On-Chain Analytics, Valuation and Decision-Making

Introduction and Context

Blockchain networks produce a distinctive form of public, time-stamped data. Transactions, contract calls, token transfers, liquidity positions and governance actions are recorded on-chain and can often be inspected without relying on a company’s private reporting system. This transparency creates new possibilities for financial analysis, operational monitoring and managerial decision-making. It does not, however, make analysis automatically reliable. Blockchain data can be incomplete, pseudonymous, difficult to interpret and distorted by automated activity.

On-chain analytics is the systematic collection, transformation and interpretation of blockchain data to answer questions such as:

• Who controls an address, and what evidence supports that conclusion?
• How does value move through a protocol or ecosystem?
• Are users growing, returning and performing economically meaningful activity?
• Is liquidity deep enough to support trading or withdrawals?
• Does protocol revenue justify a token’s valuation?
• How might a project perform under adverse market, technical or regulatory conditions?

This chapter develops a practical analytical framework using explorers, dashboards and query tools. It connects descriptive metrics to valuation, scenario analysis, stress testing and risk-adjusted performance measures. The central principle is that a metric is useful only when its definition, data source, time period and limitations are understood.

Learning Objectives

By the end of this chapter, students should be able to:

1. Explain the structure and limitations of on-chain data.
2. Use explorers, dashboards and query tools to investigate addresses, transactions and contracts.
3. Distinguish users, transactions, active addresses, volume, liquidity, fees and protocol revenue.
4. Trace token flows while accounting for bridges, routers, wrappers, exchanges and smart contracts.
5. Apply basic valuation models to protocols and tokens.
6. Construct scenarios and stress tests for liquidity, revenue and token value.
7. Calculate and interpret return, volatility, drawdown, Sharpe ratio and related risk metrics.
8. Present evidence-based investment or managerial recommendations while addressing privacy, manipulation and governance risks.

1. The Analytical Object: Blockchain Data as a Ledger of Events

A blockchain record is primarily an event record, not a conventional accounting database. A transaction may call a contract, and that contract may emit several token-transfer events. One user action can therefore generate many records. Conversely, economically important activity may occur off-chain, through a centralised exchange, custody service, order book or private agreement, and may not be visible until assets enter or leave the chain.

1.1 Layers of observable information

A practical investigation usually combines four layers:

The distinction between an address and an entity is essential. An address is a technical identifier. An entity may control many addresses, and one address may be a smart contract used by thousands of people. Explorer labels are helpful but are not proof of ownership. Attribution should be treated as a confidence judgement supported by evidence such as verified contract code, public disclosures, repeated operational patterns and known deposit or withdrawal addresses.

1.2 Explorers, dashboards and query tools

A blockchain explorer is generally best for investigating an individual address, transaction, block or contract. It can show transaction status, gas expenditure, token transfers, internal calls and contract verification. Dashboards aggregate data into time series and charts, making them useful for comparing protocols, chains and periods. Query tools provide greater flexibility: analysts can filter event logs, join tables, calculate cohorts and create reproducible datasets using SQL or equivalent languages.

A sound workflow is:

1. Define the economic question and measurement period.
2. Identify the relevant chain, contracts, tokens and address labels.
3. Inspect representative transactions manually.
4. Query or export the underlying records.
5. Clean duplicate, failed, spam and zero-value records.
6. Reconcile totals against an independent source where possible.
7. Interpret the metric in its economic and governance context.
8. Record the query, assumptions, block range and data timestamp.

This workflow prevents a common error: selecting a convenient dashboard metric before deciding what the metric is intended to measure.

2. Address and Transaction Interpretation

2.1 Reading a transaction

A transaction investigation should distinguish at least the following:

• Originator: the address that signed and submitted the transaction.
• Target: the address called by the transaction; it may be a user, contract, router or multisignature wallet.
• Native value: the blockchain’s native asset transferred directly.
• Token movements: transfers emitted by contracts during execution.
• Execution status: success or failure.
• Gas used and gas price: the computational cost paid by the originator.
• Method and parameters: the intended contract function and its arguments.
• Block time: the approximate ordering time, subject to chain-specific finality.

