Coverage
Four official f(x) pools
ETH/BTC long and short pools are tracked. Wallets are filtered to valid 0x addresses; literal NULL strings and unknown system rows are excluded from trader rankings.
Methodology
How the f(x) research dashboard calculates trader and protocol metrics
This dashboard combines public f(x) Protocol subgraph data, indexed cashflows and current position snapshots to build trader profiles, capital-flow research, protocol-health cards and ranking tables. Values are research estimates for transparency and monitoring; they are not trading, investment, tax or legal advice.
PnL source
Official order stream first
Realized PnL prefers the official f(x) order stream when the latest synced order block is at least as fresh as the latest indexed cashflow block.
Freshness guard
No stale UI PnL
ui_realized_pnl_usd is used only when ui_last_order_block >= last_cashflow_block. Otherwise the dashboard falls back to cashflow-derived realized PnL.
Data pipeline
What data is read
| Layer | Table / source | Purpose |
|---|---|---|
| Current position book | fx_current_positions | Open-position owner, collateral, debt, equity, oracle price, debt ratio and current PnL inputs. |
| Historical cashflows | fx_position_cashflows | Manager-source deposits, withdrawals, debt changes, fees and first/last activity timestamps. |
| Official f(x) positions | fx_official_positions | Canonical owner / real owner, pool metadata, closed/open state and raw position state from the official f(x) subgraphs. |
| Official f(x) orders | fx_official_position_orders | Execution-price order history used for UI-aligned entry price, realized PnL and order-count freshness. |
| Position PnL cache | fx_position_pnl | Per-position cashflow aggregates plus official-order UI PnL columns and freshness guards. |
Coverage
Tracked pools
| Market | Side | Instrument | Interpretation |
|---|---|---|---|
| ETH Long | long | ETH | wstETH collateral, long-side debt |
| ETH Short | short | ETH | fxUSD collateral, borrowed ETH exposure |
| BTC Long | long | BTC | WBTC collateral, long-side debt |
| BTC Short | short | BTC | fxUSD collateral, borrowed BTC exposure |
PnL methodology
Realized, unrealized and total PnL
Realized PnL
Realized PnL is computed from official f(x) execution prices using a weighted-average entry model over Open/Add and Reduce/Close/Liquidate/WithdrawRepay/Withdraw orders. For shorts, the calculation uses debt-token size from deltaDebts and flips the direction so falling prices are profitable.
Unrealized PnL
Open-position unrealized PnL uses current oracle prices. Shorts prefer official-order UI unrealized PnL. Long displays intentionally use the raw entry-price fallback instead of long-side ui_entry_price_usd, because long order enrichment can have misleading decimal scales.
Total PnL
Trader total PnL is realized PnL from indexed position history plus current unrealized PnL from open positions. Top Traders and profile pages use the same ownership and PnL methodology.
Order sync
How official f(x) order history is kept complete
- Official f(x) subgraph orders are fetched with skip-based pagination in 1,000-order pages, so positions with more than 1,000 orders are not truncated.
- The incremental candidate set includes positions with recent cashflows, positions whose cashflow block is newer than the last synced official-order block, and current positions missing official-order coverage.
- Order rows are de-duplicated by official order id and sorted by block number and log index before entry/PnL calculations.
- A client timeout while calling the long-running sync endpoint is inconclusive; database freshness and enrichment output are the verification source.
Trader rankings
No opaque composite score as the default
| Tab | What it ranks |
|---|---|
| Top PnL | Default ranking. Sorts wallets by total PnL = realized PnL + unrealized PnL, descending. Positive PnL ranks above negative PnL. |
| Top ROI | Sorts by total PnL divided by capital used. Wallets must have at least $1K of capital/notional to reduce tiny-wallet noise. |
| Best Realized PnL | Sorts by closed-position realized PnL only. Unrealized gains are shown separately but do not drive this tab. |
| Most Active | Sorts by total position count. This measures activity/churn, not trading performance. |
| Whale Traders | Sorts by current notional exposure. This measures size, not trading performance. |
Capital flows
How deposits and withdrawals are converted to USD
| Flow type | Conversion |
|---|---|
| Long BTC collateral | collateral_raw / 1e8 × latest WBTC pool oracle |
| Long ETH collateral | collateral_raw / 1e18 × latest wstETH pool oracle |
| Short collateral | collateral_raw / 1e18 because short-side collateral is fxUSD-denominated |
| Net flow | deposits minus withdrawals in the selected 7d / 30d / all-time window |
Research sections
What the research hub measures
Research cards
Cards surface live signals and open a drill-down brief. Each brief links to a verifiable source view such as capital flows, a trader profile, a position page or the conviction tracker.
Trader archetypes
Archetypes are rule-based classifications over current exposure, position count, notional size, directional bias and liquidation proximity. They are labels, not performance scores.
Conviction tracker
Highest Conviction Positions are ranked by large open exposure weighted by debt-ratio pressure. This measures committed exposure and liquidation headroom, not skill.
Protocol health
Protocol Health cards explain TVL, open interest, long/short ratio, average leverage, net deposits, active traders, liquidation risk and fee generation.
Known limitations
How to read the numbers safely
- Values depend on the latest indexed subgraph rows, current position snapshots and oracle prices. They are refreshed periodically, not via a live browser-side subscription.
- Official-order PnL is treated as the preferred UI-aligned source only after passing the freshness guard against cashflows.
- Cashflow fallback PnL can differ from f(x) UI execution-price PnL, especially for shorts, because short collateral movements are fxUSD-denominated rather than direct execution-price profit/loss.
- Liquidation-risk displays are debt-ratio proximity signals, not predictions that a liquidation will occur.
- Metrics may differ from other dashboards because each site can use different coverage, refresh timing, owner attribution and calculation methods.