RiskUpdated daily after close · as of 2026-09-04
Momentum Churn: How Violently the Market Is Rotating Inside a Quiet Index
We build a momentum long-short factor from our own ~4.1k-stock universe — winners minus losers, rebalanced monthly — and compare its 15-session realized volatility to SPY's. A high multiple means maximum churn under minimum index movement: last year's winners being dumped and last year's losers squeezed, with the violence cancelling out at the index level where nobody sees it.
Today's reading
As of 2026-09-04, the momentum factor's volatility runs 1.59% per day against SPY's 0.54% — a ratio of 2.96x, the 91th percentile of every reading since 2011 (1.3 standard deviations above the mean). State: Churning. The current winner and loser baskets (406 stocks each) were formed at the 2026-08-31 rebalance; today's long-short swing was +0.91%.
Sources, methodology & freshnessLast updated 2026-09-04 · Open ↓Close ↑
Source
Computed from our own ~5,500-symbol daily price database (TradeStation), common stocks only via the Polygon share-class snapshot, with SPY as the index leg.
Methodology
Month-end rebalance; 12-1 momentum ranks; equal-weight top/bottom deciles (price ≥ $3, ~400 names each); daily long-short return; 15-session population volatility of factor and SPY; ratio, with regime states at fixed percentiles (50/80/95) of the full 2011+ history. Forward returns are overlapping-window averages.
Updates
Recomputed after every US trading close as part of the daily pipeline.Last: 2026-09-04
Maintained & reviewed by Yuriy Matso — methodology shown on the page.
Momentum churnas of 2026-09-04
2.96x
Churning — 91th percentile since 2011
Factor vol
1.59%
SPY vol
0.54%
L/S today
+0.91%
Both volatilities are % per day over 15 sessions. Regime bands sit at fixed percentiles (50/80/95) of the ratio's own history: 1.4x / 2.1x / 3.4x.
01
The churn ratio — factor vol ÷ index vol since 2011
The gauge itself. A normal, two-sided market runs around 1.4x — the factor is always swingier than the index it nets out of. Spikes are rotations: December 2015, May 2018, the 2021-22 momentum unwinds. One honest mechanical note: the ratio can also stretch when the denominator collapses — a frozen index flatters the multiple — which is why the next chart shows the factor leg on its own.
Range:
02
The factor leg alone — long-short volatility, % per day
The numerator, denominated in daily percent: how much the winners-minus-losers spread is actually moving. This is the chart that separates “the index went quiet” from “the rotation got violent” — when both this and the ratio are extreme at once, the churn is real and the denominator is not the cause.
Range:
03
What happened next — SPY after each churn regime
The honest test, run before shipping. The pattern is monotonic and the edge lives in the calm corner: the quietest factor regimes preceded the strongest forward returns, while high churn preceded modestly below-baseline — but still positive — returns. A drag, not a cliff. If we had found a crash signal, this table would show it; it doesn't.
| Ratio that day | Days | Next 5 sessions | Next 21 sessions | Next 63 sessions |
|---|---|---|---|---|
| Lowest quartile (calm factor) | 1,001 | +0.38% · 63% win | +1.76% · 72% win | +5.07% · 82% win |
| Middle half | 1,931 | +0.25% · 62% win | +0.77% · 65% win | +2.40% · 73% win |
| Highest quartile (churn)Today | 996 | +0.12% · 55% win | +0.86% · 68% win | +2.42% · 73% win |
| All days | 3,928 | +0.25% · 60% win | +1.05% · 67% win | +3.10% · 75% win |
Average SPY price return and share of positive outcomes, 2011+. Overlapping forward windows — adjacent days share most of their path, so sample sizes are optimistic. Dividends excluded.
04
What this gauge cannot tell you
Three limits, stated plainly. It is not the academic factor: our series is equal-weight over our own universe from 2011 — levels are not comparable to cap-weighted 1972+ momentum series, which is why everything here is read in percentiles of its own history. The history carries survivorship: our universe prunes delisted names, so dead stocks exit the historical loser baskets; the volatility ratio is far less sensitive to this than factor returns would be, but it is not zero. It is direction-blind by design: a long-short spread says nothing about where the index goes — the base-rate table above is the only directional claim on this page, and it is a modest one. For where the rotation's damage lands, read it with the Hidden Bear Index and Bear Market Breadth.
Momentum churnas of 2026-09-04
2.96x
Churning — 91th percentile since 2011
Factor vol
1.59%
SPY vol
0.54%
L/S today
+0.91%
Both volatilities are % per day over 15 sessions. Regime bands sit at fixed percentiles (50/80/95) of the ratio's own history: 1.4x / 2.1x / 3.4x.
How Momentum Churn Works
- 1Build the momentum factorAt each month-end we rank every common stock in our universe (~4,000 names after filters) by 12-1 momentum — the return from 12 months ago to 1 month ago, skipping the latest month, the standard academic definition. The top decile are the winners, the bottom decile the losers.
