; the week's close 11 Dec 2026 |
The reasoning behind each point
- Cross-sectionAgainst its own history since 2010, volatility is most unusual in Nasdaq-100 (96th percentile); the calmest is USD/JPY.
Next most unusual: Brent crude (95th percentile); the calmest, USD/JPY, sits at the 16th. Crude 50% (WTI, USO) and 55% (Brent, BNO); gold 28.1%; the broad dollar 3.4% (H.10). Volatility of volatility is highest against its own history in USD/JPY (92nd percentile).
- EquitiesS&P 500 realised volatility is 23.2% on the 20-day window, the 89th percentile since 2010.
Over 63 days it is 15.1%. 20-day over 63-day ratio 1.53 (rising short-term volatility). Nasdaq-100 36.5%, Russell 2000 25.9%, energy 31.4% (83rd percentile).
- RatesThe 10-year yield moves 105bp a year on a 20-day basis and the 2-year 119bp.
Percentiles since 2010: 10-year 82nd, 2-year 91st. 63-day: 10-year 86bp, 2-year 98bp. Front-end volatility above long-end volatility is the signature of a policy repricing. Long Treasuries (TLT) 12.2%.
- Co-movementEquity and 10-year yield volatilities have moved together with a correlation of +0.39 over two years of weekly changes (73rd percentile).
Equity vs crude volatility +0.40, equity vs gold +0.32, equity vs euro +0.46. When volatilities rise together, diversification across markets fails at the moment it is needed.
- HorizonsAt every horizon from two weeks to a year, S&P 500 volatility sits between the 43rd and 95th percentiles since 2010.
Each horizon's reading against every window of the same length since 2010: 10 sessions 31.0% (95th, median 11.9%); 20 sessions 23.2% (89th, median 12.3%); 63 sessions 15.1% (66th, median 12.9%); 126 sessions 14.5% (60th, median 13.4%); 252 sessions 13.8% (43rd, median 15.0%). The widest short-horizon reading against its own cone is Nasdaq-100 at 10 sessions (98th percentile). Observed windows, not a forecast.
- Decision-week testThe study's tests for the 16 Sep 2026 hike week both held: UVXY closed 17.44 on Thursday (held) and 17.53 on Friday (held).
UVXY closed Thursday 17 Sep 2026 at 17.44, −4.4% from Tuesday's 18.23, less than 5.4% below it (17.27) — held; Friday 18 Sep 2026 at 17.53, not above the 18.36 top of the hike analogues' middle half — held. Held: the hike analogues remain the right reference. Held: a decay week, not a volatility event. The study, written on 16 Sep 2026 with data to the 15 Sep 2026 close, stated both in advance; they are evaluated here from the daily closes the page holds, to 16 Nov 2026. The core finding of the study reproduces from daily closes: an FOMC week costs UVXY a median −1.3% (study −1.2%) against −2.6% in ordinary weeks here (−3.0% in the study). The Board's IORB moved +25bp to 3.90% effective 17 Sep 2026, i.e. the Committee hiked on 16 Sep 2026. The study's hike analogues had put that Friday's close at 17.51 (middle half 17.19–18.36); it closed at 17.53, and on 16 Nov 2026 UVXY was at 20.45, +11.5% from Tuesday's 18.23. The study's intraday and surprise results are quoted, not re-derived.
