Home / Research Methodology, Data Sources & Limitations
Know what the research can—and cannot—establish
Costa Rica Trees publishes research tools with explicit data and validation limits. A deterministic calculation can be correct without providing evidence that an investment strategy will outperform.
Free during the research beta. Email confirmation required.
Sources and availability
Company research uses SEC filings and identified issuer disclosures where available. Market observations come from the providers named in the relevant workspace. Data may be delayed, stale or incomplete. Source links, reporting periods and availability dates are part of the evidence; an unavailable disclosure must not be silently replaced with zero.
Scoring is versioned, not a promise
Value-investing scores use fixed, explicitly versioned weights. Unknown inputs earn no contribution and missing weights are not redistributed. The current weights remain unvalidated. Screening organizes research; it does not estimate the probability of a profitable investment or establish suitability for an individual.
Historical evidence stays frozen
Paper cohorts retain their original selection rules, membership and historical scoring versions. A cohort selected under an older model cannot cleanly validate selection under a corrected model. The 15% price-trigger experiment is separate from stock-selection research and from net returns after modeled costs. These are hypothetical paper records, not brokerage executions.
Dividends and corporate actions
Yahoo events are discovery candidates. SEC, issuer and fund disclosures provide terms for review. A missing event in a feed does not prove that no event occurred. Total-return publication requires reviewed event terms, complete holdings and benchmark coverage, compatible raw prices and archived source evidence. Until those requirements are met, displayed paper results are price-only and verified total returns are withheld.
What would count as investment evidence?
The evaluation standard separates corrected-model buy-and-hold versus SPY from the exit rule versus holding the identical stocks. It calls for prospective registration, cost sensitivity, dependence-aware uncertainty, multiple market conditions and independent review. Existing cohorts are exploratory. Necessary evidence floors include five calendar years, 1,260 completed sessions and three separately registered cohorts; meeting a floor alone is not sufficient validation.
Review and corrections
This material is maintained by the Costa Rica Trees product team. Review the dated methodology, score version and source notes in the app. Financial assumptions and unsupported evidence should remain visible. Software tests check behavior under specified inputs; their count does not validate a scoring model, prove a durable competitive advantage or establish future investment returns.
Primary source references
SEC filing access and reuse · Investor.gov: dividend entitlement · SPY official fund disclosures
Start with evidence. Keep your questions organized.
Create an account to use the research workspace and save a watchlist. Confirm your email, acknowledge the research-beta terms and begin exploring the available data.
Create free account