CompanyGraph
Filtering for stochastic %K, Bollinger %B, and RSI's own Bollinger %B all reading upper-side surfaces stocks whose current price and momentum sit in the upper portion of their recent reference frames — a direction-asymmetric configuration the conventional 'overbought' label misreads as predictive.
How to use the screener to read upper-side configurations of price-within-range, price-within-Bollinger-Bands, and RSI-within-its-recent-mean — and why this is not an 'oversold' screen.
What This Article Can and Cannot Show
The screener exposes one technical position interpretation — Price Extreme Position — which reads three upper-side configurations simultaneously. The interpretation does not identify oversold conditions despite the conventional vocabulary the surrounding fields sometimes suggest. All three wired observations are sign-asymmetric: they fire only when the close (or RSI) is in the upper portion of its recent reference frame. Equivalent lower-side configurations score 0–30 on those obs, not 70+, and therefore would not activate the interpretation.
A lower-side / oversold interpretation is not currently wired. Building one requires either an inverse-stochastic observation, an inverse-Bollinger-%B observation, or a rewrite using observations that honestly fire on lower-side configurations. Until that gap is filled, screens that look for stocks at the bottom of their range are outside what this tool can answer.
A second interpretation that some readers may expect here — Drawdown Risk Profile — was deactivated (Wave 47) because it referenced an observation key (ulcer-index) that does not exist in the observation catalog. The interpretation cannot fire under any conditions and is currently inactive.
What Upper-Side Configuration Means Structurally
Each of the three wired observations reads a different facet of where the present price (or RSI) sits within its own recent history. Stochastic %K measures the close's position within the 14-week high-low range. Bollinger %B measures the close's position within the 20-week Bollinger Bands (mean ± 2 standard deviations). RSI's own Bollinger %B applies the same lens to the RSI series itself — where does the current RSI value sit within ±2 standard deviations of its 20-week mean.
The interpretation fires when all three are in their upper portions simultaneously. That is the entire claim: three independent upper-side position readings co-occur. It is not a momentum-direction reading (it does not say price is rising), it is not a mean-reversion forecast (it does not say price is about to fall), and it is not a valuation reading (the formulas know nothing about earnings or book value). It is a structural co-occurrence observation across three position metrics.
The observations are direction-asymmetric in a specific way: their underlying formulas use the upper-side reading directly, mapping high values to high scores. A close near the 14-week low scores around 0–10 on stochastic %K; that low-side configuration cannot push the score above 30 and therefore cannot enter the active set the resolver uses. Building a lower-side equivalent would require either inverse versions of these obs or different obs entirely.
Key Observations
Close Position Within Recent High-Low Range (legacy 'stochastic-standard')
What it measures: Stochastic %K — where the current close sits within the 14-week high-low range. Score 0 = current close equals the 14-week low; score 100 = current close equals the 14-week high. Direction-asymmetric: fires (70+) only when the close is in the upper portion of the range.
Data source: Weekly closing prices over the trailing 14 weeks, identifying the high and low boundaries and the current close's position between them.
Price Position Within Bollinger Bands (legacy 'bollinger-percent-b')
What it measures: Bollinger %B — where the current close sits within the 20-week Bollinger Bands (mean ± 2 standard deviations). Score 0 = current close equals the lower band; score 100 = current close equals the upper band. Direction-asymmetric in the same sense: fires (70+) only on the upper-band side.
Data source: Weekly closing prices over the trailing 20 weeks, with the moving average and standard-deviation bands computed from that window.
RSI Above Its Recent Mean (legacy 'rsi-bollinger-bands')
What it measures: Where the current RSI sits within ±2 standard deviations of its own 20-week mean — Bollinger %B applied to the RSI series rather than to price. Score 0 = RSI at its lower band (well below recent mean); score 100 = RSI at its upper band (well above recent mean). Direction-asymmetric: fires on the upper side of the RSI's own recent reference frame.
Data source: Weekly RSI values over the trailing 20 weeks, with mean and standard-deviation bands computed from that window.
Low Annualized Volatility (legacy 'inverse-vol-1y')
What it measures: Annualized realized volatility of weekly returns over the trailing year, mapped inversely so low volatility produces a high score. A high score means weekly returns have been stable; a low score means weekly returns have been wide-swinging.
Data source: Standard deviation of weekly returns over the trailing 52 weeks, annualized.
Operating Cash Flow Relative to Net Income (legacy 'earnings-quality')
What it measures: Operating cash flow divided by reported net income for the most recent annual period. A high score means reported earnings are well-backed by cash; a low score means reported earnings exceed cash generation. Single-ratio reading, not a composite.
Data source: Operating cash flow and net income from the most recent annual financial statements.