- 2Track the long-short spread dailyEach day until the next rebalance, the factor return is the equal-weight winners' return minus the losers'. When winners get dumped and losers get squeezed — a rotation — this spread swings violently even if the index is flat.
- 3Compare the two volatilitiesThe gauge is the factor's 15-session realized volatility divided by SPY's. Around 1.4x is normal. High multiples mean maximum churn under minimum index movement: the market is being rearranged from the inside while the surface stays calm.
- 4Read it against its own historyRegime states (Quiet / Elevated / Churning / Extreme churn) are fixed percentiles of the ratio's own 2011+ history, and the page shows what SPY actually did after each regime — so a scary reading can be checked against base rates.
Who Uses Momentum Churn
Momentum traders
The direct measure of how hostile the tape is to momentum books right now — extreme churn is exactly the regime in which trend-following strategies bleed regardless of index direction.
Stock pickers
High churn means leadership is being repriced: last year's winners sold, last year's losers squeezed. Position turnover risk is highest precisely when the index looks calmest.
Risk managers
Factor volatility spiking against quiet index volatility is hidden portfolio risk — books hedged with index products are unhedged against the rotation itself.
Market watchers
The under-the-surface confirmation for divergence work: when breadth splits and vol spreads widen, this gauge shows whether actual positioning is being unwound or just drifting.
Pro Tips
01
Mind the denominator
The ratio can spike simply because index volatility collapses — the December 2015 records were as much about a frozen index as a wild factor. The page shows both components separately so you can see which side is moving.
02
Calm was the real signal
In our base rates, the QUIET quartile preceded the strongest forward returns (+5.1% per quarter, 82% win rate). High churn preceded modestly below-baseline returns: a drag on returns, never a cliff.
03
Rotation, not direction
The factor is long-short — it is deliberately blind to whether the market goes up or down. It measures who is being bought versus sold inside the market. Read it with the breadth family for the full rotation picture.
04
Watch it around rebalances
The factor re-forms at each month-end. A churn spike that persists across a rebalance means the rotation is chasing the new leadership too — a deeper repricing than a single squeeze.
Common Issues & Solutions
How is this different from the academic momentum factor (1972+)?▾
Same construction idea, different universe and weighting: ours is equal-weight over our own ~4,000-stock database from 2011, while the academic factor is cap-weighted over the CRSP universe from 1972. Levels are not comparable across constructions — the percentile against this series' own history is the honest read.
Is the history survivorship-biased?▾
Partly, and we say so: our universe prunes delisted symbols, so stocks that died leave the historical loser baskets. That flatters the factor's long-run LEVEL, but the 15-session volatility ratio — the thing this tool actually measures — is much less sensitive to it.
Why equal-weight?▾
We do not have reliable historical market caps for the full universe, and equal weighting inside deciles is a standard, transparent choice. It makes our factor swingier than a cap-weighted version — another reason to read percentiles, not absolute multiples.
The ratio is extreme — should I sell?▾
The base rates on this page say no, not on this signal alone: extreme churn preceded modestly below-average but still positive forward returns. It is a regime gauge — it tells you WHAT KIND of market you are in, not where it is going.
Frequently Asked Questions
What is momentum churn?▾
The realized volatility of a momentum long-short portfolio (winners minus losers, rebalanced monthly) expressed as a multiple of the index's own volatility. High readings mean the market is rotating violently beneath a calm index — winners being dumped, losers being squeezed, with the moves cancelling at the index level.
What does an extreme reading mean?▾
That leadership is being repriced at near-record speed while the index hides it. In our 2011+ history, extreme readings preceded modestly below-baseline forward returns (a drag on returns, never a crash signal) and marked regimes hostile to momentum and trend-following strategies.
How is the momentum factor built?▾
Each month-end we rank our common-stock universe by 12-1 momentum (12-month return skipping the latest month), form equal-weight top and bottom deciles (~400 stocks each), and track the daily long-short return until the next rebalance. The gauge is that series' 15-session volatility divided by SPY's.
Why does the history start in 2011?▾
Our daily price database begins in 2010 and the first rebalance needs a full year of lookback. That is shorter than academic factor series (1972+), which is why the page reads everything in percentiles of its own history rather than comparing levels across datasets.
Is high churn bearish?▾
Not reliably — our forward-return table shows the churn quartile preceded +2.3% average 3-month SPY returns versus +3.1% baseline. The cleaner signal in the data is the opposite corner: the calmest factor regimes preceded the strongest returns.
Explore Other Tools
Volatility Premium
The implied-vol cousin: single-stock vol vs index vol, from option prices
Tech Volatility Spread
VXN minus VIX — how much extra risk the market prices into the Nasdaq
Momentum Scanner
The single-stock view of the same universe this factor is built from
Hidden Bear Index
The average stock vs the index — where rotation damage shows up
Last updated: 2026-09-04