Next: Positioning — who is exposed to the swings
Where each market's volatility sits in its own history
How far is each market moving day to day, and where does that sit in its own history since 2010? Select a row in the table to chart itExhibit 1 · Each market's 20-day volatility ranked against its own history
About this exhibit
Exhibit 2 · Realised volatility level and change for each market, by window
| Series | Level | 1w | 1m | 3m | 1y | Pct since 2010 | Level, recent | As of |
|---|---|---|---|---|---|---|---|---|
| Equities · 20-day | ||||||||
| S&P 500 (SPY) | 23.2% | +0.9% | +12.7% | +9.6% | +7.8% | 89 | 16 Nov 2026 | |
| Nasdaq-100 (QQQ) | 36.5% | +0.7% | +18.0% | +12.8% | +14.7% | 96 | 16 Nov 2026 | |
| Russell 2000 (IWM) | 25.9% | −0.5% | +11.2% | +11.1% | +2.0% | 84 | 16 Nov 2026 | |
| Developed ex-US (EFA) | 20.4% | +0.4% | +8.0% | +6.5% | +6.0% | 80 | 16 Nov 2026 | |
| Emerging markets (EEM) | 34.3% | +2.4% | +13.3% | +7.7% | +19.8% | 94 | 16 Nov 2026 | |
| Energy (XLE) | 31.4% | −0.5% | +9.7% | +6.5% | +14.8% | 83 | 16 Nov 2026 | |
| Rates · 20-day | ||||||||
| 2-year Treasury yieldbp a year | 119bp | +30bp | +22bp | +54bp | +53bp | 91 | 16 Nov 2026 | |
| 10-year Treasury yieldbp a year | 105bp | +17bp | +15bp | +30bp | +44bp | 82 | 16 Nov 2026 | |
| Treasuries 20y+ (TLT) | 12.2% | +1.5% | +0.8% | +0.5% | +4.3% | 46 | 16 Nov 2026 | |
| Credit · 20-day | ||||||||
| High yield (HYG) | 4.8% | +0.1% | +0.4% | +1.8% | +1.5% | 44 | 16 Nov 2026 | |
| Commodities · 20-day | ||||||||
| WTI crude (USO) | 50.5% | +0.3% | +11.0% | −14.4% | +25.9% | 91 | 16 Nov 2026 | |
| Brent crude (BNO) | 55.2% | +1.8% | +17.4% | −6.8% | +32.3% | 95 | 16 Nov 2026 | |
About this exhibit
Exhibit 3 · S&P 500 (SPY): realised volatility
The same reading at every horizon
Against its own windows since 2010, is today's reading unusual for the horizon it is measured over?Exhibit 4 · S&P 500 realised volatility by horizon, against its range since 2010
About this exhibit
How unstable, and how shared
The volatility of volatility, and whether volatilities rise togetherExhibit 5 · How unstable each market's volatility is, against its own median
| Market | Vol of vol | Median | Percentile | 20d / 63d |
|---|---|---|---|---|
| USD/JPY (H.10)FX | 186% | 110% | 92 | 0.70 |
| EUR/USD (H.10)FX | 136% | 97% | 88 | 1.01 |
| Brent crude (BNO)Commodities | 136% | 108% | 83 | 1.22 |
| S&P 500 (SPY)Equities | 140% | 111% | 77 | 1.53 |
| WTI crude (USO)Commodities | 121% | 98% | 77 | 1.15 |
| Energy (XLE)Equities | 113% | 97% | 76 | 1.30 |
| Nasdaq-100 (QQQ)Equities | 130% | 109% | 74 | 1.51 |
| Developed ex-US (EFA)Equities | 120% | 105% | 74 | 1.39 |
| Broad dollar (H.10)FX | 108% | 97% | 70 | 0.87 |
| Dollar bullish (UUP)FX | 101% | 94% | 66 | 1.03 |
| Russell 2000 (IWM)Equities | 107% | 99% | 64 | 1.44 |
| Emerging markets (EEM)Equities | 106% | 99% | 62 | 1.36 |
| BitcoinCrypto | 146% | 135% | 59 | 1.22 |
| 2-year Treasury yieldRates | 91% | 99% | 34 | 1.21 |
| Treasuries 20y+ (TLT)Rates | 76% | 87% | 23 | 1.09 |
| Gold (GLD)Commodities | 94% | 107% | 20 | 1.12 |
| High yield (HYG)Credit | 87% | 112% | 18 | 1.17 |
| 10-year Treasury yieldRates | 68% | 88% | 7 | 1.21 |
| Silver (SLV)Commodities | 81% | 110% | 4 | 1.08 |
About this exhibit
Exhibit 6 · Correlation of volatility between pairs of markets, rolling two years
About this exhibit
Event study: FOMC weeks
What a decision week does to a volatility product — the UVXY study of 16 September, reproduced where daily closes allow and quoted where they do not
Exhibit 7 · Volatility fund UVXY's return in each FOMC week, by Fed action
About this exhibit
Exhibit 8 · UVXY daily closes in the decision week, against preset thresholds
About this exhibit
Exhibit 9 · UVXY FOMC-week study results as published and re-derived from daily closes
