The rank is a relative momentum score and nothing else. For each
instrument the scanner measures total return over four windows — 1, 3, 6 and
12 months. Each window is scored against the other 53
instruments on the page (a percentile from 0 to 100), the four percentiles are
averaged with equal weight, and the table is sorted on that average. Rank 1 is
the strongest, rank 54 the weakest. No valuation,
fundamental, income or risk measure enters the score, and a rank improves
either because the instrument rose or because its peers fell.
Percentile columns, written p0 to p100, compare an instrument with its
own past — 504 sessions of its own history,
roughly 2 years — never with the other rows.
RV p85 means today is more volatile than 85% of that
instrument's own past two years.
Directional signals
Columns with documented predictive evidence behind them.
Read these first.
- Trend
- Where price sits against its 50- and 200-day averages.
Strong up = price above MA50 above MA200 with both averages
rising; Up = above MA200 but the alignment is incomplete;
Range = price between the two averages, or the two within 1% of
each other; Down and Strong down mirror the two upward
states.
- Rank
- The four-horizon momentum rank described above. 1 is strongest.
- ΔR 20D
- Change in that rank over the past 20 sessions, about one month.
↑ is climbing the table. A high rank that is falling and a middling rank
that is climbing are different situations.
- 12-1%
- The twelve-month return with the most recent month left out. The
skipped month is the standard convention: very recent moves tend to
reverse, so dropping it gives a cleaner read of the longer trend.
- vs 52W hi
- Distance below the highest close of the past year. 0.0 is a fresh
one-year closing high, −13.8 is 13.8% below it. The figure can never be
positive.
Risk signals
These describe how large the moves are, not which way
they go. A high reading is not bearish.
- RV pctl
- Realised volatility — 20 days of daily returns, annualised — ranked
against its own history. High means the instrument is moving more than
usual, which is a reason to size smaller rather than a direction.
- BBW pctl
- Bollinger bandwidth: how wide the 20-day price band is, ranked the
same way. Low RV and low BBW together is a squeeze — a
compressed range that often precedes a larger move of unknown
direction. Both must be low: a quiet fund with a wide band, or a wide
band on a quiet fund, is neither.
- Squeeze release
- The chip that replaces Squeeze when a compressed fund finally
moves: it is still inside the squeeze on today's percentiles, and today's
return exceeded 1.5 standard deviations of the 20 sessions before it. It
says the quiet period broke, not which way it broke or that it will
continue — the direction is in the price columns, and the release itself
is one session's evidence.
- ATR%
- The average daily trading range over 14 days as a percentage of
price. 3.5 means the instrument typically covers 3.5% of its price
between the day's high and low — a sizing and stop-distance input.
State & confirmation
Today's context. None of these carries standalone
predictive power; they confirm or contradict the directional zone.
- 1D%
- The last session's return.
- Vol ×20D
- The session's share volume against its own 20-day average. 1.0× is
normal; above 3× is flagged as an alert chip.
- RS 1M
- The past month's return minus SPY's over
the same month, in percentage points; positive means it outpaced US
equities. For bonds and commodities read it as distance from equity
beta, not leadership within the group. Click the header to switch to the
raw one-month return.
- Dev 200D
- How far price sits above or below its 200-day average. A large
positive number means price is extended; near zero means the long-term
average is close by.
ETF layer
Properties of the fund wrapper, not of the market it
holds.
- P/D%
- Premium or discount of the market price to the fund's stated net
asset value. Indicative only: the NAV feed carries no date, so a reading
cannot be aligned to its close and raises no alert — see the footer.
- 5D flow
- Change in shares outstanding over five sessions, a proxy for money
entering or leaving the fund. Blank where the feed supplies no share
count, which is roughly half the panel.
Marks, colours and the strips above the table
- Green is above zero, red below. An amber cell is an unusual
risk reading — realised volatility above p80, or volume above 3× its
20-day average. Purple marks a squeeze.
- A grey row ending in ● is more than
3 sessions behind the rest of the panel:
read its figures as stale. A row inside that threshold but still
behind carries a T-1 marker beside the ticker —
its venue has not published a later close, so its returns stop a
session earlier than the rest of the table and the ranking is that far
from like-for-like.
- ^ means the percentile used a shorter window
than the full one, so it is not strictly comparable with its neighbours.
A superscript P means the row was scanned through
a trading proxy rather than the line the sleeve holds — hover for
which.
- Overlays are the deployed risk state, read from the strategy
rather than recomputed here: the breadth gate (RISK_ON / RISK_OFF on the
share of S&P 500 constituents above trend), the emerging-market tilt,
and SPY's own one-month return as the anchor
for the RS column.
- Alerts are events on the latest session: 52-week highs and
lows, MA200 crossings, daily moves beyond 2 standard deviations, volume
above 3× normal, squeezes and their releases, RSI at or beyond
75 / 25, and rows crossing into or out of the top 10 of the
ranking. Chips name the fund and carry its ticker in the tooltip, and
every alert is shown — Show fewer collapses a
noisy day to the 12 highest-priority chips.
Click any chip to jump to that instrument's row
and open its chart. If your sleeve filter is hiding that row, the page
says which sleeve the instrument sits in and leaves the filter alone
until you ask for it.
- A rank crossing is a statement about the table, not a
judgement about the fund: the row is inside the top 10 today and was
outside it on each of the 5 sessions before, or the mirror. That
confirmation window stops a row hovering at the boundary from firing
every other day. Because the cut is fixed, a row entering usually
displaces one leaving — and a rank can improve because peers fell, so
read the crossing with 12-1% and Trend, which are absolute.
- A sigma move is one session, not a run: the last session's
return against the standard deviation of the 20 daily returns before it.
IBIT 3.3-sigma move +6.2% reads as "+6.2% in a
day, 3.3 times its own recent daily variation". The window stops at the
previous session, so a large move cannot inflate the yardstick it is
measured against.
- Every column header sorts, and the sleeve buttons filter to
A sectors, B asset classes, C thematics, D Europe or the overlay.
Reading the page in 30 seconds
- Read the pills top right: the price date, whether the panel is on
mixed dates, and the deployed gate state.
- Scan the alert strip — that is what actually changed today.
- The table opens sorted by rank. The top rows are the strongest
momentum across all four horizons at once.
- Read ΔR beside the rank: climbing ranks are emerging leadership,
falling ranks are leadership fading.
- Check the risk zone before treating a strong rank as an entry. A p90
realised volatility means the move is already large by that instrument's
own standards.
- Filter by sleeve to compare like with like — a sector against
sectors, not against gold.
- Click any row for its price chart with MA50 and MA200.
What this page is not
- Monitoring, not a strategy: no buy list, no position sizes, no
transaction costs, and no backtest behind any threshold shown.
- Every parameter is an unvalidated industry default. None has been
tested on this universe, so no number here is a calibrated
threshold.
- Ranks are relative. Rank 1 in a falling market is still losing
money — read it alongside 12-1% and Trend, which are absolute.
- A momentum rank says nothing about what an instrument holds, what it
costs, or whether it suits a portfolio.