Whale Buy Ratio Explained: How to Read On-Chain Flow Data
What buy ratio, net flow, conviction scoring, and the Whale Sentiment Index actually mean — a plain-language guide to interpreting the metrics on Deep Blue Alpha's dashboard.
Published 2026-06-20 · Deep Blue Alpha
Whale buy ratio measures the percentage of total whale trading volume that consists of purchases: buy_volume / (buy_volume + sell_volume). Above 50% means whales are net buyers. Below 50% means net sellers. Net whale flow is the dollar difference: buy_volume - sell_volume. Positive = net inflows, negative = net outflows.
Deep Blue Alpha tracks these metrics across 20,000+ Ethereum whale wallets in real time. This guide explains every core metric on the dashboard — buy ratio, net flow, the Whale Sentiment Index, conviction scoring, and multi-wallet convergence — so you can interpret what the data shows without needing to reverse-engineer the methodology yourself.
All metrics are available on the free dashboard at deepbluealpha.io. Updated June 2026.
What is whale buy ratio and why does it matter?
Whale buy ratio is one of the simplest and most informative metrics in on-chain analytics. It answers a single question: what fraction of total whale trading volume came from the buy side?
The formula is straightforward:
Buy Ratio = buy_volume / (buy_volume + sell_volume)
If whale wallets bought $8M worth of a token and sold $2M in the same 30-day window, the buy ratio is $8M / ($8M + $2M) = 80%. That is a heavily buy-dominated period. If they bought $4.5M and sold $5.5M, the ratio is $4.5M / ($4.5M + $5.5M) = 45% — a sell-leaning window.
The metric matters because whale wallets are not average market participants. These addresses hold significant positions in Ethereum-based assets and tend to move larger individual trade sizes. Their aggregate behavior can represent a meaningful share of a token's total DEX volume on any given day. When whale buy ratio climbs above 60% on a token, it reflects concentrated, sustained purchasing. When it drops below 40%, it reflects sustained distribution.
Whale buy ratio also differs from aggregate exchange data (which mixes retail, whale, and algorithmic volume indiscriminately). Deep Blue Alpha filters for whale-grade wallets specifically, excluding known bot addresses and market makers through automated detection. The result is a cleaner read on what the largest individual holders are actually doing.
Buy Ratio Interpretation Guide
| Buy Ratio Range | Label | What It Means | Historical Context |
|---|---|---|---|
| 75%+ | Heavy Accumulation | Whales strongly favoring buy side; 3:1 or higher buy-to-sell ratio | Relatively uncommon; typically seen during sharp pullbacks or after extended sell-offs |
| 60–75% | Net Buying | Buy volume meaningfully exceeds sell volume; sustained accumulation | Common during periods of growing on-chain activity on a token |
| 40–60% | Balanced / Transitional | Neither side dominates; could be consolidation or a shift between regimes | Most tokens sit here most of the time; the equilibrium state |
| 25–40% | Net Selling | Sell volume meaningfully exceeds buy volume; distribution underway | Often follows extended rallies or precedes periods of reduced whale interest |
| Below 25% | Heavy Distribution | Whales strongly favoring sell side; 3:1 or higher sell-to-buy ratio | Uncommon; typically seen during panic events or protocol-specific crises |
One important nuance: buy ratio is a rear-view mirror. It tells you what whales did, not what happens next. A 75% buy ratio over 30 days is a factual observation about completed trades. It does not imply that buying continues, that price appreciates, or that the market has reached a bottom. Treat it as data, not as a directional forecast.
How is net whale flow calculated?
Net whale flow is the dollar difference between total whale buy volume and total whale sell volume over a given time window:
Net Flow = buy_volume − sell_volume
Where buy ratio gives you a percentage, net flow gives you the absolute dollar magnitude of the imbalance. This distinction matters because a token with an 80% buy ratio on $100K of total volume is a very different situation from an 80% buy ratio on $50M of total volume. The ratio is the same; the net flow ($80K vs $40M) tells a completely different story about the scale of whale activity.
Example
Suppose DBA tracked the following LINK whale activity over the past 30 days:
- Buy volume: $5.0M across 47 trades from 19 distinct wallets
- Sell volume: $3.0M across 31 trades from 14 distinct wallets
The net flow is $5.0M − $3.0M = +$2.0M. Positive. Whales were net buyers of LINK to the tune of $2M over the period. The buy ratio is $5.0M / $8.0M = 62.5%, which falls in the "Net Buying" band.
