TakaCompassটাকা কম্পাস
DEMO DATA — NOT LIVEFictional educational examples · not current market guidance

Evidence-first investing

Educational guide

Build an investment note from dated sources, explicit assumptions, contrary evidence, and a review condition.

Start with a source you can identify and date. Separate what it says from what you infer, record what could change your view, and leave room for no action. A completed checklist is a record of a process, not a signal or an expected result.

Editorial update: 3 min readRelease evidenceIndependent financial review not recorded
Evidence path from source to reviewFour linked layers show how an identified source becomes dated context, a testable reading and a later review.01020304
A traceable learning sequence; later layers do not make the source more certain.
  1. Identify the original document or dated observation and keep its source.
  2. Add period, units, scope, capture time and what remains unknown.
  3. Compare the interpretation with contrary evidence and an observable condition.
  4. Add a new dated note when the review date or evidence trigger arrives.

Start with a dated source

For an issuer fact, begin with a dated company filing or exchange notice. Use a regulator or exchange page to check the document and security identity. A calculation can summarize a source, while commentary and a chart can suggest questions; neither replaces the underlying record. Keep the source, effective period, retrieval date, currency, unit, scope, validation status, and any transformation version beside each material observation.

Write an observation in neutral words before adding an interpretation. For example, record the reported figure and period first; label a possible explanation as an interpretation or assumption. If a field is absent, stale, cropped, or cannot be matched to the issuer, say it is unknown. A screenshot is a dated image of a chart, not a live feed or proof of the data behind its pixels.

Test both sides and define a review

Record contrary evidence—the strongest evidence against your interpretation—as carefully as the evidence for it. State an observable condition that would weaken or invalidate the idea, and name the next dated document or event you would check. Waiting, monitoring, or deciding that the evidence is insufficient are valid outcomes. Do not replace an unknown with a confident-sounding guess.

Keep a dated snapshot of the original reasoning. Add a later review as a new note with its own source, date, and observation; do not rewrite the earlier note to make it agree with what happened later. Review the reasoning separately from the outcome, and note when a planned review date or evidence trigger arrives rather than creating urgency around a price move.

Use prototype methods only as a checklist

The DSE Frontier prototype is a paper-only research project. Its examples can prompt questions about prior-volume context, liquidity, dated inputs, and unknown source status. Its internal conviction field is a model score based on proxy labels; it is not a probability of profit, a calibrated confidence measure, a DSE rule, or an order. TakaCompass uses no Frontier rank or score as an investor outcome.

For any learning exercise, inspect source identity, missing fields, the comparison period, and what would contradict the reading. Treat prototype outputs as research artifacts with their own limits. This educational guide and any demo example marked `DEMO DATA — NOT LIVE` do not connect to an exchange or execute an order.

Change one assumption. See what follows.

From reading to practice

Try the idea with fictional amounts. Keep the calculation and its limitations in view.

Open the learning toolReview assumptions in Panchayat