GEMSTONEPORTFOLIO
For stocks that have fallen hard

Every falling stock looks like a bargain. Most of them are traps.

GemstonePortfolio reads the evidence underneath a crash: whether the business still stands, whether the selling is running out, and whether it is being absorbed, trade by trade. Then it tells you where that evidence stands, including when the honest answer is not yet.

Read a company, from $9Why are you here? →
118,780trades read in a single week237,579calculations on that week alone$0subscription, pay per company
The Buy Lens for the invented company SAMPLE
StructureBase building, 33 weeks between $8.94 and $13.40
Not yetEach share sold now does 20% more damage than at the start of the window
The business
Discounted
Piotroski 6/9, clean books, realistic DCF +83%
The structure
PB · building
capitulation 36 weeks ago, supply 64% dried
The tape
Recovery 50.6%
low reclaimed and held 8 of 8 sessions
The verdict
Start buying · 44%
a partial stage, not a full position: the absorption is not confirmed

One company, read three independent ways, then weighed into one verdict. SAMPLE is an invented company; every screenshot on this page is the real system run on its generated data.

Why are you here?

Nobody lands on a page like this by accident. Something happened to a stock you own or watch. Here is what the system answers for each of the usual reasons.

It fell 70%. Bargain or trap?
A cheap price proves nothing. Companies go to zero cheaply every year.
What you getBankruptcy, health and manipulation tests for each of five years, three distress gates that strip the "cheap" credit from a failing business, and a check on whether the price still means anything at all.
See the business test →
I own it and it keeps falling.
Average down, hold, or cut? Every forum says something different.
What you getThe holder's voice: Thesis breakingHoldAdd, the price whose loss breaks the case, and whether the sellers are actually running out, trade by trade.
See the tape →
It bounced 25%. Am I late?
Chase it, or wait for a pullback that may never come?
What you getWhere the move sits in the base: still inside, testing, or already trending above it, plus Room to Run and Wait for pullback when the low-risk entry has passed.
See the structure →
It's up. When do I sell?
Sell too early out of fear, or ride it all the way back down?
What you getThe Sell Lens: stretch, trend, distribution into strength and the value left, resolved into five phases from Holding to Sell now.
See the exit →
Who is buying all these shares?
"Institutions are buying" means nothing if it is an index fund rebalancing.
What you get13F holders split into those who chose the stock and the portfolios that hold it by rule, and whether the two moved the same way last quarter.
See the holders →
Is the market even with me?
The right stock in the wrong market still loses money.
What you getThe market regime: eight index engines, credit, volatility, rates, crowding and five regime detectors, and the most neglected corner of the market worth watching.
See the market →

What a screener will never show you

Scores and ratios are everywhere, many of them free. This is what sits on top of them, and it is where the decision is actually made.

Every trade
The tape, not the candle
Bottom Absorption reads each individual trade in the window: who hit the bid, who lifted the offer, what traded off-exchange, and what price did afterwards.
18 · 10
Tape signals, market states
From climax decay to VWAP deflection and chopped-up hidden orders, resolved into states such as capitulation, distribution, absorption and recovery.
PA → PE
The structure of a base
Capitulation, the rally that sets the ceiling, the test, the breakout: each event found in the price itself, and cancelled when price takes it back.
Not yet
A verdict that argues with itself
Every reading prints the evidence against it next to the evidence for it, and names the price that would prove it wrong.
Chose ≠ hold
Deliberate against routine holders
A share held by a 40-stock manager is not a share held by an index. The 13F view scores every holder for how deliberate its position is.
Q ≠ Y
The quarter-adjusted year
When the latest quarters have moved away from the annual figures, the page's own DCF, Piotroski, Altman and Beneish are recomputed on the fresh quarters.

A bottom is a process, not a price

The system does not guess where a falling stock will stop. It measures the transition, stage by stage, and shows you which stage the evidence supports.

1
Selling pressure
Company Detail, Bottom Depth
2
Exhaustion
Bottom Depth, the capitulation week
SAMPLE IS HERE
3
Absorption
Bottom Absorption, trade by trade
4
The base
Market Phase PB, supply drying
5
Demand returns
PC and PD: spring, sign of strength
6
Markup
PE, then the Sell Lens takes over

Why read the bottom at all? Because the move is made before the crowd arrives.