| Result | Study | Here | Status |
|---|---|---|---|
| Scheduled decisions, Jan 2015 – Jul 2026 | 92 | 92 | check |
| Median, T−1 close to week end, all meetings | −1.2% | −1.3% | check |
| … hikes (n) | −1.0% (20) | −0.9% (20) | check |
| … holds (n) | −1.1% (63) | −1.4% (63) | check |
| … cuts (n) | −5.6% (9) | −5.4% (9) | check |
| … projections meetings | −1.8% | −1.8% (45) | check |
| Median, ordinary weeks (n) | −3.0% (477) | −2.6% (505) | shown |
| Mann–Whitney p, FOMC vs ordinary | 0.08 | 0.15 | shown |
| Abnormal median vs term-structure-matched weeks (90% CI) | +1.5% (+0.4 to +2.6) | – | quoted |
| 14:00 five-minute bar, multiple of an ordinary week | 3.2× | – | quoted |
| Rank correlation, surprise vs statement window / week end | 0.51 / 0.17 | – | quoted |
| Week end, most hawkish / most dovish third of surprises | +0.8% / −4.3% | – | quoted |
| Hike analogues: Friday close, median (middle half; 10th–90th) | 17.51 (17.19–18.36; 14.37–20.11) | – | quoted |
About this exhibit
Exhibit 10 · The conclusions of the UVXY FOMC-week study
- FindingAn FOMC week is a milder decay week.
Median −1.2% against −3.0% for ordinary weeks; the statement bar carries the event premium, released at 14:00.
- FindingThe surprise, not the action, sets the sign.
Hikes traded like holds; the most hawkish third of surprises ended the week +0.8%, the most dovish −4.3%.
- ReadHedges are regime insurance, not a meeting trade.
The study treats long volatility as insurance against oil, the term premium and the November calendar carried by October–November VIX futures.
About this exhibit
Appendix
How each measure is calculated, why the indexes are absent, and what the study can and cannot reproduce28 checks, all pass
Realised volatility — how far each market actually moves day to day — against its own history since 2010, for every horizon from two weeks to a year; how unstable volatility is, and whether volatilities rise together. Implied volatility and skew join when the options data are added; the UVXY study is the site's own view of what an FOMC week does to volatility products.
Exhibit 11 · Definitions
- RealisedFrom daily returns.
Realised volatility = √252 × sample standard deviation of the last 20 or 63 daily log returns, in percent a year. For Treasury yields it is the standard deviation of daily changes in basis points × √252 (bp a year, normal volatility). ETF returns are on NYSE sessions, Treasury yields on Treasury business days and H.10 rates on the Board's business days; bitcoin trades every day but its returns are taken on its own calendar.
- Vol of volInstability of volatility.
Volatility of volatility = √252 × standard deviation of daily log changes of the 20-day realised volatility over the last 63 days, in percent a year. Consecutive 20-day windows overlap by 19 days, so the measure is smoother than the volatility of an implied index; compare it with its own history (percentile), not with VVIX.
- Co-movementCorrelation of volatilities.
Correlation of volatilities = correlation of weekly log changes in 20-day realised volatility over 104 weeks. High values mean shocks raise volatility in both markets together.
- ConeRealised volatility by horizon.
The realised-volatility cone: for each horizon (10, 20, 63, 126, 252 sessions), realised volatility over every window of that length ending on a session since 4 Jan 2010 — its minimum, 10th, 25th, 50th, 75th and 90th percentiles and maximum — with today's reading and its percentile among them. The windows overlap, so the count of independent windows at 252 sessions is about sixteen, not the thousands of windows counted. Observed history, not a forecast: it says whether today's volatility is high for the horizon it is measured over.
- ImpliedNot used here yet.
