Aggregated RWA Price Feeds Methodology

Version 1.1 · 6 October 2026

Blocksize publishes aggregated RWA price feeds derived from crypto-native venues as part of its Bid/Ask data feed. The feeds aggregate quotations for tokenized representations of real-world assets, using the established Bid/Ask aggregation methodology and technical infrastructure.

The service runs 24/7, including outside traditional market hours and on weekends, as long as eligible crypto-native quotations are available. Prices may be provided with full redistribution rights, depending on the applicable licence.

At a glance

• 38 published feeds for tokenized equities, identified by the underlying ticker and the suffix _AGG.

• Sources are quotations from centralised exchanges, including tokenized-asset markets. DEXs and pools were researched and are not part of the published aggregate.

• During traditional market hours, every quote is checked against an internal TradFi benchmark before it can contribute.

• Bid and ask are aggregated separately, weighted by quoted size.

• In the tested market sessions, the aggregated feeds achieved a P99 absolute price deviation below 20 basis points (0.20%).

1. Published feeds

The current publication universe comprises 38 equity-related feeds. Each aggregate asset identifier uses the underlying ticker followed by _AGG. Quote currencies are specified separately through the Bid/Ask interface.

2. Source selection and price normalization

2.1 Eligible sources

The published RWA feeds use quotations from centralized exchanges (CEXs), including eligible tokenized-asset markets. Source suitability is assessed for each individual tokenized asset and exchange market. Different tokenized representations of the same underlying asset may contribute where their prices have been normalized to a comparable basis.

Initial research and testing also covered decentralized exchanges (DEXs), individual pools and pool-state prices. These observations inform source assessment. The published feeds aggregate CEX quotations only.

The exchanges associated with each instrument are disclosed through the bidask_instruments endpoint. The set of quotations accepted at any moment depends on live eligibility and data availability.

2.2 Price normalization

Quotations are normalized before comparison and aggregation. Live stablecoin-to-US-dollar conversion follows the Bid/Ask approach used across all pairs. For Ondo tokenized assets, token-to-underlying adjustments, including applicable share multipliers, align quotations with the price of the underlying asset.

2.3 Dynamic selection during market hours

Source selection is dynamic during traditional market hours, when a contemporaneous TradFi benchmark is available. Each incoming bid/ask quote is assessed by comparing its midpoint, (bid + ask) / 2, with the midpoint of the corresponding internal equity benchmark feed.

A quotation is eligible only if its absolute price deviation is within the maximum permitted deviation and it meets the Bid/Ask data-quality requirements. Eligibility is reassessed as quotations arrive, so the contributing sources can change from one update to the next.

The benchmark is used to assess eligibility. The published price is derived from accepted crypto-native CEX quotations only.

3. Aggregation

The RWA feeds use the Bid/Ask aggregation logic. Accepted bid and ask observations are aggregated separately, each weighted by its quoted size in units of the underlying asset. The Bid/Ask data-quality and contingency framework applies alongside the RWA-specific source selection.

The full calculation rules are described in the Bid/Ask methodology.

4. Testing and validation

Validation runs in two stages: first the individual data sources, then the resulting aggregated feeds. Source suitability and aggregate performance are assessed separately.

4.1 Testing individual data sources

Each tokenized-asset market is evaluated on its own against traditional-market reference data, on a common underlying-asset and currency basis. CEX observations use valid best-bid/best-ask midpoints. DEX research evaluated individual pools using trade prices and, where available, nonzero pool-state prices.

For CEX source testing, the latest valid midpoint is sampled at five-second boundaries, with a maximum quote age of ten seconds. DEX trades are evaluated within their own five-second buckets using USD-notional-weighted prices, without carrying trades into inactive buckets. Pool-state observations use freshness limits suited to the relevant network.

The assessment covers absolute price deviation, the distribution of deviations, freshness and coverage. Duplicate trades are removed, and observations are normalized for currency and token-to-underlying adjustments. Sources without enough comparable reference data stay inconclusive until enough evidence is available.

4.2 Testing aggregated price feeds

Each published aggregate feed is then tested against its internal equity benchmark feed. Aggregate and benchmark midpoints are recomputed from valid bid and ask values. Quotes must be finite and positive, with the bid no greater than the ask.