A successful transaction does not necessarily mean that the intended economic outcome occurred. A swap can succeed while receiving a poor price; a bridge transaction can succeed on one chain while its destination execution remains pending; and a contract can transfer assets to an unexpected address if the user approved a malicious spender.

For a transaction with gas consumption G and gas price p, the fee in native units is:

Here C_g is the gas cost, G is gas used, and p is the price per unit of gas. If the gas price is expressed in a smaller denomination, such as gwei, it must be converted before multiplying by a native-token price. The fiat cost is:

where P_N is the fiat price of the native asset at the selected valuation time. This is a realised transaction cost, not necessarily the total economic cost, because slippage, approval transactions and opportunity cost may also matter.

2.2 Failed transactions and duplicated economic events

Failed transactions usually consume gas but do not complete the intended state change. Analysts should not count them as successful swaps, deposits or user actions. However, they may be relevant to user experience and network demand. Event logs should be filtered using transaction status and contract-specific rules.

Aggregators and routers create another problem. One swap may pass through several pools and produce multiple transfer events. Summing all transfer events can double-count volume. A defensible volume definition identifies the economic boundary—for example, the amount entering and leaving a user-facing swap—or uses a protocol’s documented trade event rather than every intermediate movement.

3. Measuring Activity, Users, Liquidity and Revenue

3.1 Activity metrics

Important metrics include:

• Transactions: submitted or successful transactions, depending on the definition.
• Active addresses: distinct addresses interacting during a period.
• New addresses: addresses observed for the first time under a stated rule.
• Returning addresses: addresses active in both a current and previous period.
• Volume: value exchanged, transferred or settled under a specified boundary.
• Retention: the proportion of a cohort that returns in later periods.

An address is not necessarily a person. One user can control several wallets; one exchange or bot can control many addresses. Consequently, address-based user metrics are proxies rather than direct population counts.

For a cohort of N₀ addresses first active in period 0, retention in period t is:

where N_t is the number of cohort addresses active in period t. The formula assumes that the cohort definition and activity rule remain consistent. Retention should not be compared across dashboards using different definitions of “active.”

3.2 Liquidity and total value locked

Liquidity may refer to pool reserves, lending supply, borrowable assets, staked assets or assets deposited in a protocol. Total value locked (TVL) is generally calculated by valuing eligible units at a selected price:

Here q_i,t is the quantity of asset i at time t, and P_i,t is its valuation price. TVL is therefore sensitive to both net deposits and price changes. An increase in TVL does not necessarily indicate new capital, and a decrease may result from falling token prices rather than withdrawals.

For a constant-product automated market maker, the idealised invariant is:

where x and y are reserve quantities and k is constant during a fee-free swap. This model assumes a particular pool design and does not fully represent concentrated liquidity, fees, oracle-based markets or dynamic-weight pools. Price impact depends on pool depth relative to trade size; therefore, TVL alone is not a sufficient measure of executable liquidity.

3.3 Fees, revenue and earnings quality

A protocol may collect gross user fees, distribute part of them to liquidity providers, pay incentives, incur operating costs and direct a remainder to a treasury or token holders. These are not interchangeable concepts.

A useful decomposition is:

The exact definition must follow the protocol’s economic design. For example, a lending protocol’s interest paid by borrowers may be divided between suppliers, reserves and a treasury. A token buyback funded by revenue is a use of revenue, not automatically additional revenue.

An analyst should report at least: gross fees, protocol revenue, revenue recipient, incentives, one-off items and the period used. Annualising a short period can be misleading, especially during unusual market conditions.

4. Token-Flow Analysis

Token-flow analysis follows assets between addresses, contracts and chains. It is useful for treasury monitoring, concentration analysis, bridge risk, liquidity assessment and investigation of unusual activity.

A basic flow table records:

Net flow into an address or entity over a period can be represented as:

where F_in and F_out are measured in the same asset and period. If multiple assets are combined, each must first be converted using a stated price basis; otherwise the result has no coherent unit.

Interpretation requires caution. A transfer to an exchange does not prove selling, and a transfer from an exchange does not prove buying. A market maker may move assets for inventory management. A bridge may lock an asset on one chain and mint a representation on another, creating apparent outflows and inflows that are economically linked. Wrapped tokens, rebasing tokens, liquidity-provider tokens and vault shares also require protocol-specific treatment.