Implied-volatility indexes (Cboe VIX, VIX3M, VVIX, OVX, GVZ; ICE BofA MOVE) and option skew are not yet read by the site, so the page uses only volatility computed from prices and published yields; what protection costs joins when the options data do. VIX-futures ETPs (UVXY) are traded products whose prices are shown with their attribution.
- UVXY studyReconstructed and quoted.
The UVXY study (the site's own research, 16 Sep 2026) follows each scheduled FOMC decision from the close before the decision day to the last session of that week, with pre-28 Feb 2018 log returns scaled by 0.75 for the change from 2× to 1.5× leverage. Reconstructed here from Yahoo daily closes and the Board's calendars and IORB: meeting counts by action, medians, share lower. Not reconstructible without paid data and quoted from the study: the five-minute statement-window results, the surprise correlations (Bauer–Swanson and a two-year-yield proxy), the term-structure-matched abnormal return and its bootstrap interval, the analogue scores and cone. The ordinary-week control here has 505 weeks against the study's 477 (the study's exclusions are not specified), so its median and the Mann–Whitney p differ and are shown, not checked.
- LicenceVendor prices; official yields and FX.
Prices are Yahoo Finance data, shown with their attribution; Treasury, H.10, policy-rate and FOMC calendar data are free to use. Realised measures and the reconstruction are computed here.
Sources
- licensedYahoo Finance chart API, daily adjusted closes (SPY, QQQ, IWM, EFA, EEM, XLE, TLT, HYG, USO, BNO, GLD, SLV, UUP, BTC-USD, UVXY)
Prices: Yahoo Finance. · https://query2.finance.yahoo.com/v8/finance/chart/
- publicU.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates (2- and 10-year)
Source: U.S. Department of the Treasury · https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?type=daily_treasury_yield_curve
- publicBoard of Governors of the Federal Reserve System: H.10 (broad dollar, EUR, JPY), policy rates (IOER/IORB), FOMC meeting calendars
Source: Board of Governors of the Federal Reserve System · https://www.federalreserve.gov/
- computed hereThe site's own study: UVXY through the decision week (92 FOMC meetings, 2015–2026), project note UVXY_FOMC_EVENT_STUDY_2026-09-16.md — quoted results
The site's own research, 16 Sep 2026; intraday bars and Cboe indexes behind it are paid data and not shown.
Pages this one copies from
| Page | Title | Updated | Data to | Arrays | Values | Source checks |
|---|---|---|---|---|---|---|
| Policy rates | Policy rates and money markets | 2026-11-17 | 13 Nov 2026 | 1 | 0 | 19 pass |
Reproduction checks
| Check | Published | Computed | Tolerance | Result |
|---|---|---|---|---|
| Realised volatility = √252 × sample s.d. of the last 20 or 63 daily log returns (bp changes for yields), recomputed with numpy (2280 asset × window × date cells; max abs difference) | 0 | 0 | 0 | pass |
| Annualisation identity: σ²_annual ÷ 252 equals the daily sample variance of the S&P 500 ETF's last 20 returns (%², absolute difference) | 2.13 | 2.13 | 0 | pass |
| Volatility cone: today's 20- and 63-session readings are the page's own realised volatilities (19 markets; max abs difference) | 0 | 0 | 0 | pass |
| Volatility cone: min ≤ 10th ≤ 25th ≤ median ≤ 75th ≤ 90th ≤ max at every horizon (5) and market (violations) | 0 | 0 | 0 | pass |
| Volatility cone: the S&P 500's 63-session median recomputed window by window with numpy (4,244 windows since 2010; abs difference) | 12.923 | 12.923 | 0 | pass |
| Sanity: 63-day bp volatility of the 10-year yield (Treasury) and 63-day volatility of TLT (licensed price) correlate at least 0.8 since 2010 | 0.8 | 0.854 | 0 | pass |
| Sanity: 63-day volatility of EUR/USD (H.10) and of UUP (licensed price) correlate at least 0.8 since 2010 | 0.8 | 0.85 | 0 | pass |