Most traders buy at stage 5, when demand has visibly returned: the breakout, the strong close above the ceiling, the chart that finally looks safe. By then the stretch from the floor to the ceiling has already been travelled without them, and the point that proves them wrong now sits a long way below their entry.

Take SAMPLE's own levels: a floor at $8.94, a ceiling at $13.40, and the price at $10.62 inside the base.

Reading the bottom

Staging in inside the base, at $10.62
The move up to the $13.40 ceiling+26% ahead of you
Distance to the $8.94 floor that breaks the case16% below
What you are waiting forevidence, not the crowd

Waiting for demand to return

Buying the breakout, at $13.40 at the earliest
The move from $10.62 to the ceiling+26% already gone
Distance to the same $8.94 floor33% below
What you are buyingconfirmation, at a higher price
The honest trade-off. The earlier you act, the less has been confirmed. That is exactly why the system never says "all in" at the bottom: it names the stage the evidence supports, sizes it (SAMPLE: a stage of 44% of intended size), names the price that would prove it wrong, and tells you what would have to print before the next stage. On SAMPLE today: the low at $10.02 was reclaimed and retested twice, each retest weaker than the last, but absorption efficiency is getting worse and the sellers are not exhausting. So: a partial stage, and watch.

Two depths. Pay once, per company.

No subscription, no dashboard to babysit. Pick the company, pick the depth.

Company Detail

$9
one company, one-time payment
  • Piotroski, Altman and Beneish: their sub-tests drawn for each of five years
  • Twelve more five-year rows: valuation, returns, cash, balance sheet
  • Missing figures named in red, never filled in
  • Six trading setups, eight technical systems, the confluence engine
  • Roughly 1,400 calculations
Get Company Detail, $9
THE FULL SYSTEM

Everything, for one company

$39
one company, one-time payment, Company Detail included
  • Bottom Absorption: every trade in the window, 18 signals, 10 market states
  • Buy Lens: the phase, the entry trigger, five boxes with their evidence, nine lenses, Suggested Allocation
  • The base model: PA to PE, with what advances it and what breaks it
  • Three standalone checks: Hallucination, 13F Holdings, Quarter vs Annual
  • Sell Lens with five exit phases, and Market Sentiment
  • About 105,000 calculations
Get the Full System, $39
Why you're here · 1

It fell 70%. Bargain or trap?

A cheap price proves nothing. The first question is whether there is still a business underneath it.

The whole business on one screen$9

Company Detail opens with the price inside its five-year and 52-week range, the three forensic scores, and three written blocks that read the page for you: price and tape, the business, and value, risk and outside view, from the realistic DCF to insider filings, shelf registrations and analyst ratings.

Five pass/fail cards answer the first questions a careful investor asks, each with the numbers that decided it: financial health, price dislocation, seller exhaustion, capital protection and earnings quality.

On SAMPLE: an Altman Z of 2.67 in the grey zone, Piotroski 6 of 9, clean books, earnings backed by cash, the price at 3% of its five-year range.

Company Detail header for SAMPLE

The real Company Detail header on SAMPLE.

Piotroski sub-tests for SAMPLE drawn as bars over five years

Every Piotroski sub-test on SAMPLE, five years side by side. Good grows up from the middle line, bad grows down.

Are the books clean, every year?$9

A score shown for this year only hides the year the trend turned. Here the sub-tests are drawn for every one of five years: eight of Piotroski's nine tests of financial health on their own rows, Altman's five bankruptcy components, and all eight of Beneish's manipulation indices, including TATA, the gap between reported profit and the cash actually generated.

Each cell is coloured by its level and marked by its direction, so a healthy reading that slipped, a weak one that is improving and a weak one getting worse all look different. The last one is exactly what the page is built to surface early.

Nothing is faked to fill a cell. A figure the company did not report is named in red on that cell, and for the F-Score a missing input counts as a failed test, so the score can only be understated. A ratio whose denominator collapsed is flagged with the figure that caused it. A figure carried forward is marked with the year it came from.