If instead sell volume had been $7.0M, the net flow would be $5.0M − $7.0M = −$2.0M. Negative. Whales were net sellers despite still buying $5M worth. The buy ratio would be $5.0M / $12.0M = 41.7% — transitioning into the "Net Selling" band.
Net flow is displayed on every token page on Deep Blue Alpha, alongside inflows, outflows, buy ratio, whale count, and trade count. It is also the primary metric used in the hourly whale report graphics and the trending tokens dashboard.
What does the Whale Sentiment Index (WSI) mean?
The Whale Sentiment Index is a composite 0–100 score that combines two complementary views of whale behavior across the entire tracked universe:
- Trade Sentiment — what percentage of individual whale trades (by count) are buys vs sells. If 60 out of 100 whale trades in the past hour were buys, the trade sentiment component is 60.
- Volume Sentiment — what percentage of total whale dollar volume is buy-side vs sell-side. If whales traded $20M total and $13M was buy-side, the volume sentiment component is 65.
The WSI blends these two components. The result is a single number that reflects the overall buying or selling posture of the tracked whale universe at any given moment.
WSI above 50 = net buying activity across tracked wallets. WSI below 50 = net selling activity. WSI at 50 = balanced.
Why two components? Because trade count and dollar volume can diverge meaningfully. A scenario where 80% of trades are buys but 70% of dollar volume is sell-side tells you that many whales are making small purchases while a few are executing large sells. Trade sentiment would read bullish; volume sentiment would read bearish. The composite WSI captures both dimensions, preventing either signal from masking the other.
The WSI is displayed on the main dashboard and updates in real time as new whale transactions are detected on-chain. It applies to the entire tracked universe — all tokens, all wallets — giving a macro-level read on whale positioning. For individual token sentiment, use the per-token buy ratio on the token page instead.
How does conviction scoring work?
Not all whale wallets carry equal informational weight. A wallet that has historically timed entries and exits effectively provides a different kind of signal than one that trades frequently with mixed results. Deep Blue Alpha's conviction scoring system addresses this by evaluating individual wallets based on their historical trading patterns.
Conviction scoring examines factors including:
- Historical consistency — has the wallet's past buying activity on a token preceded periods of appreciation, and has its selling activity preceded periods of decline? Wallets with consistent directional accuracy receive higher conviction scores.
- Position sizing behavior — does the wallet take concentrated positions (larger individual trades) or spread activity across many small trades? Concentrated positioning, when historically accurate, carries higher conviction.
- Holding duration — wallets that hold positions for meaningful periods (as opposed to rapid in-and-out trading that resembles bot behavior) are scored differently from short-duration traders.
The practical application: when a high-conviction wallet makes a large trade, it registers as a stronger signal in the DBA system than the same-sized trade from an unscored or low-conviction wallet. This filtering layer is part of the Intelligence Suite available to Pro tier subscribers and above.
Conviction scoring is backward-looking. A high score reflects past patterns; it does not guarantee future accuracy. Whale wallets can change strategies, rotate operators, or encounter different market conditions that render historical patterns less informative. The score is a data input, not a recommendation.
What is multi-wallet convergence?
Multi-wallet convergence is one of the strongest signal types in whale flow analysis. It occurs when three or more independent whale wallets execute trades on the same token within a short time window — typically hours, sometimes a single day.
Why does independence matter? A single whale buying $10M of a token is notable, but it could reflect a portfolio rebalance, an OTC settlement, or a strategy unique to that wallet. When three, five, or eight unrelated wallets all buy the same token within a few hours, the signal is structurally different: multiple independent actors arrived at the same conclusion through their own analysis.
Multi-wallet convergence: 3+ independent whale wallets trading the same token within a short window. Stronger signal than a single wallet acting alone. Detected automatically by DBA's Intelligence Suite.
The Intelligence Suite flags convergence events and surfaces them to subscribers. The system checks for independence by ensuring the wallets have no shared transaction history (i.e., they have not sent tokens to each other, which would suggest they are controlled by the same entity or operating in coordination).
Convergence works in both directions. When multiple wallets sell the same token simultaneously, it is a convergence event on the sell side. The metric is direction-neutral by design — it highlights concentration of whale activity, not a specific directional call.