| Study reconstruction: scheduled FOMC decisions Jan 2015 – Jul 2026 from the Board's meeting calendars = 92 (count) | 92 | 92 | 0 | pass |
| … of which hikes 20, holds 63, cuts 9, classified from the Board's IOER/IORB change (differences in counts) | 0 | 0 | 0 | pass |
| … median UVXY log return, T−1 close to week end, all meetings: study −1.2% (daily closes vs the study's last 5-minute bar) | -1.2 | -1.34 | 0.35 | pass |
| … medians by action (hikes −1.0, holds −1.1, cuts −5.6) and at projections meetings (−1.8), max abs difference, pp | 0 | 0.283 | 0.35 | pass |
| … share of weeks ending lower: hikes 65%, holds 56% as printed (max abs difference, pp; the study rounds to whole percent) | 0 | 0.444 | 0.51 | pass |
| … no ordinary (control) week contains a scheduled or unscheduled decision (overlaps) | 0 | 0 | 0 | pass |
| UVXY Yahoo daily close on 11 Sep 2026 as quoted in the study (18.02) | 18.02 | 18.02 | 0.005 | pass |
| UVXY Yahoo daily close on 15 Sep 2026 as quoted in the study (18.23) | 18.23 | 18.23 | 0.005 | pass |
| IORB move in basis points: the published bp equals (after − before) × 100 of the published levels (abs difference, bp) | 25 | 25 | 0 | pass |
| IORB move effective date: the Board's series here and rates.policy's copy date the last change the same day (2026-09-17; days apart) | 0 | 0 | 0 | pass |
| IORB after the move equals rates.policy's level on the effective date (3.90%) | 3.9 | 3.9 | 0.005 | pass |
| Dated tests: the Thursday threshold is the study's 5.4% below the 15 Sep 2026 close (18.23 × exp(-0.054) = 17.27) (identity) | 17.272 | 17.272 | 0 | pass |
| Dated tests: every observed close is UVXY's Yahoo close on the test's date (2 evaluated; max abs difference) | 0 | 0 | 0 | pass |
| FOMC calendar: every decision with a statement since 2021 (47, the last 28 Oct 2026) is the last day of a meeting on the Board's calendar page as parsed for the next decision week (decisions not found) | 0 | 0 | 0 | pass |
| Falsifier: the next decision on the release calendar (2026-12-09) is the one the Board's meeting page gives (2026-12-09) (days apart) | 0 | 0 | 0 | pass |
| Falsifier: the next session is on the business-day calendar fetched on or before 2026-11-17 (1 = found) | 1 | 1 | 0 | pass |
| Falsifier: every date it names is after the panel's last close (16 Nov 2026) — a dated test whose date has passed is evaluated in the read instead (dates named on or before it: none) | 0 | 0 | 0 | pass |
| Dated tests: every test dated on or before 16 Nov 2026 is evaluated and its outcome printed in the read (2 evaluated; tests not printed) | 0 | 0 | 0 | pass |
| The Policy rates page: the source page is published with all its checks passing (dated 17 Nov 2026, 19 checks) | 1 | 1 | 0 | pass |
| The Policy rates page: the site still serves the page this one copied (dated 17 Nov 2026) | 1 | 1 | 0 | pass |
| The Policy rates page: every copied observation equals the source page, matched by date (86 values in 1 arrays; mismatches) | 0 | 0 | 0 | pass |
- Yahoo Finance chart API, daily adjusted closes (SPY, QQQ, IWM, EFA, EEM, XLE, TLT, HYG, USO, BNO, GLD, SLV, UUP, BTC-USD, UVXY) — Prices: Yahoo Finance.
- U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates (2- and 10-year) — Source: U.S. Department of the Treasury
- Board of Governors of the Federal Reserve System: H.10 (broad dollar, EUR, JPY), policy rates (IOER/IORB), FOMC meeting calendars — Source: Board of Governors of the Federal Reserve System
- The site's own study: UVXY through the decision week (92 FOMC meetings, 2015–2026), project note UVXY_FOMC_EVENT_STUDY_2026-09-16.md — quoted results — The site's own research, 16 Sep 2026; intraday bars and Cboe indexes behind it are paid data and not shown.