Parabola Radar · a Synthos Research surface · July 25, 2026
The Parabola Radar
Technology is accelerating — and in markets, acceleration has a visible signature: the parabola, a price curve that bends upward. This page explains what parabolas actually are, why they are becoming more common, and how Synthos hunts them.
This is the explainer — the reasoning, from first principles, before the first call. The live radar board publishes here as signals accrue and get graded, and every radar call will be logged and graded in public, like everything else Synthos publishes. Educational research, not investment advice.
What a parabola actually is
A parabola is not a chart pattern, and spotting one is not chart-reading. It is the price signature of a specific physical situation: an exponential demand process colliding with supply that cannot respond. Demand compounds — every year more buyers need more of the thing than last year. Supply is inelastic — the fab, the mine, the power plant takes years to build, so for years the only variable free to adjust is price. Price does all the adjusting, and the curve bends.
Then a third force takes over: reflexivity. In most markets a rising price is self-correcting — it attracts supply and repels buyers. In a reflexive market a rising price improves the fundamentals themselves: capital gets cheaper, the best people join, customers treat the winner as the safe choice. Better fundamentals justify a higher price, which improves fundamentals again. The loop runs until something breaks it.
Full description: the left half of the diagram is a price-versus-time chart. A solid curve labeled in words “exponential demand — adoption compounding year over year” rises slowly at first and then increasingly steeply. A dashed, nearly vertical line labeled “inelastic supply — capacity takes years to respond” cuts across it. Where they meet, a marked point labeled “the collision — price does all the adjusting” shows that because supply cannot expand, price is the only variable free to move. The right half shows the amplifier, reflexivity, as a three-step feedback loop drawn with arrows: price rises, which brings cheaper capital, talent, and customer trust, which makes the fundamentals actually improve, which feeds back into price rising — each turn feeds the next.
This is not a rare curiosity. The same physics has produced many of the defining moves of recent years:
The five phases — and the only window that matters
Full description: a single price curve over time, divided into five labeled regions. In stealth the line is nearly flat — nobody is watching. In ignition it starts to bend upward. In acceleration it steepens sharply as the crowd arrives. In saturation the rise flattens near the top — everyone already owns it — and an annotation warns that arriving here is the fatal mistake. In blowoff the line goes briefly vertical, then breaks: a second line falls steeply, labeled “typical retrace: minus 60 to minus 80 percent.” A shaded band over the ignition and acceleration regions is labeled “the tradable window.”
- 1 · Stealth
- The constraint already exists but almost nobody is looking. Price drifts. The evidence lives in lead times, order books, and the things specialists say to each other — not in headlines.
- 2 · Ignition tradable window opens
- The curve starts to bend. Informed money notices the supply problem; the first earnings surprises land; the story is still contrarian. This is where being early is worth the most.
- 3 · Acceleration tradable window
- Steepening. The narrative goes mainstream, the reflexive loop engages, and the move compounds. Holding through the noise here is the hard, valuable skill.
- 4 · Saturation the fatal mistake is arriving now
- Everyone who can own it does. New supply is finally arriving; the marginal buyer is the least-informed one. The price may still rise — but the edge is gone and the exit is crowded.
- 5 · Blowoff
- Briefly vertical, then over. When parabolas end, they do not correct politely: across the storied peaks — 1929, gold 1980, Nasdaq 2000, uranium 2007, Bitcoin’s cycle tops — the typical outcome is a 60–80% retrace of the move. That number deserves respect, stated plainly, before the first position — not after.
The tradable window is ignition into acceleration. The fatal mistake — the one that ruins most people who chase these moves — is arriving at saturation, when the story is loudest and the remaining upside no longer pays for the retrace that follows.
The singularity lens
You do not need to predict AGI timelines to use this page. You only need the direction of one variable: technology adoption cycles are compressing. Railroads took roughly fifty years to saturate. Electricity took about thirty. The internet, twenty. Mobile, ten. AI is on pace for three to five. Whatever your view on where it ends, the direction of travel has been one-way for two centuries.
Full description: five S-shaped adoption curves side by side, each rising from ten percent to ninety percent adoption. The width of each curve is drawn to scale with how long that adoption took, so each successive curve is visibly narrower and steeper: railroads, about fifty years, roughly 1830 to 1880; electricity, about thirty years, roughly 1890 to 1920; the internet, about twenty years, roughly 1990 to 2010; mobile, about ten years, roughly 2007 to 2017; and AI, on pace for three to five years, 2023 onward. The calendar gaps between waves are compressed for legibility; the printed year ranges carry the real dates.
