Applied FTE workflow: Blockchains, cryptoassets, coins, tokens, stablecoins and DeFi

Learning objectives

• Explain definitions, network role and asset category using a defined system and time boundary.
• Evaluate provenance, completeness and uncertainty for evidence about Blockchains, cryptoassets, coins, tokens, stablecoins and DeFi.
• Produce a reviewable FTE source pack without advice-style conclusions.

Applied FTE workflow: specialist subject brief

Build an FTE identity card that records network role, issuer or protocol, contract where applicable, settlement dependency and category-specific failure modes.

Applied FTE workflow: the research boundary

Blockchains, cryptoassets, coins, tokens, stablecoins and DeFi is not a single data point; it is a research boundary. The first professional task is to identify which system, asset, venue, contract, period and stakeholder the claim concerns. In this module the governing evidence focus is definitions, network role and asset category. A precise boundary prevents a familiar error: combining observations that share a label but not a definition, timestamp or scope.

Use plain language, define every technical term and show each arithmetic step. The learner should be able to explain what the evidence can establish, what must be calculated, and what remains interpretation. This separation is essential because a technically correct figure can still mislead when its coverage, unit, observation window or dependency is omitted.

Method: collect before interpreting

Begin with the module source pack. Record publisher, document title, stable URL, publication or update date, retrieval time and the exact claim supported. For definitions, network role and asset category, capture raw observations before creating ratios or narrative. If an interface supplies a derived number, locate its methodology or mark the transformation as provider-defined rather than silently treating it as raw evidence.

Apply the FTE source pack: one row or node per material claim, linked to its source and calculation. Evidence quality improves when identity, time, units and lineage are explicit. It does not improve merely because several websites repeat the same upstream value. Shared provenance is one evidence chain, not independent confirmation.

Interpretation and uncertainty

Assess completeness separately from confidence. A source may be authoritative for the fields it publishes yet incomplete for the research question. Conversely, broad coverage may contain uncertain labels or inferred classifications. For module 1, document both dimensions and explain the consequence of each gap instead of converting uncertainty into a false numerical precision.

When sources disagree, preserve the conflict. Compare definitions, as-of times, chain coverage, venue coverage and transformation rules. Select a value only when the selection rule is defensible; otherwise present the range or unavailable state. The analyst's job is to make disagreement inspectable, not to hide it behind an average.

Applied FTE control

In the FTE workflow, each material statement receives an evidence label: observed, calculated, inferred, historical, fictionalised, unavailable or provider-gated. The output also records source, data timestamp, retrieval timestamp, method and warning state. These labels let a reviewer distinguish the facts collected from the analyst's synthesis without reading every working note.

The final narrative must remain non-advisory. Describe source-backed data points, historical patterns and evidence-based risk indicators. Do not convert the exercise into a trade instruction, unsupported forecast or claim of certainty. Correlation does not prove causation, and an absence of recorded incidents does not prove an absence of risk.

Quality assurance and hand-off

Before hand-off, reproduce one calculation from the recorded inputs, test one alternative definition and verify every citation resolves to the supported claim. Ask whether a reader could recreate the result without private context. If not, add the missing assumption, formula, field definition or evidence reference rather than relying on analyst memory.

Conclude with three lists: established observations, bounded interpretations and open questions. This structure creates an auditable stopping point and directs the next research action. A professional note is not one that sounds certain; it is one whose evidence, limitations and review status are proportionate to the decision context.

Measurement design and sensitivity

A measurement is useful only when its construction matches the question. For definitions, network role and asset category, write the numerator, denominator, aggregation rule, inclusion criteria and time convention before calculating. Note whether the value is a stock, a flow, a point-in-time balance or a rolling estimate. These distinctions determine which comparisons are valid and which apparent movements may be artefacts of the method.

Run a bounded sensitivity check. Change one defensible assumption, such as the observation window, coverage threshold or treatment of missing records, and compare the result. The purpose is not to manufacture a range of outcomes. It is to show whether the interpretation depends on a fragile choice and to identify the assumption that a reviewer should challenge first.

Risk, dependencies and counter-evidence

Map dependencies that sit between the source observation and the claim: index construction, exchange or oracle coverage, address labelling, contract upgrades, bridge accounting, API caching and analyst transformation. A dependency can introduce delay, duplication or classification error even when the upstream publisher is credible. Record the dependency owner and the failure state that would invalidate the result.

Actively search for counter-evidence. Ask what observation would weaken the proposed interpretation of Blockchains, cryptoassets, coins, tokens, stablecoins and DeFi, then check whether the source pack can observe it. If it cannot, record that blind spot in the risk register. Professional scepticism means designing a test that could change the conclusion, rather than collecting only material that confirms the first narrative.

Review cadence and change control

Evidence expires at different speeds. Transaction data may require a fresh as-of time; protocol documentation changes after upgrades; regulatory material follows formal publications and guidance updates. Assign each source a next-review date and an immediate-review trigger. For definitions, network role and asset category, an identity change, methodology revision or material dependency event should reopen the evidence record before reuse.

Version the resulting note, workbook and assessment references together. A correction should preserve the earlier record, state what changed and explain whether the change affects any material claim. This protects certificate compatibility and allows another reviewer to reconstruct what the learner knew at the recorded time rather than silently judging historical work against later information.

Professional review questions

A reviewer should be able to ask: Is the identity canonical? Are units and timestamps comparable? Does the cited source support this exact claim? Is the transformation reproducible? Are confidence and completeness justified independently? Are conflicts visible? Does the language distinguish historical observation from inference? Each answer should point to a workbook field or evidence record, not rely on an undocumented conversation.

The learner's final check is proportionality. The strength and prominence of a statement should not exceed the evidence behind it. Where evidence is partial, the limitation belongs beside the claim rather than in a remote footnote. Where evidence is absent, the correct output is unavailable or an open research question. That discipline is the foundation of trustworthy FTE research at every course level.

Fictionalised review: Blockchains, cryptoassets, coins, tokens, stablecoins and DeFi

fictionalised data · as at 2026-06-30

A fictional analyst receives two observations about definitions, network role and asset category: Source A covers the full period but publishes a weekly value; Source B is current to the hour but covers only part of the relevant market.

The analyst records both observations, labels their different coverage, reproduces the comparable calculation and leaves the combined result unavailable because the datasets cannot be reconciled without an unsupported assumption.

Common errors and evidence-quality controls

FTE application

• Create claim-level evidence records.
• Apply observed, calculated, inferred and unavailable labels.
• Run the non-advisory language gate before publication.