Returns, cash and the balance sheet$9

Twelve more five-year rows: net current asset value, the Graham Number, price to free and to operating cash flow, the realistic DCF upside, return on capital after tax, gross and net margin, revenue, free cash flow per share, debt to equity and the quick ratio. 33 rows in all.

Where a ratio has no meaning it says so: a negative free cash flow leaves no multiple, negative equity leaves debt to equity undefined, and a price below net current asset value, the classic Graham net-net floor, gets a banner of its own.

Fundamental trajectory for SAMPLE

SAMPLE's fundamental trajectory, the latest twelve months on the left.

Asset Floor lens for SAMPLE

The Asset Floor lens on SAMPLE, every row with the points it earned.

Cheap for a reason is caught$39

In the full system, three fundamental lenses decide whether there is a business case at all: Asset Floor (book, liquidation value and earnings power against the price), Quality Dislocation (a sound business the market has not caught up with) and Insider Conviction (the people closest to it, and whether the share count is shrinking or being diluted).

Three distress gates. An Altman Z in the distress zone cuts the asset floor to a quarter: the stock looks cheap because it is failing. A Beneish M pointing at manipulation cuts the quality reading to a quarter. A Piotroski below 3 cuts both. And negative equity is never counted as a discount.

Without a value case, the verdict is Not yet, however good the chart looks.

Has the price left the business behind?$39

The Hallucination view reads nothing but the company's own price history and its own statements. No peer group, no analyst target, no sentiment. It scores how far and how abnormally the price moved (40 points), the gap between price and business (20), and whether the statements still agree with each other (40): do gross profit, operating income and cash flow move with revenue, does the physical base keep up, is the margin in line with its own history?

The answer sits on one axis: Punished Anchored Inflated. A punished price on statements that still agree is a discount. A price tracking the fastest line in statements that no longer agree is something else, and the card says which.

Hallucination view for SAMPLE

Hallucination on SAMPLE: ANCHORED, 37.0. Price and business tell broadly the same story.

Why you're here · 2

I own it and it keeps falling. Is anyone buying?

A stock can have a beautiful base and still fall through it. The chart cannot tell you who is taking the other side of the selling. The trades can.

Bottom Absorption on SAMPLE: RECOVERY 50.6%

Bottom Absorption on SAMPLE's invented two-week tape: RECOVERY, 50.6% of a fixed 735-point maximum, after a week of DISTRIBUTION.

Every trade, not the daily candle$39

Every other tool works from daily bars: one open, one close, one volume. Bottom Absorption pulls every individual trade in the window, tens of thousands in a single week, up to 118,780 so far, and asks what daily bars can never answer: what did the selling actually accomplish?

If heavy selling keeps producing lower prices, sellers are still in charge. If the selling keeps coming while price stops falling, the same low holds, sell volume shrinks, buyers step in and a new low is reclaimed, the selling is still there but becoming less effective. That is absorption, and it is what the page is built to detect.

Nothing guesses at intent. Every panel states an observable fact, then classifies it by what price did next, and where it cannot tell it says PENDING rather than inventing a number.

Eighteen signals, three tiers

  • Volume structure: climax decay, the cumulative buy-minus-sell arc since its trough, seller exhaustion over three, five and ten sessions, absorption efficiency, net-buy sessions
  • Price under selling: how long price hugged the low without breaking, whether the cost base has converged, whether a new low was reclaimed and held, the quality of every retest, closing strength, recovery speed, and whether sell-heavy minutes still closed above their own average
  • Hidden flow: off-exchange volume judged only by its price response, block trades located against the low and the high, large orders chopped into small prints and reconstructed, the aggressor ratio, urgent sell sweeps across venues

The score measures how much evidence accumulated. The state is detected separately, from specific clusters of conditions: CapitulationBreakdownDistributionLate momentumMarkupRecoveryAbsorptionBase buildingMixed.

Tier 2 of Bottom Absorption on SAMPLE

Tier 2 on SAMPLE: the $10.02 low reclaimed and held 8 sessions of 8, two retests, each weaker than the one before.

