On-Chain Analytics

When to Activate

• Analyzing blockchain data for trading signals (whale movements, exchange flows, network health)
• Evaluating crypto asset fundamentals using on-chain metrics
• Building or interpreting dashboards from Glassnode, Nansen, Dune Analytics, or similar platforms
• Assessing market structure through wallet clustering and entity behavior

Core Concepts

Blockchain Data Layers

1. Raw transaction data — inputs, outputs, gas, timestamps, contract interactions
2. Derived metrics — aggregated statistics computed from raw data (active addresses, transfer volume)
3. Entity-level analytics — clustering addresses into wallets, labeling exchanges, funds, whales
4. Behavioral signals — interpreting entity actions as bullish/bearish indicators

Key On-Chain Metrics

Supply Metrics:

• Circulating supply — tokens that have moved within a lookback window (e.g., 1 year)
• Realized cap — sum of each UTXO valued at the price when it last moved; smooths out speculative premium
• MVRV ratio — Market Value / Realized Value; above 3.5 historically signals overheating, below 1.0 signals deep value
• NUPL (Net Unrealized Profit/Loss) — aggregate unrealized gains across all holders; >0.75 = euphoria, <0 = capitulation

Network Activity:

• Active addresses — unique sending/receiving addresses per day; proxy for network adoption
• Transaction count and volume — raw throughput; distinguish economic transfers from spam/self-transfers
• NVT ratio (Network Value to Transactions) — market cap / daily on-chain transfer volume; crypto's P/E ratio equivalent
• NVT Signal — uses 90-day MA of transfer volume for smoother signal; >150 historically overvalued for BTC

Mining/Staking:

• Hash rate — total computational power securing PoW networks; sustained drops signal miner capitulation
• Hash ribbons — 30-day MA crossing below 60-day MA of hash rate; historically strong buy signal when ribbons recover
• Staking ratio — percentage of supply staked; higher ratio reduces liquid supply but may indicate complacency

Exchange Flow Analysis

Inflow/Outflow Framework:

• Exchange inflow — tokens moving to exchange wallets; signals potential selling pressure
• Exchange outflow — tokens leaving exchanges; signals accumulation/cold storage
• Net flow — inflow minus outflow; sustained negative net flow is structurally bullish
• Exchange reserve — total tokens held on exchanges; declining reserves reduce available sell-side liquidity

Whale Tracking:

• Define whale thresholds by asset (BTC: >1000 BTC, ETH: >10,000 ETH)
• Track whale-to-exchange transfers as potential distribution signals
• Monitor whale accumulation addresses (exchange-to-whale flows)
• Nansen "Smart Money" labels track wallets with historically profitable behavior

Wallet Clustering and Entity Analysis

Clustering Methods:

• Common input heuristic — addresses used as inputs in the same transaction likely belong to the same entity
• Change address detection — identifying change outputs to group with sending addresses
• Timing analysis — addresses that consistently transact together
• Contract interaction patterns — wallets interacting with same DeFi protocols in similar patterns

Entity Categories:

• Exchanges (labeled by Chainalysis, Nansen, Arkham)
• Mining pools and miners
• DeFi protocols and their treasuries
• Institutional custodians
• Known fund wallets (a16z, Paradigm, etc.)

Methodology

Signal Construction Process

1. Data ingestion — pull raw data from node, indexer (The Graph), or analytics API
2. Metric computation — calculate derived metrics with appropriate lookback windows
3. Normalization — z-score or percentile rank metrics against historical distribution
4. Signal generation — define thresholds or regime boundaries for actionable signals
5. Confluence — combine multiple on-chain signals; require 3+ confirmations for high conviction

Practical Analysis Workflow

Step 1: Check exchange net flows (are tokens leaving exchanges?)
Step 2: Review MVRV and NUPL (are we in overvalued or undervalued territory?)
Step 3: Examine active address trends (is adoption growing or shrinking?)
Step 4: Monitor whale behavior (accumulating or distributing?)
Step 5: Cross-reference with funding rates and derivatives data
Step 6: Synthesize into a directional bias with confidence level

Data Source Hierarchy

• Tier 1 (most reliable): Running your own full node, parsing raw blocks
• Tier 2: Indexed data providers (The Graph, Bitquery, Flipside)
• Tier 3: Analytics platforms (Glassnode, Nansen, CryptoQuant, IntoTheBlock)
• Tier 4: Social/aggregated dashboards (Dune community queries, DefiLlama)

Examples

Example 1: BTC Accumulation Signal

Observation:
- Exchange reserves declining for 30+ consecutive days
- MVRV at 0.85 (below realized value)
- Whale addresses (>1000 BTC) increasing in count
- Hash ribbons recovering after miner capitulation

Assessment: Strong accumulation phase. Multiple on-chain metrics align
with historical bear market bottoms. High conviction long bias.

Example 2: ETH Distribution Warning

Observation:
- Large exchange inflows from top-100 wallets (3 consecutive days)
- NVT Signal at 180 (historically overvalued zone)
- NUPL at 0.72 (approaching euphoria)
- New address growth flattening despite price increase

Assessment: Distribution signals emerging. Price appreciation not
supported by network growth. Reduce position size, tighten stops.

Example 3: Stablecoin Supply Expansion

Observation:
- USDT and USDC combined market cap increasing $2B in 7 days
- Stablecoin exchange reserves rising (dry powder accumulating)
- Stablecoin dominance declining from recent highs

Assessment: Capital inflow into crypto ecosystem via stablecoins.
Historically precedes risk-on rallies. Watch for deployment into
BTC/ETH as confirmation.

Quality Gate

Before acting on on-chain signals, verify:

• Data source is reliable and up-to-date (check for indexing lag, typically 1-6 blocks)
• Metrics are calculated with appropriate lookback windows (avoid recency bias)
• Whale movements are confirmed across multiple tracking platforms (avoid false labels)
• Exchange flow data accounts for internal transfers and wallet rotations (exchanges regularly reshuffle hot/cold wallets)
• On-chain signals are corroborated by at least one other data domain (derivatives, sentiment, or price action)
• Contextual factors considered — protocol upgrades, airdrops, and bridge activity can distort on-chain metrics
• Time horizon matches the signal frequency — on-chain metrics are typically medium-term (weeks to months), not intraday
• Account for chain-specific nuances — UTXO vs account model, L1 vs L2 activity