At each one-second sampling boundary, the latest aggregate and benchmark observations at or before that boundary are paired. Neither observation may be more than ten seconds old, and future observations are excluded. Repeated observations with the same ticker and timestamp are consolidated using the median midpoint.

Absolute price deviation, in basis points:

deviation (bps) = abs(aggregate midpoint / benchmark midpoint - 1) × 10,000

P99 is the nearest-rank 99th percentile of these absolute deviations. Price tails are kept without outlier removal or winsorization. Coverage and the number of valid paired observations are assessed alongside the deviation statistics, and a five-second sampling check supplements the one-second comparison. Insufficient reference data is treated as an evidence gap.

4.3 Results

In the tested market sessions, Blocksize achieved a P99 absolute price deviation below 20 basis points (0.20%) for the aggregated feeds. This is an observed result for the evaluated feeds, valid paired observations and tested conditions. It does not establish that every published feed meets this threshold in every market session.

The result is a historical test statistic and not a guarantee of future performance. It does not establish equivalent deviation outside the tested conditions, or during periods when no contemporaneous TradFi benchmark is available.

5. Market hours and data contingencies

Live comparison with the TradFi benchmark runs during traditional market hours, when that benchmark is available. Benchmark-based source selection is not performed when no contemporaneous benchmark is available.

Pricing continues outside traditional market hours, including weekends, using eligible crypto-native CEX quotations under the Bid/Ask aggregation and data-quality framework. These prices reflect the crypto-native market information available in those periods.

If no eligible crypto-native quotations are available, Blocksize does not publish a calculated price for the affected feed. The 24/7 service therefore depends on eligible sources being available.

6. Access

The RWA feeds are delivered through the existing Bid/Ask endpoints and response formats. Existing Bid/Ask clients can retrieve or subscribe to the aggregate instruments through the same infrastructure, with no additional implementation.

The bidask_instruments endpoint lists the available instruments and their exchanges. Use the aggregate identifiers from section 1, such as AAPL_AGG or NVDA_AGG, together with the quote currency and the instrument details returned by that endpoint. See the API documentation.

For access to the RWA feeds, contact Blocksize or write to connect@blocksize-capital.com.

7. Methodology maintenance

Blocksize reviews source suitability and feed quality under its Bid/Ask framework. Source coverage and the publication universe may change as eligible markets and further underlying assets become available. Current instrument and exchange information is available through bidask_instruments.

Potential additional coverage

The following 69 underlying assets are candidates for future feeds. Inclusion in this list does not mean a feed is published, and it is not a launch commitment. Aggregate identifiers are assigned when a feed is introduced. Gold (XAU) and silver (XAG) are published as RWA spot feeds. The remaining candidates are equity-related.

1810.HK Xiaomi · ABBV AbbVie · ABNB Airbnb · ABT Abbott · ACN Accenture · ADI Analog Devices · ALAB Astera Labs · AMAT Applied Materials · AMGN Amgen · ANET Arista Networks · APP AppLovin · ARM Arm Holdings · ASML ASML Holding · AXP American Express · BA Boeing · BE Bloom Energy · BIDU Baidu · BLK BlackRock · BRK.B Berkshire Hathaway · CAT Caterpillar · CEG Constellation Energy · COHR Coherent · COST Costco · CRWD CrowdStrike · DELL Dell · DIS Disney · EQIX Equinix · ETN Eaton · FCX Freeport-McMoRan · GE General Electric · GEV GE Vernova · GLW Corning · GS Goldman Sachs · HON Honeywell · HPE Hewlett Packard Enterprise · IBM IBM · INTU Intuit · ISRG Intuitive Surgical · JNJ Johnson & Johnson · KEYS Keysight Technologies · LITE Lumentum · LLY Eli Lilly · LMT Lockheed Martin · MA Mastercard · MCD McDonald's · MRNA Moderna · MSTR Strategy · NBIS Nebius Group · NET Cloudflare · NFLX Netflix · NOK Nokia · NOW ServiceNow · NU Nu · OXY Occidental Petroleum · PANW Palo Alto Networks · PBR Petrobras · PFE Pfizer · PG Procter & Gamble · REGN Regeneron Pharmaceuticals · SBUX Starbucks · SHOP Shopify · SNOW Snowflake · SPOT Spotify · STX Seagate · TCEHY Tencent Holdings · UNH UnitedHealth · UPS UPS · VRT Vertiv · WMT Walmart