And it is allowed to say the story is still wrong

For it
  • low reclaimed, held 8 of 8
  • two retests, each weaker
  • net buying rising 6 sessions
  • cost base converged
  • 171 minutes on the floor
Against it
  • absorption efficiency worse by 20%
  • sellers not exhausting
  • recovery too slow to count
  • off-exchange flow neutral
  • blocks with no buy lean

Its own summary: the rally is real, but the sponsorship evidence behind it is thin. That is not a failure of the system. That is the information.

What the holder hears

Hold the stock and the Buy Lens speaks in a holder's voice: Thesis breaking when the value case is gone, Hold when the case stands and no stage is due, Let this leg cool, Let it run, and Add only when a test has held, a spring has printed, or price is at or below your own entry inside a structure that allows it. A bottom fisher adds lower or on evidence, never higher on nothing.

Two slower panels ask what happened afterwards: after a heavy off-exchange day, did the dips get shallower or deeper? After every large block, did price trade above it one, three and five sessions later, below it, or neither?

Detected behaviour on SAMPLE's tape

Detected behaviour on SAMPLE, each pattern tagged STRONG, MODERATE or WEAK.

Why you're here · 3

It bounced 25%. Am I late?

A bounce is not a bottom. What matters is where the move sits in the structure of a base, and how much room is left.

The Buy Lens for SAMPLE

The Buy Lens on SAMPLE: START BUYING, entry trigger 53 of 100, base building, stage 44%.

Five questions, each with its evidence inside the box$39

  • Bottom Depth: how bottomed is the tape? Four technical lenses that must agree, lifted when institutional absorption confirms a genuine drawdown
  • Market Phase Model: how far through the base-building process is the stock?
  • Room to Run: nine oscillators on three timeframes, each against this stock's own history: how much space before the tape is extended?
  • Buyer Control: four money-flow measures: are buyers actually taking control?
  • Value Available: the business case, after the distress gates

One vocabulary of 37 named measurements runs through all five, each with its window in its name. Open the evidence panel and every signal says where it was computed, what it scored there and what it means for this company today, followed by what this is, what this is not and the weaknesses that are real.

A trigger that behaves like a chain

The entry trigger is a confluence of value, bottom depth, the turn in the structure and room to run, value weighted most. One weak link drags the whole number down: a cheap stock that has not turned and a turning stock that has already run both score modestly.

The phase name is read from what the structure actually printed: Not yetWatch closelyStart buyingBuy nowWait for pullback. If you are late, it says so: once price has left the range, the low-risk entry has passed and the page waits for the pullback that holds.

The grip says how large a stage the situation allows, what flips this read names the price that breaks the case, and Suggested Allocation answers how much: quality against payoff, a quarter of the Kelly result. On SAMPLE: 14.1%, confidence 65%, odds 4.2 to 1.

Where the stock stands in forming a base

Read on the weekly chart, from the price series itself. The capitulation is the week of greatest volume and spread at a genuinely new low; the ceiling is where the first rally tops out; the floor is the lowest price defended in between. A breakout is measured against the base's own typical week, never against the panic. Events that fail are cancelled, and two weekly closes below the floor cancel the base.

StepWhat it meansEvents the model looks for
PA · Stopping the downtrendThe fall is arrested. A heavy, wide week marks the panic low; a rally off it sets the ceiling.PS preliminary support, SC the capitulation, AR the first rally, ST a quieter retest
PB · Building the causeSideways inside the range while supply is absorbed. A base is time spent absorbing the decline.SUPPLY drying against the panic; weeks contained in the range
PC · The testA final shakeout under the floor that fails to attract sellers.Spring, and a quiet test of that low
PD · Trend within the rangeDemand has taken the ceiling and holds above it.SOS, LPS, higher lows
PE · Trend outside the rangeThe base is behind; price is trending.markup against the height of the base
On SAMPLE: PB. A capitulation 36 weeks ago on three times normal volume, a ceiling at $13.40, a floor at $8.94, 33 weeks in the range, supply 64% dried. It advances on a spring below $8.94 that reclaims the range, or a close above $13.40 on expanding volume. It breaks on two weekly closes below $8.94.
Why you're here · 4

It's up. When do I sell?

Buying is the easy decision. Most people sell too early out of fear, or hold too long out of hope.