Run that direction through the parabola machinery and the lens makes a prediction. Faster adoption means demand curves bend harder. Physical supply chains — fabs, mines, grids — do not speed up nearly as much, so the demand–supply collisions get more violent. Under acceleration, the lens expects parabolas to become more frequent, larger, and faster. That is a thesis we test in public, not a law.
It also reframes a prior most investors inherited from a slower century. In a world where big moves are rare accidents, mean-reversion is the posture that fits: fade the extreme, expect the average to reassert. In a world where big moves are the recurring signature of compounding technology, the posture that fits is structural long-convexity: one in which the rare, huge outcome pays for many small misses — because the rare, huge outcome keeps happening.
Bottleneck migration — where the next one comes from
Parabolas do not appear at random; they appear where the constraint binds. And the constraint moves in a predictable way: solving one bottleneck funds the demand that exposes the next. Chips got built — and the binding constraint became memory. Memory capacity came — and it became power. Power contracts got signed — and it became grid equipment, then the materials that grid equipment is made of, and further down the stack, the robots and biotech tooling that all of this eventually pays for. The migration is predictable in sequence; the lag at each layer is where the opportunity lives.
Full description: six boxes stacked vertically, top to bottom: compute chips (GPUs and accelerators); memory, meaning high-bandwidth memory; datacenter power; grid equipment such as transformers and switchgear; materials such as copper; and robotics and bio. A downward arrow beside the stack is labeled: the constraint migrates down the stack — each solved layer funds the demand that exposes the next. Live expert readings beside each layer: Compute chips: experts read Bullish, net +34 · 62 claims · full-coverage reading. Memory (HBM): experts read Bullish, net +49 · 40 claims · full-coverage reading. Datacenter power: experts read Very Bullish, net +71 · 21 claims · thin sample — provisional. Grid equipment: no live reading (not yet a tracked topic). Materials: experts read Very Bullish, net +86 · 7 claims · very thin sample — treat as anecdote. Robotics / physical AI: experts read Very Bullish, net +91 · 12 claims · thin sample — provisional.
Readings are the net stance of the graded expert voices Synthos tracks, over the claim window 2026-06-27 to 2026-07-25 — the same data behind The Delta. Where the sample is thin, the reading says so; a strong-looking number on a handful of claims is an anecdote, not a signal.
How the radar hunts
The radar is a synthesis discipline, not a secret indicator. Conceptually, it watches four things and demands that they agree. (What follows is the full public description — we publish our reasoning and our graded record, not our internal recipe.)
1 · Narrative acceleration
Is conviction building across our graded expert panel — the voices whose past calls actually held up — and is it building faster, not just louder? A story spreading through people with measured skill is different from a story spreading through a feed.
2 · Price physics
Curvature, not levels. A high price tells you nothing; a price whose rate of climb is itself climbing is the signature we care about. The radar reads the shape of the move, not its headline number.
3 · Fundamental confirmation
Is the business going parabolic, or only the price? Revenue, backlogs, and delivery lead times bending alongside price is a supply collision. Price bending alone is a bubble — the radar's job is to tell them apart, in public.
4 · Supply inelasticity
How long until supply can actually respond — in the words of the experts closest to it? Stated lead times for fabs, mines, turbines, and transformers put a floor under how long the collision can last.
Every radar call will be logged with a date, graded against what actually happened, and left on the record — wins and losses alike — on the same terms as the rest of the Synthos prediction record.
The honest part
Most ignitions fail. For every parabola that completes, many curves bend and then flatten back into nothing. Expected hit rates here are a minority of calls — the economics work only because the winners are large, not because the calls are usually right. If someone shows you a parabola-hunting record with a high hit rate, look for the bodies.
This is the most survivorship-poisoned corner of markets. The parabolas everyone remembers are the ones that worked; the graveyard is invisible. Our source data includes the graveyard on purpose — the failed calls and faded narratives stay in the record, because a radar trained only on survivors is a story-teller, not an instrument.
Scope: long-only, in publicly accessible instruments — stocks, ETFs, and tokens. No leverage math, no derivatives, nothing a reader cannot actually access. And as everywhere on this site: this is educational research, not investment advice.
Where is the board? Launching this page as an explainer first is deliberate. The live radar board — named candidates, phase assessments, and their grades — publishes here as signals accrue and survive grading. The reasoning had to be on the record before the first call, so you can judge the calls against it. The radar’s expression as a portfolio — the Parabola model portfolio — lives on the model portfolios page; the CALL board itself still arrives here only after grading.