How to read the DBA live feed
The live whale feed is the most granular view Deep Blue Alpha offers. It shows individual whale transactions as they happen on-chain, updating in real time via WebSocket. Each row in the feed contains:
- Timestamp — when the transaction was confirmed on Ethereum. Displayed in your local timezone.
- Wallet Address — the whale wallet that executed the trade. Shortened (e.g.,
0x7a25...f93d), clickable to the wallet's full page on DBA, which shows that wallet's complete trading history and conviction score. - Token — the token that was traded. Clickable to the token's DBA page, which shows aggregate flow data for that asset.
- Action — either
SWAP_BUY(the wallet purchased the token, swapping ETH/WETH/stablecoins for it) orSWAP_SELL(the wallet sold the token, swapping it for ETH/WETH/stablecoins). - USD Value — the dollar value of the trade at the time of execution, derived from on-chain pricing data.
- Sentiment Tag —
BULLISHfor buy-side trades,BEARISHfor sell-side trades. This is a factual label based on trade direction, not a prediction.
The feed also supports filtering by token, by wallet, and by minimum trade size. Free tier users see the full live feed with the most recent trades. Pro and Alpha tier users get extended history and additional filtering options.
DBA Core Metrics Reference
| Metric | Formula | Where to Find | What to Watch For |
|---|---|---|---|
| Buy Ratio | buy_vol / (buy_vol + sell_vol) |
/token/TICKER page | Sustained readings above 60% or below 40% |
| Net Flow | buy_vol − sell_vol |
/token/TICKER page, hourly reports | Magnitude relative to token's market cap |
| Whale Sentiment Index | Composite of trade + volume sentiment | Main dashboard (/) | Divergence between trade count and volume |
| Conviction Score | Historical accuracy weighting | /wallet/ADDRESS page | High-conviction wallets trading against the trend |
| Multi-Wallet Convergence | 3+ wallets same token in short window | Intelligence Suite alerts | Convergence on tokens outside the top 20 |
| Trade Count | Count of individual whale swaps | /token/TICKER page, /feed | High count with small avg size may indicate bots |
| Whale Count | Distinct wallets trading the token | /token/TICKER page | Growing whale count = broadening interest |
How to read the token page flow data
Every token tracked by Deep Blue Alpha has a dedicated page at /token/TICKER (e.g., deepbluealpha.io/token/LINK). These pages consolidate all whale flow data for that asset into a single view. Here is what each section contains:
The narrative summary
At the top of each token page, DBA generates a plain-English narrative sentence that summarizes the 30-day whale flow. A typical reading looks like:
"Over the past 30 days, DBA tracked 19 whale wallets trading LINK. Inflows totaled $5.0M against $3.0M in outflows, producing a net flow of +$2.0M — a 62% buy ratio."
This sentence is auto-generated from the live data. It provides the essential story at a glance: how many wallets were active, the total dollar volume on each side, the net result, and the buy ratio.
The flow breakdown
Below the narrative, the page displays:
- 30-Day Inflows — total USD value of whale buy-side trades. This is the raw buy volume before netting.
- 30-Day Outflows — total USD value of whale sell-side trades. The raw sell volume.
- Net Flow — inflows minus outflows. Positive means net buying; negative means net distribution.
- Buy Ratio — inflows divided by total volume (inflows + outflows), expressed as a percentage.
- Whale Count — the number of distinct whale wallets that traded this token during the period.
- Trade Count — the total number of individual swap transactions recorded.
Each of these figures traces directly to the on-chain transaction data in DBA's database. The numbers update as new whale trades are detected in real time.
Interpreting the numbers together
No single metric tells the full story. A few patterns worth noting:
- High net flow + high whale count = broad-based whale interest. Multiple wallets contributing to the flow independently.
- High net flow + low whale count = concentrated activity. One or two wallets driving most of the volume. Not inherently better or worse, but structurally different.
- Low net flow + high trade count = lots of activity with minimal net directional bias. Whales trading back and forth or different whales on different sides.
- Negative net flow + high buy ratio is impossible by definition, since a buy ratio above 50% means buy volume exceeded sell volume. If you ever see conflicting numbers, check the time window — the feed and the token page might be showing different periods.
