Every screen on this desk is an argument, not a number. Here is the argument — the four things we believe about markets, and how each instrument was built to express them. Read this once and every chip, band and percentile on the platform stops being decoration and starts being a case you can make to a client.
We don’t forecast — we measure. No one knows what the market does next, but we do know, with real numbers, what has followed thousands of times from a given setup. So every read here is a base rate and a distribution, never a call. “When a name has been this stretched, the next 21 days ran +7% two times out of three” is a defensible sentence. “This goes higher” is not. We only make the defensible kind.
“70 is overbought,” says the textbook. Overbought for what? A utility and a semiconductor don’t share a nervous system. Every reading here is scored as a percentile of that specific name’s own five-year distribution — so a 67 that’s the 96th percentile for it is a genuine extreme, and a 67 that’s the 40th is noise. We threw out the universal thresholds. They flatter the eye and lie to the portfolio.
An index can rise while it rots underneath — a handful of names carrying a market that’s quietly narrowing. So we start under the surface: what share of stocks sit above their own trend, across four horizons at once. When near-term participation runs ahead of the primary trend, the advance is broadening — the healthiest tape there is. When it rolls beneath, the index is standing on fewer and fewer shoulders. That’s the pre-wobble tell, and you only see it if you read breadth before you read price.
The middle of a distribution is a coin flip, and we don’t trade coin flips. The information lives at the tails: washed-out enough that reversion has historically paid, or stretched enough that chasing has historically hurt. Every tool here is built to find those edges — and, just as important, to say plainly when nothing is at one. “Nothing actionable today” is a real answer. Most desks can’t bring themselves to give it.
How far has this name run from its own trend, and the last hundred times it was here, what followed?
Distance from a moving average is old news. Pairing it with the forward-return distribution conditioned on that exact level of stretch is the part that matters. We don’t just show you the rubber band — we show you what happened the prior times it was pulled this tight.
The line is % above/below the 65-day trend; the shaded bands are ±2σ of its own history; the dot is today. The snapback table beneath is the payoff — forward 5/10/21-day median, win rate and range, computed only from days it was as stretched as it is now.
a timing trigger. A name can stay stretched for months. It’s a base rate, not a buy button.
Which sectors and sub-sectors are washed-out (potential adds) vs extended (crowded) — and are they turning this week?
The same stretch lens as the names, aimed at the 11 sectors, sub-sectors, factors and fixed income — because that’s where an advisor actually tilts. The 1-week return on every tile tells you whether a washed-out sector is already starting to turn.
Percentile of each ETF’s distance from trend — green ≤ 25th (washed-out), red ≥ 75th (extended). The “1w” figure is the trailing five-day return, color-coded. Click any tile for its rolling-return history.
a buy/sell list. It’s where the odds and the crowding sit; sizing and timing stay yours.
Is this name winning or losing versus its peers — and is that relative move at an extreme?
Absolute RSI tells you if a price is stretched. We care about relative strength, so we run Wilder’s RSI on the ratio of a name’s cumulative return to a leave-one-out basket of its peers (cap-weighted for sectors, equal-weighted for sub-sectors) — then percentile it against its own history. A 67 RSI is nothing; a 67 at the 96th percentile of relative strength is a crowded leadership trade.
RSI at 21 and 65 days on the relative line, each with its own percentile. High = outperforming and stretched; low = lagging and washed-out.
a price call. A name’s relative line can be at an extreme while its price does very little.
Is a sector leading, improving, weakening or lagging the S&P — and which direction is it headed?
A snapshot of relative strength misses direction. RRG plots relative strength against its own momentum, with a tail, so you read rotation as motion — what’s rotating in before it’s obviously strong, and rolling over before it’s obviously weak.
X = relative strength vs the S&P (100 = on par); Y = the momentum of that strength. Leading (top-right), Improving (top-left, rotating in), Weakening (bottom-right, rolling over), Lagging (bottom-left). The tail is the path it took to get there.
precise timing. Quadrants can whipsaw near the center. It’s a map of flow, not an entry.
If I had one screen before a client call, what do I need to know?
Everything above, synthesized. Adds vs trims, the rotation map, and the concentration read in one place — because the point of rigor is a fast, confident answer, not more tabs to open.
Left: where the odds favor adding vs trimming. Right: the rotation picture. Top: the one-line house read of the tape.
a portfolio. It’s a disciplined starting point for judgment, not a substitute for it.
A client asked something specific — what does the data actually say?
Every entry is an event study: define a trigger, measure the forward-return distribution across decades, report the base rate and the honest caveats. It’s the archive of ‘probability over prediction’ — and it’s where your bespoke, co-branded questions become permanent research.
Searchable by theme across 100+ studies. Each one: the trigger, forward returns, win rate, base rate, and the caveats spelled out.
a guarantee. Base rates are odds, not promises — and regimes change.
Is the tape healthy or hollow? The Board
Where’s the rubber band — name and sector? Stretch History · Sector Compass · Peer-Relative RSI
Which way is leadership moving? RRG
When we’ve been here before, what followed? Research Library