Guide
How to use every page and featureOverview
The /news page shows live price predictions generated from the most recent earnings filings (10-K, 10-Q, and 8-K forms) as they are published on the SEC website. Each entry represents the latest prediction for one company, using our proprietary equilibrium price model. This target is the price our algorithm believes the market should converge toward, based on the company's reported financials. Companies are selected as "winners" or "losers", when they signifacantly outperform or underperform compared to all other companies in the sector.
The page is split into three tabs; Latest, Winners, and Losers. It also includes a filter panel to help you narrow results by prediction quality, sector, and filing type. A timestamp in the top-right corner shows when the data was last refreshed.
Filter Panel
At the top of the page sits a filter card. Every filter applies to all three tabs simultaneously, and all filters use AND logic. A prediction must pass every active filter to remain visible. The controls below are fully interactive; try clicking them.
FILTER PREDICTIONS BY:
FILTER BY METHOD, SECTOR, AND FORM TYPE:
Each filter bar has three clickable segments. Red (left segment) = no filter active, all predictions are shown. Amber (middle segment) = low-quality predictions are hidden. Green (right segment) = only the highest-quality predictions remain. Click the bars in the mock card above to see them in action.
Uncertainty: lower is better
Measures how much the underlying financial data deviates from quarter to quarter. A lower value means the company's reported figures are stable and consistent, which produces more reliable model inputs and therefore a more trustworthy prediction. A high uncertainty score signals that the data feeding the model varies a lot. Treat those predictions with more caution. Moving the filter right progressively removes high-uncertainty predictions and leaves companies with steadier financials. Whether that is desirable depends on your strategy; some traders prefer the predictability of low-uncertainty stocks, while others seek the volatility of high-uncertainty ones.
Volatility: higher is better
Measures how volatile this stock has been historically, in the days after earnings reports came out. Some companies are not susceptible to big price swings after earnings. Their stock barely reacts to new information, for better or worse. Other companies are very reactive (more volatile). Their stock jumps or drops significantly after earnings. This filter does not remove predictions because they are "wrong", but we consider predictions for more volatile stocks to be more actionable, because they indicate a stronger market reaction to earnings news. Moving the filter right removes low-volatility (steadier) stocks, keeping only the most dynamic movers. Whether that is desirable depends entirely on your strategy.
Fit Error: lower is better
Shows how close the equilibrium model converged to a perfect solution. A lower fit error means the algorithm found a tight, internally consistent price that satisfies all the financial constraints simultaneously. In other words, a more precise prediction. A high fit error means the model had to make larger trade-offs to reach a solution, so the resulting price target is less precise. Moving the filter right removes high-error predictions, keeping only the cleanest results.
Method: Long Term vs Short Term
This toggle controls how historical earnings reports are weighted when calculating the equilibrium price.
Long Term weights older filings more heavily, giving each reporting period roughly equal influence. This can surface deep value; companies that have been structurally undervalued or overvalued for an extended period, regardless of short-term noise. A Long Term target is more likely to sit further from the current market price, reflecting a thesis that takes time to play out.
Short Term prioritises the most recent filing, producing predictions that tend to sit closer to the current market price and react faster to changes in the business. A Short Term target is more likely to agree with recent analyst consensus, but may miss longer-running valuation discrepancies.
Neither method is universally better. Long Term suits value-oriented or fundamental analysis; Short Term suits momentum or event-driven approaches. When in doubt, compare both; the most reliable predictions will be those where the two methods are in agreement.
Sectors & Form Types
The two dropdown buttons let you restrict results to specific sectors (e.g. Technology, Healthcare, Energy) and specific SEC form types. 10-K is the annual report, reported once a year. 10-Q is the quarterly report, reported three times a year after each fiscal quarter other than the fourth. The 10-K report is essentially the last quarterly report of the fiscal year. 8-K is a current report for material events such as earnings releases or major announcements. We only process the 8-K filines with complete earnings releases. These earnings releases often come out a couple of days before the 10-Q/10-K, so they can be useful for getting an early read on the model's reaction to the new financial data. By default, all sectors and all form types are selected.