Common misinterpretations of whale flow data
Whale flow metrics are descriptive, not prescriptive. They report what happened on-chain. Misinterpretation usually comes from adding a predictive layer that the data does not support. Here are the most common traps:
1. Treating buy ratio as a price predictor
A 70% buy ratio does not mean price appreciation follows. It means whales bought more than they sold over the measured window. Price is determined by the full market — retail traders, market makers, algorithmic bots, CEX order books, derivatives, macro flows, and more. Whale DEX activity is one input among many. Historical periods of high buy ratio have sometimes preceded price appreciation and sometimes preceded further declines. The relationship is not reliable enough to use as a standalone signal.
2. Ignoring wallet concentration
A +$10M net flow that came from a single wallet is qualitatively different from +$10M spread across 15 independent wallets. The aggregate number is the same, but the signal structure is different. One wallet could be rebalancing for portfolio reasons unrelated to a directional view on the token. Fifteen wallets converging independently is harder to explain as noise. Always check the whale count alongside the net flow figure.
3. Confusing time horizons
A token with a 70% buy ratio over 30 days and a 30% buy ratio over the last 24 hours is not contradicting itself. It is showing a shift: whales were net buyers for most of the month but have recently shifted to net selling. The 30-day figure includes the 24-hour figure. Both are correct; they describe different windows. When reading DBA data, always note which time frame you are looking at and whether the short-term trend diverges from the longer-term picture.
4. Not accounting for bot and market-maker filtering
Whale flow data is only as clean as the wallet universe it tracks. Deep Blue Alpha applies automated bot detection to filter out market makers, arbitrage bots, and MEV searchers from the tracked wallet set. This includes: a net-imbalance filter (wallets with less than 20% net direction on high volume are flagged as market makers), a frequency filter (wallets executing more than 100 trades in a window are flagged as bots), and an average trade-size filter (wallets with small average trades across many executions are flagged as small-trade bots). If you are comparing DBA data to another platform's whale metrics, differences in bot filtering methodology can produce meaningfully different readings on the same token.
5. Correlation is not causation
Even when whale buying and price appreciation occur in the same window, one did not necessarily cause the other. Both could be driven by a third factor (a protocol upgrade announcement, a partnership, a macro event). The data shows co-occurrence, not causation. Responsible analysis acknowledges this limitation explicitly.
Signal Strength Hierarchy
| Signal Type | Description | Reliability | Example |
|---|---|---|---|
| Multi-wallet convergence (high conviction) | 5+ high-conviction wallets buying same token within hours | Strongest | 8 independent wallets accumulated AAVE in a 6-hour window |
| Multi-wallet convergence (mixed conviction) | 3+ wallets of varying conviction scores on same token | Strong | 4 wallets bought LINK over 2 days; 2 high-conviction, 2 unscored |
| Single high-conviction wallet, large position | One historically accurate wallet taking a concentrated position | Moderate | Top-ranked wallet bought $3M of UNI in a single trade |
| High buy ratio, broad wallet base | 70%+ buy ratio across 10+ distinct wallets over 30 days | Moderate | PENDLE showed 72% buy ratio from 14 wallets over 30 days |
| Single wallet, large position | One unscored wallet making a large trade | Weaker | Unknown wallet bought $5M of ENS; no conviction history |
| High buy ratio, low wallet count | Strong ratio driven by 1–2 wallets | Weakest | 90% buy ratio on token X, but only 2 wallets contributed |
None of these signal types constitute a recommendation to trade. The hierarchy describes relative informational weight — how many independent data points support the observation. Even the strongest signal type (multi-wallet convergence from high-conviction wallets) has produced losses in past instances. Past patterns are not predictive of future price movements.
Bottom line
Whale flow data provides a window into what the largest on-chain participants are doing — how much they are buying, how much they are selling, and whether independent wallets are converging on the same assets. Deep Blue Alpha distills this into a set of core metrics: buy ratio, net flow, the Whale Sentiment Index, conviction scoring, and multi-wallet convergence.
The metrics are descriptive. They report completed on-chain transactions, not predictions. A 70% buy ratio tells you that whales bought more than they sold over the measured period. It does not tell you what happens next. Responsible interpretation means treating these numbers as one layer of research — combining them with fundamental analysis, market context, risk management, and your own independent judgment.
Everything described in this guide is available on the free tier at deepbluealpha.io. The live feed, token pages, and dashboard sentiment require no signup and no payment. Advanced features — conviction scoring, multi-wallet convergence alerts, and extended historical data — are available through the Intelligence Suite on the Pro tier and above.
Watch whale activity in real time
The DBA live feed shows every whale trade as it happens on-chain. Free, no signup required.
Open the live feed →