Concentration is often summarised using a Herfindahl–Hirschman Index:

where s_i is entity i’s share of a chosen population, expressed as a proportion between 0 and 1. A higher HHI indicates greater concentration. The result depends entirely on the classification: counting contracts rather than controlling entities may understate concentration, while treating exchange omnibus wallets as one holder may overstate it.

5. Valuation Frameworks for Protocols and Tokens

Valuation is not the same as price prediction. It is an explicit relationship between assumptions and an estimated value. Blockchain assets require particular care because governance rights, fee rights, token supply and protocol usage may not be legally or economically connected.

5.1 Market capitalisation and fully diluted valuation

Circulating market capitalisation is commonly approximated by:

where S_c is circulating supply and P_T is token price. Fully diluted valuation is:

where S_f is the specified fully diluted supply. These measures are meaningful only if supply definitions are consistent. Locked, vested, bridged, burned and treasury-held tokens may be treated differently by different data providers.

A high FDV relative to current market capitalisation may indicate substantial future unlocks, but it is not by itself evidence of overvaluation. The effect depends on demand growth, vesting schedules, liquidity and the economic rights attached to the tokens.

5.2 Cash-flow and fee-based approaches

If token holders have a credible claim on future distributable cash flows, a discounted cash-flow model may be appropriate:

Here V₀ is present value, CF_t is distributable cash flow in period t, r is the discount rate, T is the explicit forecast horizon and TV_T is terminal value. The formula assumes cash flows are defined consistently, the discount rate reflects relevant risk and terminal value is not double-counted.

A perpetual-growth terminal value can be written as:

where g is the perpetual growth rate. This special case requires r > g and is highly sensitive to both assumptions. It should not be used when cash flows are unstable, rights are uncertain or perpetual growth is economically implausible.

When token holders do not receive cash flows, a fee-based multiple may be used descriptively:

The metric might be revenue, fees or earnings. Annualisation should be based on an economically representative period, and comparisons require similar token rights, protocol maturity, chain costs and accounting definitions.

5.3 Network and usage indicators

Price-to-sales-style ratios, price-to-fees ratios, token velocity and user growth can support a valuation narrative, but none is a universal valuation law. Velocity may reflect payment usage, speculation, treasury transfers or market-making. Active addresses may be inflated by bots. A responsible valuation combines financial, operational, market-structure and governance evidence.

6. Scenario Analysis and Stress Testing

Scenario analysis changes a set of assumptions coherently. Stress testing examines the effect of adverse but plausible shocks. Both are preferable to presenting a single precise forecast.

A simple protocol-value scenario may be represented by:

where M is a selected valuation multiple and A is the relevant annual metric. This is not a fundamental law; it is a scenario model. A more complete scenario table could vary users, volume, fee rate, revenue share, token price, discount rate and liquidity.

For a lending protocol, an adverse stress test might include a 30% collateral-price decline, increased liquidations, a 40% reduction in borrowing demand, a stablecoin depeg and a specified withdrawal rate. The analyst should model the order of events rather than simply subtracting percentages. For example, a price decline can reduce collateral value, trigger liquidations, increase slippage and reduce the value of protocol reserves.

Useful stress-test outputs include:

• reserve coverage after losses;
• proportion of withdrawals that can be met without severe price impact;
• collateralisation or utilisation after the shock;
• treasury runway under lower revenue;
• token unlocks relative to average market liquidity;
• bridge or oracle dependencies exposed by the scenario.

Sensitivity analysis varies one assumption while holding others constant. If a model’s result changes dramatically when the discount rate moves from 20% to 25%, that sensitivity should be disclosed rather than hidden behind a point estimate.

7. Risk-Adjusted Performance and Decision Metrics

A decision should consider return relative to risk, liquidity and the investor’s benchmark.

For an investment with initial value V₀, final value V_T and cash distributions D, simple holding-period return is:

This assumes values and distributions are measured consistently and ignores interim timing. For multiple periods, the geometric annualised return is:

where T is the number of years. This is appropriate for a compounded investment without external cash flows; money-weighted methods are more suitable when deposits and withdrawals occur at different times.

