Methodology

What does a rules-based basket built on PIInneh’s research look like, and how does it behave?

Eleven indices in eight families, each built for one purpose: QuantGDI 50, 150 and 450, diversified baskets of companies selected for low correlation; QuantAIP 100 and 200, machine-learning evidence on fundamentals, projections and forensic indicators; QuantBIOT 100, biotechnology companies assessed on their pipelines, chances of success and cash runway; QuantUV 100, companies trading below intrinsic value whose finances are sound; QuantHM 50, the strongest and most persistent momentum confirmed by the Superradiance trend; QuantDIV 100, dividends supported by cash generation; QuantGTH 100, the strongest projected growth; and QuantLEADER 100, the leaders of each industry. Every index is weighted with Active Floating Market Cap Weighted + 10% Cap (until float data is connected, by full market capitalization), is reconstituted every quarter, and is checked weekly and daily for diversification.

Nine grades on one scale, for every kind of company.

Every company is rated on one nine-notch scale, from πAAA (strongest) to πC (weakest). The rating comes from the distress model that fits the company’s sector, so a bank, a manufacturer and a clinical-stage biotech are each judged by a test built for them, and the results remain comparable on one scale.

A rating moves when its model crosses a zone line, and the previous rating is kept, so every upgrade and downgrade is visible. How the zones map to the nine grades will be published in the technical specification with the first live ratings.

Each company is judged by a test built for its kind of balance sheet.

• Operating companies: Altman Z, checked with Beneish M, Piotroski F, Ohlson O and Zmijewski. PIInneh’s zone lines: safe above 2.99, grey 1.2 to 2.99, distress below 1.2.
• Financial companies: Merton distance to default and probability of default. PIInneh’s zone lines: safe at 4σ or more, grey 1.2σ to 4σ, distress below 1.2σ.
• Biotechnology companies: cash runway (DD_Bio) with the catalyst calendar. Safe at 18 months or more, grey 6 to 18, distress under 6.

The corporate models are never applied to financial or biotechnology companies: their balance sheets are not what those models were built for.

End-of-day data, recalculated after every close once live data is connected.

Ratings, signals, theses, probabilities and index figures are recalculated after every close, and each report names the close it was computed on.

Until live market data is connected, every figure on this site is sample data, generated to show how the methods work; it is not market data.

Which companies pass the tests that matter to me, and how would they fit together?

You set filters in seven groups (universe, forensic health, valuation and rating, forward estimates, industry-relative ratios, trend and portfolio fit), or start from a ready-made screen: Quality at fair value, Deep value, Distress watch, Hedge finder, Biotech catalysts and Index candidates. Financial health is judged by the model that fits each company: Altman Z, Merton distance to default or cash runway. Deep value adds a value-trap guard that removes cheap companies whose finances are weak. The companies that pass can be built into minimum-risk and maximum-Sharpe portfolios, long only with a 10% limit per company unless you choose otherwise.

What is this company, how strong is it, how is it trending and what is it worth?

One page brings together the company’s π rating and distress model, the PIInneh Quant chart, the Superradiance trend, statistics against its industry, its financial statements and history, financial projections, insider and activist activity, and the Correlation Lab. The Superradiance trend score combines how consistently the stock has beaten its risk over the past year, how much of the year it spent in an uptrend, and its three-month momentum; its strongest reading also needs all four of its confirming checks. An index symbol opens an index view with the company-only parts removed.

How does this company compare with its own industry, measure by measure?

About 100 measures in eight groups: the company against its industry, market-based valuation, the income statement, the balance sheet, cash flow, risk, return and volatility, return-series risk, and advanced return-series measures. For each one PIInneh shows the company’s figure, the industry median and average, the gap to the median, and the company’s percentile among its industry peers, with a one-line reading of what it means. The same peer percentiles drive the Screener’s industry-relative filters. The Financial Ratios API will deliver these measures to your own systems.

What is the pipeline worth after its risks, and does the cash last until the next catalyst?

Each program is valued on its own: what it would earn over its commercial life, weighted by its chance of success, less what it still costs to develop. The programs are added together, a reserve is held when value is concentrated in one program, and cash is added to reach a value per share. Cash runway is projected month by month against a safety reserve, not against zero, and compared with the next catalyst date. A scenario lab shows the value if the lead program fails, approval is delayed, the price or market share is lower, or costs rise. Biotech Company Research is the company page for biotechnology: cash runway takes the place of the corporate distress models, and it adds the next catalyst, the funding gap, expected dilution and the Biotech Link in the Correlation Lab.

What will the business earn, and what is that worth per share?

Machine-learning models forecast what drives the business, such as growth and margins. Accounting rules then build revenue, earnings, cash flow and the balance sheet so that they tie together, each with a low and a high case. The projection becomes an intrinsic value per share, using the valuation model that fits the company. Biotechnology companies are valued program by program instead (see above).

Where is the price likely to be after a chosen number of bars, and how sure is the model?

Each indicator gives a reading between down and up. The model judges the market’s condition as chances across six regimes, such as a persistent trend, a sideways range or liquidity stress. It replays the recent past, issuing test forecasts and scoring each one only once its outcome is known, so it never learns from information it would not have had. Each indicator and factor earns weight by how well it has predicted this stock, at this horizon and in this regime; indicators that repeat their family are weighted down, and direction is decided by weighted evidence, not by a majority of indicators.

Which of my stocks move together, where is my risk concentrated, and which mixes and hedges spread it best?

The Lab measures how every pair of stocks moves together and reads the result through four lenses. Overview: the matrix and Market Link, a stock’s typical strength of connection with the whole market, placed in the market’s thirds. Risk: how connections tighten on days the market falls 1.5% or more, how distress could spread, and a shock simulator. Structure: hidden clusters of stocks that move as one, why stocks are connected, how stable the connections are, and which stocks lead others. Action: equal-weight, minimum-risk, maximum-Sharpe and most diversified mixes (long only, at most 10% per company; with fewer than ten companies that limit cannot be met, so those mixes have no limit and the page says so), the efficient frontier, and the best hedge for each holding. The biotech version adds the Biotech Link, a stock’s connection with the biotechnology industry.

What changed in each stock after the close, and do PIInneh’s methods agree?

Five method groups each give a direction for a stock: up or down. PIInneh counts how many agree, lists the stocks on which three or more agree, and runs four screens: value trap, quality at a discount, smart money and distress watch. Rating migration shows the upgrades and downgrades.

Is the reason I own this stock still true?

Each condition of a thesis is checked after every close. When all pass, the thesis is Intact; when one fails, Weakening; when two fail or the rating is cut, Broken. For each condition PIInneh finds the price at which it would break and the chance of that within one and three months. The list is also checked for stocks that move too closely together and run through the Black Swan scenarios.

What do I need to know to decide about this stock today?

The Decision report answers seven questions on one page: why own it, what it is worth, what would break it, how likely that is, what a crisis would do, who agrees, and what to watch next. The factsheet, Signals, forensic, thesis and Black Swan reports each go deeper into one part. You can read a report in the Investor or the Analyst view, print it, take it to Excel, and record what you decided and why.

If an extreme event happened, which index would hold up best?

Each company takes the shocks through its sector’s sensitivity to each factor, cushioned by its quality and deepened by its fragility, and no loss can exceed 100%. An index’s result is the weighted sum of its members’. PIInneh adds the cost of trading through the stress at your account size, a low-to-high range for the fall, and the time to recover. A resilience score from 0 to 100 combines the fall, the trading cost, the recovery time, concentration and quality. Two broad-market benchmarks take the same shocks for comparison.