The Sell Lens only gets loud when the tape and the business agree$39

Five boxes: Price Stretch (how far price has run across daily, weekly and monthly candles), Trend (how close the move is to its exhausted end), Seller Control (the same four money-flow measures, read from the selling side), Market Pressure (heavy volume that moves price very little, the classic sign of distribution) and Value Remaining.

The exit trigger weighs the core stretch-and-overbought reading most, then two distribution signals built to catch large holders selling into strength, and lightest of all the fundamentals, as the tie-breaker. For a position you hold it resolves into five phases:

HoldingWatchingSell into strengthScaling outSell now

Real value still left on the table keeps a stretched stock at Scaling Out rather than a full exit. A system that cries wolf on every red day is worse than no system at all, so the bar is deliberately high.

The Sell Lens for SAMPLE

The Sell Lens on SAMPLE: no position held, exit trigger 53 of 100.

Why you're here · 5

Who is buying all these shares?

"Institutions are buying" means nothing if it is an index fund rebalancing. And an annual report can already be out of date.

13F Holdings for SAMPLE with invented holders

13F Holdings on SAMPLE, every holder name invented: ACCUMULATED, 82.5.

Who chose the stock, and who holds it by rule$39

Four times a year every manager running more than $100 million reports its holdings to the SEC. This view prints only what is in those reports, and scores every holder for how deliberate its position is, from four things the report discloses: index funds inside the holder, options in its book, how much of its portfolio this stock is, and how many stocks it holds. Forty is a choice; twenty thousand is a rule.

Then: who holds the shares, who bought for the first time, who sold out, who added and who trimmed, which way the options lean, and the row that weighs most, did the deliberate holders and the routine ones move the same way? A manager that changed its SEC identity is matched, not read as a sale and a new buyer.

On SAMPLE: the holders who chose the stock bought into the fall; the selling came from portfolios that hold it by rule.

Do the latest quarters confirm the annual picture?$39

Almost every number rests on annual statements and a trailing-twelve-month column that can still carry quarters most of a year old. Quarter vs Annual compares the latest quarter with the same quarter a year earlier, on revenue, gross margin, core operating margin, cash conversion, the balance sheet and working capital, each against this company's own usual change.

Nine flags catch the quarters that would mislead: one-off items, one-offs that recur, a pending cash outflow, receivables or inventory running ahead of sales, profit without cash, buybacks on credit, an unreliable derived quarter checked line by line against the SEC's own figures, and restatements. Then it rebuilds the year on the fresh quarters and re-runs the page's own DCF, Piotroski, Altman and Beneish on it.

Together with Hallucination these are three views that give no verdict and feed nothing. That is deliberate: a warning that fed the score would be averaged away. Standing apart, a warning stays a warning.

Quarter vs Annual for SAMPLE

Quarter vs Annual on SAMPLE: IMPROVING, 75.0.

Why you're here · 6

Is the market even with me?

The same trade that works in a broad, healthy market loses in a narrow, late-cycle one.

Market Sentiment dashboard

The real Market Sentiment page, with market data as of 25 September 2026.

The market's vital signs on one wall$39

Most investors never measure the market's mood; they feel it, usually too late, from their own losses. No single vital tells the story, but a fast pulse with falling blood pressure is an emergency. Every box scores its own theme from 0 to 100 and lands in one of seven zones from oversold to overbought, coloured as temperature, not as good and bad, so the eye goes to the extremes.

  • Eight index engines, the trend backbone, combined into one composite
  • Crowding and rotation: who is beating whom, and how crowded the winners are
  • Volatility, rates and the curve, the dollar, commodities, credit and valuation
  • Five regime detectors and a one-sentence regime line
  • Board-wide extremes: any single reading at the 10th or 90th percentile, even inside an average-looking box
  • A buy recommendation: the most beaten-down area showing evidence its selling is exhausted, or an honest empty card

Two views of the same bottom

The structure asks what stage of the base-building process are we looking at? The tape asks what are the actual trades doing right now? They are built from different data and never feed each other, which is what makes their agreement worth something.

