Which sectors lead and lag SPY, and what moved.
Everything on this page is ranked on price. The options-derived sector axis is still under validation and is not released for ranking.
As of Thu, Oct 15, 2026 close
Rank, 20d · how many places that sector has gained (▲) or lost (▼) in the RS-Ratio ranking over the last 20 sessions. A flat arrow (→) means no change; a dot (·) means it has not been tracked long enough to compare. This is the price relative-strength ranking, a DIFFERENT ranking from the row's own numbered position in this table, never positioning.
The bar · each sector's 20-session price relative strength trend, drawn from a centre line. To the right of the line that sector's relative strength has been improving; to the left it has been deteriorating. Bar length is relative to the widest move in today's field, so it compares sectors to each other on this day, not this day to other days.
What “relative strength improving” means · each sector's price is measured against SPY and scored against its own recent range; the bar shows the direction that score has been moving over 20 sessions. It is not the same as how much ground the sector gained on SPY. A sector can be improving here while still trailing SPY over the window, and the two do come apart.
There is no positioning column today. It carries one of four words (crowded, heavy, light or quiet) and appears once those readings can be ranked. Until then it would print the same placeholder on every row, so it is left out rather than filled with a repeated non-answer.
Positioning is not fund flows. It is measured from options activity and is deliberately independent of the day's price move.
Ranked by 20-session relative-strength trend, improving at the top. Not the same as return versus SPY.
Across (price) · each sector's strength against SPY, scored against its own recent range. The vertical line is that sector's own 65-session average, not parity with SPY: right of it the sector is stronger against SPY than it has recently been, left of it weaker. The axis is fixed at 100 ± 7.94 every day, so two days can be compared directly.
Up and down (options model) · a relative score derived from ETF options activity. It is adjusted for the day's price move and remains under validation. It does not measure which institutions bought or sold.
Weak price · heavy positioning · this is the corner this page exists to surface.
Positioning is not fund flows. No money is being counted in or out of a fund here.
Why there is one axis today · The options-model axis is under validation and is not ranked. Until then each mark sits on price relative strength alone.
No trails on this map: each session is plotted on its own.
Today's field spans 23% of the full scale.
Zoomed in · 98.5 to 102.1
What this is · one sector at a time, and the instrument it is measured through. Every sector is read through a single ETF, so the ETF is named rather than left implied.
Price relative strength · how this sector's price is doing against SPY, scored against its own recent range. A sector can rise and still read weak here, if SPY rose more.
Options-model axis · derived from ETF options activity and adjusted for the day's price move. It does not identify customer side or fund flows. It remains under validation.
Nothing selected, so this is the top of the ranking. Pick a sector to change it.
Consumer Staples ranks 1st of 19 on 20-session price relative strength trend, measured through the XLP ETF.
XLPrank 1 of 19sector
What this line is · Consumer Staples's relative strength, scored against its OWN trailing 65-session norm. 100 marks where a session sits exactly at that norm. The underlying ratio is priced against SPY, but this line never claims to beat it; every reading is relative to the sector's own recent history. The axis is fixed at the same [92, 108] range every day so two different sessions stay comparable.
Colour · follows the sign of the 20-session trend, the same green/red the leaderboard bar uses. It can disagree with the line's recent direction on the chart: the two are different quantities.
One continuous line · most of this window comes from a historical backfill, verified to reproduce live readings bit-for-bit. It is the same measurement throughout, so there is no seam to mark.
Consumer Staples has spent 12% of the last 120 sessions with relative strength above its own 65-session norm.
Every sector is ranked each session by RS-Ratio — relative strength scored against its own recent range, not its price against SPY. This region reports the largest changes in that ranking, plus any sector entering the top third after a long absence.
On a typical session about thirteen of the nineteen sectors move enough to qualify, so this is the top few by size of move, not everything that qualifies.
A move between thirds is only reported when it clears the boundary by more than a typical day's movement. Two thirds of boundary crossings sit inside a single session's noise and a third are back where they started the next day.
These are descriptions of what has already happened. Nothing here forecasts a return.
vs SPY · this sector's price return minus SPY's over the same two sessions, in percentage points. It is a DIFFERENCE between two returns, not a ratio of two prices — a ratio is a relative-strength level, which is the quantity the leaderboard above already ranks the trend of.
Price return, not total return · dividends are not reinvested. A sector that went ex-distribution inside the window reads low by roughly the payout, which on a high-yielding member can move it many places in the ranking.
Windows are trading sessions · 1W is 5 sessions, 1M is 21, 3M is 63. Year to date is anchored on the last session of the prior calendar year, not on 1 January.
