Metrics & Methodology

perpdex.io records listing discovery prospectively from the production monitoring baseline. Current listed markets count only markets whose latest status is OBSERVED. Historical discovered is the non-decreasing unique-market total: removed instruments leave the current count, while discovery, removal and reappearance events remain in the ledger. BASELINE_OBSERVED means a market already existed when monitoring began; a first-seen event means it first appeared in a later complete market set. Neither is an official exchange listing timestamp. FIRST_VALID_TRADABLE requires a mark and a public two-sided quote no older than 600 seconds. NO_LONGER_OBSERVED requires absence from three consecutive complete discovery snapshots, preventing a single collection failure from being reported as a delisting.

perpdex.io calls decimals() and balanceOf(0x74bb…f2cd) on the native USDC contract at the same Arbitrum block, then applies the verified six-decimal scale. This is only the on-chain USDC held by that wallet and is labelled ‘Arbitrum OLP Vault Balance (excluding capital currently hedging)’. It excludes capital already deployed to hedges, held elsewhere or off-chain, and is not total OLP capacity. Results are cached for 60 seconds; an RPC failure may retain the last success for up to 900 seconds with an explicit STALE state.

After each UTC day boundary perpdex.io reads the Arbitrum native-USDC balance of 0x5e91…d645 and writes one PostgreSQL daily snapshot. “Positive treasury balance change (experimental)” equals max(current-day balance − prior-day balance, 0). Deposits, operating transfers and other treasury movements can also increase the balance, so it is not audited accounting revenue. The value-capture view shows 90 days and keeps a rolling 92-day daily-snapshot window.

Variational's latest public statement says that after TGE it intends to use 100% of revenue directed to the treasury to buy and burn VAR, with detailed execution rules to come before TGE. perpdex.io uses the same UTC daily treasury-balance snapshots shown in the platform overview. The 7D and 30D annualized references require a complete consecutive window and equal the window total divided by window days, multiplied by 365. Balance change per USD 1B equals the same-day positive treasury balance change divided by the sum of Perp and Swap trailing-24h volume, multiplied by USD 1 billion. This ratio is descriptive and does not prove that either business line caused the observed wallet change; the page shows only the latest 90 days.

signed_cashflow_usd > 0 means the simulated account receives cash; < 0 means it pays. For cross-venue comparison, cost_usd uses the opposite sign: positive is a cost and negative is income or a cost reduction. Cards, cumulative charts, API responses and CSV exports all use this convention.

Variational publishes separate Long and Short financing values for Swaps. The two sides are generally asymmetric and either may pay or receive. The Perp rule that a positive funding rate means Long pays applies only to Perps and must not be used to interpret directional Swap fields.

Index and metal Swaps currently use basis=rates. Signed Long/Short annual rates directly represent cashflow for that position holder, with no second inversion for Short. Daily financing uses ACT/360.

Net financing cashflow and implicit quote cost are different measurements. perpdex.io derives a two-sided reference carry and half-spread from the published Long/Short values. The result is descriptive mathematics labelled DERIVED_NOT_OFFICIAL_FEE; it is not presented as official SOFR, a platform fee or broker revenue. A Short may receive net financing while still receiving less than the derived midpoint.

Oil Swaps currently use basis=points. Positive points mean that holder receives; negative points mean it pays. The amount depends only on fixed position quantity and signed directional points. Close price and notional are audit context and do not enter this formula.

Swap financing is applied at the market close. The study books only uniquely identified latest_applied events after entry. Pre-entry events are not backfilled, missing intervals are never written as zero, and weekend or holiday multipliers are taken only from the source value.

Financing, slippage and fees live in an append-only event ledger. A formula or semantic correction creates a new version that points to the superseded event; it never edits the old row. Public totals, charts and equity use only is_effective=true events, while CSV retains the full audit history.

Without independent Long/Short direction evidence, perpdex.io shows raw values but does not generate directional costs. A single rate defaults to SINGLE_RATE_AMBIGUOUS. Any unverified schema, direction, settlement time or source identity stops booking.

TradeXYZ consumes a same-time L2 book to compute fixed-notional VWAP, partial last-level fills, slippage and zero-discount fees. Variational has no public order book, so perpdex.io uses the OLP's public two-sided indicative RFQ prices at the corresponding size. The $100K monitor treats adverse deviation from Mark as estimated cost and favorable deviation as price improvement; its immediate round-trip spread buys at ASK and sells the same quantity at BID. Indicative quotes are not firm fills, and stale quotes or a non-open market suppress the estimate.

The main long-hold chart accumulates only verified financing/funding settlements. Entry slippage and trading fees are excluded and shown separately on position cards. Positive USD values mean paid cost; negative values mean financing income received.

CSV is generated deterministically from the authoritative PostgreSQL event ledger. supersedes_event_id points to the replaced event, superseded_by_event_id identifies its replacement, and is_effective controls inclusion in current totals. Historical rows are never removed from the download.

Advantage = TradeXYZ cost − Variational cost; a positive value means Variational is cheaper. Statistics use overlapping complete dates only. Fewer than 30 complete days blocks preliminary linearity conclusions; fewer than 90 days or coverage below 95% blocks venue stability comparisons.

perpdex.io is for independent research and information only, not investment, trading, legal, tax or accounting advice. An official-source label identifies provenance and does not imply endorsement of perpdex.io calculations. Data may be delayed, missing or corrected, and material decisions require independent verification. Data, address-label or methodology issues can be reported with public evidence through Telegram @demiyv1 or X @0xkuncai.