On SAMPLE the structure says PB, building the cause, the tape says RECOVERY, 50.6% with deteriorating efficiency. That does not mean buy. It means the stock is structurally inside a base, the transaction evidence is beginning to show features of absorption, and the confirmation is still incomplete.

No need to pick the exact penny. The useful question is not whether $10.02 is the bottom. It is what happens around $10.02 if the stock comes back there: a retest on reduced selling with buyers growing more aggressive makes the case stronger; a break through it on expanding sell volume says it was never a bottom.

Confidence comes from convergence, not from one signal

  • Structure: a base, supply drying, time in the range
  • Volume: the climax passes, volume declines, selling fades
  • Transaction flow: buyers increasingly dominate, net buying rises
  • Price response: selling becomes less effective, the low is reclaimed
  • Retest: price returns to the low and the selling fails to break it
  • Demand: a sign of strength, a pullback that holds, higher lows

The goal is not to predict the bottom. It is to read the transition.

From selling pressure, to exhaustion, to absorption, to a base, to demand returning, to confirmation. The earliest stages offer the greatest asymmetry and the greatest uncertainty, so the system does not hide that uncertainty behind a green button. It shows you where the evidence stands.

Stop asking has it fallen enough? Start asking is the selling exhausted, is someone absorbing it, is price becoming resilient, is demand taking control?

Read a company, from $9Back to why you're here

Ready? Pick the depth.

One company, one payment, no subscription.

Company Detail

$9
one company, one-time payment
  • Piotroski, Altman and Beneish: their sub-tests drawn for each of five years
  • Twelve more five-year rows: valuation, returns, cash, balance sheet
  • Missing figures named in red, never filled in
  • Six trading setups, eight technical systems, the confluence engine
  • Roughly 1,400 calculations
Get Company Detail, $9
THE FULL SYSTEM

Everything, for one company

$39
one company, one-time payment, Company Detail included
  • Bottom Absorption: every trade in the window, 18 signals, 10 market states
  • Buy Lens: the phase, the entry trigger, five boxes with their evidence, nine lenses, Suggested Allocation
  • The base model: PA to PE, with what advances it and what breaks it
  • Three standalone checks: Hallucination, 13F Holdings, Quarter vs Annual
  • Sell Lens with five exit phases, and Market Sentiment
  • About 105,000 calculations
Get the Full System, $39

Questions

What is the difference between Company Detail and the full system?
Company Detail ($9) is the five-year forensic workup of the business: the three scores with their sub-tests for each of five years, 33 five-year rows, the confluence engine, the trading setups and the technical systems. The full system ($39) includes all of that, plus Bottom Absorption, the Buy Lens with its base model and three standalone views, the Sell Lens and Market Sentiment.
Why does the full system cost $39?
Bottom Absorption works on every individual trade in the window, tens of thousands in a single week, up to 118,780 so far. That trade data is bought for each company you analyse, and it is the largest part of the price.
Does it tell me when to buy?
It tells you where the evidence stands. It names the situation, prints the evidence for and against it, and says what would advance the read and what would break it. It is allowed to say not yet, and often does. The decision stays yours.
Is this a subscription?
No. Every purchase is a one-time payment for one company. There is no recurring charge.
How long do I have access, and can I keep the pages?
After purchase you receive a personal link. When you open your first view, your 24 hours begin on that browser. Within them you can open and reopen every view in your package as often as you like. The link works on one browser only and can't be shared.

Every page is built to stand on its own. Save each one to your computer, whether Company Detail, Buy Lens, Sell Lens, Bottom Absorption or Market Sentiment, and it stays yours for good, complete, with every figure and chart. Save your pages before your 24 hours end.
How current is the data?
Fundamentals come from the company's own reported statements, including its latest quarters. Prices, the lenses and the trade analysis are calculated fresh at the moment you choose your company after purchase, from the latest data available that day.
Is SAMPLE a real stock?
No. SAMPLE is an invented company. Every price, statement figure, holder and trade behind it was generated for this page, and no real company's data was used. The pages in the screenshots are the real pages, run on that invented data.
Is this financial advice?
No. GemstonePortfolio provides data, calculations and analysis to inform your own decisions. It is not personalised investment advice.
Read a company from$9Start