Volatility is commonly estimated as the standard deviation of periodic returns. For observations R₁, R₂, …, R_n, the sample standard deviation is:

where R̄ is the arithmetic mean and n is the number of observations. The annualisation rule depends on the return frequency and assumptions about independence; multiplying daily volatility by √252 is a convention for approximately 252 trading observations, not a universal blockchain law.

The Sharpe ratio is:

where R_p is portfolio return, R_f is the risk-free or benchmark return over the same period, and σ_p is portfolio volatility. The ratio is informative only when the comparison period, compounding convention and risk-free proxy are consistent. It may be unreliable for assets with illiquid prices, non-normal returns or stale observations.

Maximum drawdown measures the largest peak-to-trough decline. If W_t is portfolio value and P_t = max(W₀,…,W_t), drawdown is:

Maximum drawdown is the minimum value of D_t over the period. It captures path risk that volatility may obscure. Other useful measures include downside deviation, liquidity-adjusted return, value-at-risk and expected shortfall, but these require careful distributional and liquidity assumptions.

8. An Applied Illustrative Case: Evaluating the Meridian Lending Protocol

The following is a fictional illustrative case designed for educational use.

Background

Meridian is a lending protocol deployed on two compatible smart-contract networks. It allows users to supply stablecoins and borrow against liquid crypto-assets. Its governance token, MRD, has a circulating supply of 80 million and a market price of $1.50. The protocol dashboard reports $120 million TVL and $18 million of quarterly gross fees.

Problem

An investment committee must decide whether to purchase MRD and whether Meridian should expand to a third network. The committee is concerned that reported activity may include incentives, duplicated addresses and temporary deposits.

Analysis

The team first inspects contracts and confirms that the dashboard includes supplied assets but excludes outstanding loans. It separates stablecoin supply, collateral supply and borrowed assets. It then queries successful supply, borrow, repay and withdraw events, removes internal treasury transfers and groups known exchange addresses separately from user addresses.

The investigation finds:

• quarterly gross fees of $18 million;
• $11 million distributed to suppliers;
• $2 million paid in incentives and rebates;
• $5 million retained by the protocol before operating costs;
• 42,000 active addresses, but only 9,800 addresses returning in the following quarter;
• 61% of supplied value associated with the ten largest attributed entities;
• 35% of TVL attributable to a temporary liquidity-mining programme;
• a third-network bridge that depends on one external messaging system.

The protocol’s net retained quarterly revenue is:

If the committee uses a simple four-quarter annualisation, retained revenue is $20 million. This is a scenario assumption, not a guaranteed forecast. The token’s market capitalisation is:

The resulting market-capitalisation-to-annualised-retained-revenue ratio is:

However, MRD holders have no contractual claim to all retained revenue. Governance may change the distribution policy, and expansion may increase costs. Therefore, the ratio is only a screening indicator, not proof of fair value.

Scenario approach

The committee models three cases:

The downside case also assumes a 25% fall in collateral prices and a 15% withdrawal of supplied stablecoins. The risk team tests whether liquid assets and reserves can meet withdrawals without disorderly liquidation. It also tests a bridge outage. The result is that Meridian remains solvent under the modelled price shock, but expansion increases dependence on the external messaging system and reduces operational simplicity.

Approach and decision

The committee does not reject Meridian solely because retention is lower than active-address growth. Instead, it treats the retention result as evidence that incentives may be acquiring short-lived users. It requests disclosure of entity concentration, incentive-adjusted revenue, oracle dependencies and bridge controls. It approves a limited pilot on the third network, subject to withdrawal limits, independent monitoring and a reserve policy.

Outcome and learning points

The case demonstrates five principles. First, TVL must be decomposed into organic and incentivised capital. Second, gross fees are not protocol revenue. Third, address counts are not unique human-user counts. Fourth, valuation multiples depend on token rights and forecast quality. Fifth, a decision should combine base-case valuation with operational and technical stress tests.