Revenue Growth Composite (legacy 'growth-consistency')
What it measures: A composite reading of revenue growth that combines median annual growth rate, share of positive-year observations, and stability across the lookback window. A high score means revenue has been growing positively and consistently across recent fiscal years.
Data source: Annual revenue from the income statement across the configured number of fiscal years.
Interpretations That Emerge
Price Extreme Position (Upper Side)
Constituent observations: Close Position Within Recent High-Low Range (legacy 'stochastic-standard'), Price Position Within Bollinger Bands (legacy 'bollinger-percent-b'), RSI Above Its Recent Mean (legacy 'rsi-bollinger-bands')
What emerges: Three upper-side position readings co-occur. Stochastic %K shows the current close in the upper portion of the 14-week high-low range, Bollinger %B shows the current close in the upper portion of its 20-week Bollinger Bands, and RSI is above its recent 20-week mean (Bollinger %B applied to the RSI series itself). The configuration describes where price and RSI sit within their respective recent reference frames; it does not claim mean reversion follows.
Limits: All three readings record present position within a recent reference frame; none measures momentum direction, predicts mean reversion, or indicates timing. The conventional 'overbought' framing is predictive; the formulas make no such claim. Prices and RSI can remain in upper portions for extended periods. The legacy interpretation name 'Price Extreme Position' is direction-asymmetric — only the upper extreme is observable from this configuration. A lower-extreme / oversold reading would require different observations.
Stable Foundation
Constituent observations: Low Annualized Volatility (legacy 'inverse-vol-1y'), Operating Cash Flow Relative to Net Income (legacy 'earnings-quality'), Revenue Growth Composite (legacy 'growth-consistency')
What emerges: Weekly-return volatility has been low over the trailing year, reported earnings are well-backed by operating cash, and revenue has been growing consistently. The combination describes a business with stable market-perceived risk, cash-supported earnings, and positive consistent revenue growth — a structural anchor on the fundamental side that does not depend on where price currently sits within its range.
Limits: All three observations are backward-looking. Low realized volatility over the trailing year does not guarantee future stability; cash-supported earnings in the most recent annual period can deteriorate in subsequent quarters; revenue growth consistency reads what happened, not what comes next.
Using the Screener
Upper-Side Position Screen
Select Price Extreme Position to find stocks where stochastic %K, Bollinger %B, and RSI's own Bollinger %B all sit in their upper portions simultaneously. This is a present-configuration screen, not a mean-reversion forecast. The companies passing are at a confirmed upper-side technical position across three independent readings.
To overlay a fundamental anchor, add Stable Foundation. Companies passing both screens have upper-side technical position alongside low realized volatility, cash-supported earnings, and consistent revenue growth — a present configuration where the technical and fundamental surfaces both read favorably.
Drawdown-Recovery Configuration
For stocks at upper-side technical position that arrived there through a recovery rather than an unbroken uptrend, combine Price Extreme Position with Drawdown Recovery Position from the screener's full catalog (covered in the finding-stocks-that-have-fallen-with-fundamentals-intact guide). Drawdown Recovery Position reads drawdown-from-peak alongside fundamental stability — pairing it with Price Extreme Position narrows the result set to stocks that experienced a meaningful drawdown, have since recovered to upper-side technical position, and retain stable fundamentals.
Boundaries
What These Observations Cannot Tell You
The wired observations read present position within recent reference frames. They do not predict the next move. A stock at the upper-side technical position can continue higher, plateau, or decline — the configuration provides no directional forecast. The conventional 'overbought = mean reversion follows' framing is a prediction layered on top of the position observation; the formulas themselves make no such claim.
Lower-side / oversold configurations are not currently observable. All three wired position obs are direction-asymmetric: they fire only on the upper side. A stock trading near the bottom of its 14-week range, near the lower Bollinger Band, with RSI below its recent mean would score 0–30 on each — outside the active range. Building an inverse observation that fires on the lower-side equivalent is a wave-4 candidate.
Drawdown-based screens that some readers may expect (drawdown depth, drawdown duration, ulcer index) are not currently wired. The earlier Drawdown Risk Profile interpretation referenced a phantom observation key and was deactivated. The screener exposes drawdown-from-peak through the Drawdown Recovery Position interpretation (see finding-stocks-that-have-fallen-with-fundamentals-intact), which pairs drawdown with fundamental stability rather than with volatility-regime / duration obs.
Finally, the wired position observations carry no valuation information. A stock at the upper-side technical position can be cheap by P/E and book-value metrics, expensive by those same metrics, or mid-range — the configuration says nothing about price-to-fundamentals. That assessment lives on a different surface entirely.