Your strategy finds signals. Stock Trends puts them in context.

Your strategy sees what it was built to see

Most strategies are, by design, narrow. They resolve one view of the market extremely well and are largely blind to the conditions surrounding it. That narrowness is usually a strength — but it leaves a gap.

• A momentum strategy sees momentum. It measures the move. It does not measure how many other names are participating in it.
• A value strategy sees valuation. It measures the discount. It does not measure whether the market is currently rewarding that discount.
• A machine-learning model sees its features. It resolves the relationships it was trained on, within the regimes present in its training data.
• Stock Trends sees the surrounding environment. Trend state and maturity, relative performance, breadth, leadership concentration and rotation, and regime structure — classified the same way for decades.

None of these views is superior to the others; they answer different questions. The practical value of a consistent, long-horizon classification layer is comparability: the same market states, encoded the same way, across many different market environments. That is expensive for any single team to build and maintain alongside its own strategy.

Stock Trends does not replace your strategy. It helps your strategy understand the market in which it is operating.

How the context is built

1. Step 1 · Input

   Weekly market observations

   Price and volume across NYSE, NASDAQ, AMEX, TSX, and index data, recorded on a weekly cadence.

2. Step 2 · Encoding

   Stock Trends classification framework

   One consistent doctrine that turns raw observations into named, comparable market states — applied the same way across decades.

3. Step 3 · Context layers
   • Instrument Trend classification, trend persistence, relative performance and direction, volume behaviour
   • Breadth & leadership Market and sector breadth, leadership concentration, leadership rotation
   • Market regime Regime structure and participation: bullish, bearish, transitional, or selective
   • Portfolio Evaluation, construction, comparison, and alignment with broader conditions
4. Step 4 · Applications
   • ST-IM Probabilistic forward-return distributions conditioned on classified states
   • Systematic strategies Declared rules and long-running model-portfolio histories
   • Portfolio workflows Evaluate, construct, and compare portfolios against market conditions
5. Step 5 · Delivery

   Trading systems, portfolio managers, and AI agents

   Structured, agent-native access over the Stock Trends API — the delivery mechanism for every layer above.

Applied evidence exposed through the API

The portal documents three evidence pillars: historical classification provenance, realized statistical outcomes, and strategy and model-portfolio evidence. The two applied evidence families below have dedicated API records that agents and researchers can inspect. The Evidence & Validation page sets out all three pillars in full, including the classification provenance that underpins them.

Stock Trends exposes its record as endpoints rather than as marketing claims, so the historical outcomes and the rules and provenance behind them can be read directly instead of treated as a black box. The two families answer different questions and should not be combined into a single performance number.

Realized statistical outcome evidence

Mature ST-IM Select realized outcomes — observations that have already completed their measurement window — reported across 4-week, 13-week, and 40-week forward horizons against base-period means. The default response covers a trailing ten-year window (2016-03-04 through 2026-02-27) and includes 156,868 mature 4-week, 156,889 mature 13-week, and 139,743 mature 40-week observations.

Average returns exceed the base-period mean at all three horizons, while the outperform-base rate at 13 and 40 weeks sits below 50% — the distribution is positively skewed rather than uniformly favourable. The Evidence page documents how to read that distribution and its limitations.

• GET /v1/selections/stim-select/outcomes/summary — mature realized outcome summary
• GET /v1/meta/stim — ST-IM base-period means and modelling assumptions

Strategy and model-portfolio evidence

Official Stock Trends model portfolios expose their metadata, declared strategy rules, return history, closed-position history, and portfolio-to-strategy provenance. Return histories for the longest-running model portfolios begin in 1981 and continue to the current week — a span that starts before the 1993 weekly publication record, so it should not be read as a contemporaneously published history.

• GET /v1/stocktrends/portfolios — list official Stock Trends model portfolios
• GET /v1/stocktrends/portfolios/{port_id} — portfolio metadata
• GET /v1/stocktrends/portfolios/{port_id}/returns — chronological return history
• GET /v1/stocktrends/portfolios/{port_id}/summary — public history summary and annualized ROI
• GET /v1/stocktrends/portfolios/{port_id}/positions/history — closed-position history
• GET /v1/stocktrends/strategies — strategy definitions and economic assumptions
• GET /v1/stocktrends/strategies/{strategy_id} — declared buy/sell conditions and stop-loss settings
• GET /v1/stocktrends/portfolios/{port_id}/strategy — portfolio-to-strategy provenance
• POST /v1/portfolio/compare — compare two portfolios under the same evaluation logic

Stock Trends model portfolios are rule-based model records with declared transaction-cost and stop-loss assumptions; current live holdings are excluded from public history endpoints. They are not audited brokerage-account performance, and they are not the same evidence family as ST-IM realized outcomes or Monte Carlo process simulations. Historical evidence supports framework evaluation and research — not investment advice, guaranteed outcomes, or price targets.

The agent journey, start to finish

A new agent should not have to reverse-engineer this sequence. Everything up to step 5 is public and credential-free, so an agent can understand the service and price the work before it spends anything.

1. Discover GET /v1/ai/tools — capability and workflow discovery
2. Understand GET /v1/ai/context — indicator definitions and dataset grounding
3. Inspect evidence GET /v1/ai/proof/market-edge — signal structure before payment
4. Resolve GET /v1/workflows and GET /v1/instruments/resolve
5. Estimate cost GET /v1/pricing/catalog and GET /v1/cost-estimate
6. Choose a surface REST API for services and systems, MCP for agent environments
7. Choose a rail Subscription, standard x402 V2, or MPP funded sessions
8. Execute GET /v1/agent/screener/top — a recommended first premium call

ST-IM in context

ST-IM is one application of the broader Stock Trends classification framework, not the whole of it. Its contribution is probabilistic forward-return context: return distributions conditioned on how an instrument is currently classified.

Like any single layer, it is most useful read alongside the others — breadth, leadership, regime, and portfolio context — rather than on its own. Detailed methodology, probability interpretation, strengths, and limitations are set out on the ST-IM page.

Important disclosures

Stock Trends provides structured market context and probabilistic research intelligence. Its outputs are not investment advice, price targets, direct buy/sell commands, or guarantees of future performance, and they are not a substitute for your own investment process, risk controls, or professional advice.

Historical depth does not eliminate uncertainty. The three evidence pillars — historical classification provenance, realized statistical outcome evidence, and strategy and model-portfolio evidence — rest on different methodologies and cover different time spans, as do the supporting weekly publication record and Monte Carlo process simulations. None of them should be combined into a single performance claim. Historical relationships may weaken or change in future market regimes, and real-world implementation involves transaction costs, taxes, execution slippage, liquidity constraints, and market risk, including the risk of loss.

Stock Trends model portfolios are rule-based model records with declared cost and stop-loss assumptions. They are not audited brokerage-account performance.