9. Governance, Ethics and Regulatory Considerations

On-chain analytics can expose financial behaviour, even when addresses are pseudonymous. Analysts should minimise unnecessary publication of personal information, avoid doxxing and distinguish public evidence from speculation. Data protection obligations may apply when addresses are linked to identifiable individuals.

Analytics can also be used for sanctions screening, fraud detection and market surveillance. Such uses require documented procedures, false-positive review and escalation controls. An address label should not be treated as conclusive evidence of criminal conduct. Automated risk scores can reproduce classification errors, especially where exchange omnibus wallets, mixers, privacy tools or shared custody obscure ownership.

Governance analysis should examine who can upgrade contracts, pause withdrawals, change fees, mint tokens or control oracles. A protocol that appears decentralised in transaction volume may retain substantial administrative control in a small multisignature wallet. Key questions include signer concentration, timelocks, emergency powers, voting participation, delegation and the practical ability of token holders to challenge decisions.

Regulatory treatment varies by jurisdiction and may depend on token rights, marketing, custody, lending, derivatives, money transmission and consumer-protection rules. On-chain metrics do not determine legal classification. Professional analysis should therefore involve qualified legal and compliance specialists when decisions have regulatory consequences.

10. Limitations, Data Quality and AI-Assisted Analysis

The principal limitations of on-chain analytics are:

• Pseudonymity: addresses do not directly identify people.
• Sybil activity: one actor can create many addresses.
• Bots: automated trading can inflate transactions and volume.
• Incomplete visibility: off-chain activity and private transactions may be absent.
• Cross-chain fragmentation: economic ownership may be distributed across networks.
• Price dependence: TVL and market capitalisation change with volatile prices.
• Label error: third-party entity classifications may be outdated or wrong.
• Protocol complexity: event logs may not represent economic substance without code-level interpretation.
• Data revisions: indexing providers can correct historical data or change definitions.

AI tools can help classify transactions, generate SQL, detect anomalies and summarise governance proposals. They should not be treated as authoritative interpreters of smart-contract behaviour. A professional control process requires human review of contract code, query logic, sample transactions, data provenance and unusual outputs. Analysts should preserve prompts, code versions and validation results when AI contributes to a decision.

11. Best-Practice Decision Framework

A robust on-chain decision memo should contain:

1. Question: the precise investment or management decision.
2. Scope: chains, contracts, assets, entities and dates included.
3. Definitions: active user, volume, revenue, TVL and supply.
4. Data provenance: explorer, dashboard, query, block range and retrieval date.
5. Reconciliation: comparison with protocol reports or independent datasets.
6. Economic interpretation: what the metric does and does not mean.
7. Valuation: model, assumptions, rights and sensitivity analysis.
8. Stress test: shocks, sequencing, liquidity and operational dependencies.
9. Risk assessment: technical, market, governance, regulatory and privacy risks.
10. Recommendation: action, conditions, monitoring triggers and decision owner.

The most useful output is rarely a single score. It is a transparent chain of evidence from raw event to metric, from metric to model, and from model to decision.

Chapter Summary

On-chain analytics converts blockchain events into evidence for financial, operational and governance decisions. Explorers support transaction-level investigation; dashboards support comparison and monitoring; query tools support reproducible analysis. Analysts must distinguish addresses from people, transactions from economic actions, TVL from liquidity, gross fees from protocol revenue and token price from token value.

Core quantitative tools include gas-cost calculation, retention, TVL decomposition, net flows, concentration measures, market capitalisation, FDV, discounted cash flow, valuation multiples, scenario analysis, stress testing, return, volatility, Sharpe ratio and maximum drawdown. Each metric depends on definitions, assumptions and data quality.

A professional analysis combines on-chain evidence with contract interpretation, entity attribution, market data, governance review and risk assessment. Privacy, false attribution, manipulation, cross-chain complexity and regulatory uncertainty must be addressed explicitly. The final decision should be conditional, reproducible and sensitive to adverse scenarios rather than based on an attractive dashboard chart alone.

Key Terms

Active address: An address that satisfies a stated interaction rule during a specified period.

Address attribution: The process of associating a technical address with an entity or role.

Automated market maker: A protocol that uses programmed liquidity and pricing rules to facilitate exchange.

Cohort retention: The proportion of a defined group that remains active in later periods.