Tabs
Below the filter panel are three tab buttons. The active tab is highlighted in blue / amber; inactive tabs are muted. All filters stay active across tabs; switching between them does not reset your filters.
Shows every prediction for the last few days, sorted by the most recent filing timestamp. The most recent filings are found at the top. This is the broadest view: it includes both positive (upside) and negative (downside) predictions across all sectors and industries, with no additional selection criteria beyond what you set in the filter panel.
Clicking any row immediately loads that company in the Screener, where you can explore its full price target chart, historical filings, and detailed sector comparisons.
Freely accessible to all users.
A curated subset of predictions where the company outperforms every other company in its sector on nearly all financial metrics simultaneously. This is not simply "a positive prediction". It means the company ranks near the top of its sector across revenue growth, profitability, balance sheet strength, and other fundamentals, and the model still predicts a positive price move. Both signals must be aligned.
Winners are updated each time a new batch of filings is processed. The list will typically be much shorter than Latest, because the sector-wide comparison sets a deliberately high bar. A company appearing here is considered a structural outperformer, not just a short-term mover.
Premium feature — requires an active subscription. Non-subscribers can see the column layout but rows are hidden.
The mirror image of Winners. Here we list companies that underperform every other company in their sector on nearly all fronts, and for which the model predicts a further negative price move. Both the fundamentals and the model are aligned on the downside.
Losers are useful for identifying potential short candidates, for avoiding positions in deeply underperforming companies before a further decline, or simply for understanding which companies in a sector are lagging structurally.
Premium feature — requires an active subscription. Non-subscribers can see the column layout but rows are hidden.
Reading a Row
Each row represents one company's latest prediction. Below are two example rows. One strong prediction, one weak one. Scroll right on small screens if needed.
Ticker
The company's stock ticker symbol, rendered in blue / amber. Click anywhere on the row to open that company in the Screener, where you can explore its full price target history, historical filings breakdown, and a detailed comparison against every other company in its sector.
Target Price & Implied Return
The model's equilibrium price (top), and the percentage return implied by that price relative to the most recent market price (bottom). Green = upside potential; red = downside risk. The sign and magnitude together tell you both the direction and the size of the predicted move.
Confidence (±%)
The combined uncertainty of the prediction expressed as a ± range around the target price. A lower number means the model is more certain. Green ≤ 10%, amber 10–15%, red > 15%. For example, a target of $100 with ±8% means the model's range is roughly $92–$108.
Quality Bars: Uncertainty · Volatility · Fit Error
Three fill bars visualising the same three metrics controlled by the filter sliders. The color coding follows the same logic: green = good, amber = moderate, red = poor. Scan prediction quality at a glance without reading numbers. All three green means a clean, reliable prediction. Any red bar is a signal to investigate further before acting.
Sector & Industry
The sector and sub-industry classification of the company, following the standard GICS taxonomy. This classification also determines which companies the Winners and Losers rankings are benchmarked against. The comparison is always within the same sector.
Warnings & Fails
Two counts shown next to the warning icon column. The top number (amber icon) is warnings: the company is underperforming compared to similarly sized sector peers, but the gap is within one standard deviation. A yellow flag worth watching. The bottom number (red icon) is fails: the company has fallen outside that one-sigma range entirely. A red flag. For a positive prediction, a high count suggests the stock may still underperform sector peers despite the bullish signal. For a negative prediction, a high count suggests it may outperform despite the bearish signal. Low counts on both = fewer contradictory signals.
Timestamp & Form Type
The date and time (UTC) at which the prediction was generated from the filing, and the SEC form type that triggered it. 10-K is the annual report, the most comprehensive filing with the broadest financial data. These tend to produce the most stable predictions. 10-Q is the quarterly report, with updated figures and a narrower scope. Useful for tracking momentum. 8-K is the current report, filed for material events such as earnings pre-announcements or major corporate actions. The list is sorted newest-first, so fresh predictions always appear at the top.