Withheld numbers · a move larger than a market moves in one session is a split or a bad close, not a finding. Those windows show an em dash rather than a number, and say so on hover.
Closes Oct 8 to Oct 15
| # | Sector | vs SPY, 1 week | vs SPY | Return |
|---|---|---|---|---|
| 1 | Health CareXLV | +5.65 pts | +5.37% | |
| 2 | Consumer StaplesXLP | +4.18 pts | +3.91% | |
| 3 | MaterialsXLB | +2.89 pts | +2.62% | |
| 4 | Real EstateIYR | +2.78 pts | +2.50% | |
| 5 | China TechKWEB | +2.73 pts | +2.46% | |
| 15 | HomebuildersXHB | +0.35 pts | +0.07% | |
| 16 | SoftwareIGV | +0.29 pts | +0.01% | |
| 17 | IndustrialsXLI | -0.03 pts | -0.30% | |
| 18 | TechnologyXLK | -4.07 pts | -4.34% | |
| 19 | SemiconductorsSOXX | -8.35 pts | -8.62% |
Method · Contract-day options model balance and dark-pool dollars for each sector ETF on 2026-10-15. The signed options balance uses whole-day closing prices versus VWAP and weighted contract-day volume, not matched bid/ask prints or executed net premium. Customer side is unknown. ETF activity includes hedging and overlay positioning; an XLK print is not money moving into Apple. The model excludes some contracts using their last recorded trade condition; that does not classify every print in those contracts. Trades listed are the largest single trade in each ETF that session — not a record or a historical ranking.
VWAP-based options balance · a model-weighted difference between call and put contract-day buckets based on their closing option price versus whole-day VWAP. The sign is a model result, not a matched bid/ask trade or executed net premium. It does not establish who bought, sold, opened, or hedged; customer side is unknown.
Withheld balances · on some sessions too little of an ETF's premium meets the model assignment rule for a balance to mean anything. Those rows show an em dash rather than a confident figure, and say so on hover. A net smaller than $1.0M is also withheld: it is inside the gap between adjacent sectors, so ranking it would rank rounding.
Advanced columns · proxy-assigned share is the fraction of model-weighted contract-day notional assigned to a price-location bucket. Proxy concentration measures how uneven those assigned buckets are; it is not an accuracy probability or the share of individual trades with a verified side.
Dark-pool dollars · the off-exchange share of the session's tape multiplied by its dollar volume. It is an APPROXIMATION — the dollar volume is closing price times share volume, not a volume-weighted average, because no session VWAP is stored. "Heavier" and "lighter" mark a share at least 1.5 standard deviations from this ETF's own recent normal, in that direction specifically.
One session, no history · most of these ETFs have under three weeks of options readings, so there is deliberately no trend, no sparkline and no week-over-week comparison here. A thirteen-point line looks exactly like a thousand-point one.
| # | Sector | Options balance (model) | Balance | Dark pool (approx.) |
|---|---|---|---|---|
| 1 | SemiconductorsSOXX | +$40.2M | $3.41B | |
| 2 | Gold MinersGDX | +$28.7M | $607.9M | |
| 3 | FinancialsXLF | +$6.7M | $986.0Mheavier | |
| 4 | SoftwareIGV | +$4.2M | $360.0M | |
| 5 | EnergyXLE | -$3.6M | $636.8M | |
| 6 | China TechKWEB | +$3.5M | $117.6M | |
| 7 | Consumer StaplesXLP | -$3.5M | $682.8M | |
| 8 | Health CareXLV | -$2.0M | $606.8M |
Large single printsSOXX $231.4M at 03:34 PMIGV $56.1M at 10:23 AMXBI $82.5M at 12:06 PMXLB $48.3M at 03:32 PMKRE $29.6M at 11:13 AM
Options balance is a model estimate from closing prices against VWAP, not matched trades. Who bought or sold is unknown.
What this shows · one row per sector, one cell per trading session, coloured by that sector's rank against the rest of the field THAT day. Green is the strongest third, red is the weakest, and a quiet neutral marks the middle. It reuses the same relative-strength history as the trail above, just across every sector at once instead of one at a time.
Hatched cells · this sector had no reading that session. It is drawn distinctly from a weak (red) reading on purpose: an absence is not the same thing as a bottom-third finding.
Changes · how many times a sector's third flipped over the window shown. A sector near a rank boundary can flip on a small move, so this is a measure of how CONTESTED a sector's standing has been, not of how much its price moved.
Each row is one sector's rank against the field over the last 120 sessions, most recent on the right.