Discounted cash flow: A valuation method that converts expected future cash flows into present value.

Entity: A person, organisation, exchange, treasury or other economic controller inferred from one or more addresses.

FDV: Fully diluted valuation based on a specified fully diluted token supply.

Gas fee: The computational fee paid for executing a blockchain transaction.

Gross fees: Total fees paid by users before distributions and pass-through costs.

Liquidity: The ability to trade, withdraw or exchange assets with limited price impact.

Market capitalisation: Token price multiplied by a specified circulating supply.

Maximum drawdown: The largest peak-to-trough decline in portfolio value over a period.

Net flow: Inflows minus outflows for a defined asset, entity and period.

On-chain analytics: Systematic analysis of blockchain records and related datasets.

Protocol revenue: The portion of economic fees retained by a protocol under a stated definition.

Sharpe ratio: Excess return divided by return volatility, subject to consistent measurement assumptions.

Stress test: Analysis of performance under an adverse but plausible shock.

Total value locked: The value of eligible assets deposited or controlled by a protocol under a stated methodology.

Token-flow analysis: Examination of asset movements among addresses, contracts and chains.

Valuation multiple: Market value divided by a selected financial or operational metric.

Multiple Choice Questions with Answers

Question 1

Which statement best describes an address on a public blockchain?

A. It always identifies one human user

B. It is a technical identifier that may represent a contract, entity or wallet

C. It is legally equivalent to a registered company

D. It records only successful economic transactions

Correct Answer: B. It is a technical identifier that may represent a contract, entity or wallet

Question 2

A transaction uses 100,000 gas at 20 gwei. What is the gas cost in the native asset, using 1 gwei = 10⁻⁹ native units?

A. 0.000002 native units

B. 0.0002 native units

C. 0.002 native units

D. 2 native units

Correct Answer: C. 0.002 native units

Question 3

Why can summing every token-transfer event overstate decentralised-exchange volume?

A. Failed transactions never consume gas

B. Routers and multi-hop swaps can generate several intermediate transfers for one economic trade

C. Token transfers are not recorded on-chain

D. Liquidity pools cannot emit events

Correct Answer: B. Routers and multi-hop swaps can generate several intermediate transfers for one economic trade

Question 4

If a protocol has 80 million circulating tokens priced at $1.50, its approximate circulating market capitalisation is:

A. $12 million

B. $53.3 million

C. $120 million

D. $800 million

Correct Answer: C. $120 million

Question 5

Which item is most directly included in a total-value-locked calculation?

A. The number of governance proposals

B. Eligible asset quantity multiplied by a stated valuation price

C. All failed transaction fees

D. The protocol’s social-media followers

Correct Answer: B. Eligible asset quantity multiplied by a stated valuation price

Question 6

A protocol reports $18 million in gross fees, distributes $11 million to suppliers and pays $2 million in incentives. Before other costs, retained revenue is:

A. $5 million

B. $7 million

C. $11 million

D. $31 million

Correct Answer: A. $5 million

Question 7

What is a central limitation of using active addresses as a measure of users?

A. Addresses cannot interact with contracts

B. One user may control multiple addresses and automated actors may create activity

C. Active addresses include only off-chain users

D. Every address is permanently linked to a legal identity

Correct Answer: B. One user may control multiple addresses and automated actors may create activity

Question 8

In a discounted-cash-flow model, the discount rate primarily reflects:

A. The number of blockchain confirmations

B. The risk and time value associated with expected future cash flows

C. The token’s total supply only

D. The number of contract events

Correct Answer: B. The risk and time value associated with expected future cash flows

Question 9

Which metric focuses specifically on peak-to-trough portfolio loss?

A. Gross margin

B. Retention rate

C. Maximum drawdown

D. Fully diluted valuation

Correct Answer: C. Maximum drawdown

Question 10

Why should a bridge outage be included in a cross-chain protocol stress test?

A. It can reveal dependence on external messaging and restrict movement or settlement of assets

B. It guarantees that token prices will rise

C. It eliminates all governance risk

D. It converts gross fees into protocol revenue

Correct Answer: A. It can reveal dependence on external messaging and restrict movement or settlement of assets