The first bare-metal data warehouse dedicated to T+1 market history.
1 Petabyte of Raw OPRA, SIP, and Fundamental Data.
Direct SQL access. Zero abstraction layers. Flat-rate OpEx.
Most cloud infrastructure fails because it treats financial time-series like standard web logs. We built our stack around the specific physics of market microstructure to eliminate the "Cloud Tax."
Moving 1PB of data to your compute (S3 to EC2) is impossible for a boutique firm. Moving your compute to the data is instant.
The Solution: We give you bare-metal SQL access directly on the storage node. Zero egress fees. Zero network latency.
Standard SQL joins use "Equality" (Date = Date). Markets move in nanoseconds. This creates slippage in backtests.
The Solution: We use vectorized ASOF JOINS to find the prevailing price at the exact millisecond your signal fired.
Vendor-provided "Greeks" are static snapshots based on their black-box assumptions.
The Solution: Calculate your own Implied Volatility surfaces on the fly using our native QuantLib integration.
Stop fighting with Spark clusters. We normalize disparate feeds into a high-compression ClickHouse architecture optimized for massive aggregations and point-in-time lookups.
Query: Join 1-Day of Options to Underlying, calc IV.
| Platform | Speed | Cost |
|---|---|---|
| Standard Cloud SQL | Timeout | $15+ / scan |
| AWS Athena (S3) | ~45 Seconds | ~$5.00 / scan |
| Strasmore (Sovereign) | 0.05 Seconds | Included |
-- The "Sovereign" Backtest: Detecting Historical Arbitrage -- Replaying Market State from March 7, 2022 WITH underlying_feed AS ( SELECT ticker, (bid_price + ask_price) / 2 AS spot, sip_timestamp FROM stocks_quotes WHERE sip_timestamp BETWEEN '2022-03-07 14:30:00' AND '2022-03-07 14:35:00' ) SELECT extract(o.ticker, '^O:([A-Z]+)') AS root, toDecimal64(extract(o.ticker, '([0-9]{8})$'), 3) / 1000 AS strike, -- The Exact Stock Price at that Millisecond u.spot AS underlying_price_at_quote, (o.bid_price + o.ask_price) / 2 AS opt_mid_price, -- Calculate Intrinsic Value (Spot - Strike) (u.spot - strike) AS intrinsic_diff FROM options_quotes AS o ASOF LEFT JOIN underlying_feed AS u ON extract(o.ticker, '^O:([A-Z]+)') = u.ticker AND o.sip_timestamp >= u.sip_timestamp WHERE o.sip_timestamp BETWEEN '2022-03-07 14:30:00' AND '2022-03-07 14:35:00' AND root = 'TMUS' AND extract(o.ticker, '[0-9]{6}([CP])') = 'C' ORDER BY o.sip_timestamp DESC LIMIT 5 >> Result: TMUS | Spot: $123.00 | Strike: $75.00 | Price: $47.70 >> Status: Cache MISS (Cold Run) | Time: 0.473s | Rows: 13.8M >> Arbitrage: Option trading $0.30 BELOW intrinsic value.
Normalized from >1PB of raw logs. No packet loss. 100% Data Completeness.
| Asset Class | Coverage | Depth | Tier Availability |
|---|---|---|---|
| Aggregates (OHLCV) | Options, Stocks, Crypto | 1-Minute Bars | ALL TIERS |
| US Equities (SIP) | CTA / UTP | Tick-Level Trades & Quotes | Portfolio+ |
| US Options (OPRA) | All Chains | Full L1 Quotes (44B/day) | SOVEREIGN ONLY |
We segment tiers by data granularity (Resolution), not just capacity.
Protect your capital by paying only for the physics you need.
For rapid backtesting and signal discovery.
For prop desks and execution modeling.
Total infrastructure ownership.
Storing 1 Petabyte of OPRA data typically requires a $200k+ hardware investment or ~$23,000/mo in AWS S3 Standard fees. Strasmore converts this heavy CapEx into a predictable, flat-rate OpEx subscription.
Strasmore provides Historical Data only. This generally exempts clients from "User Reporting" requirements and real-time display fees associated with terminals like Bloomberg or Refinitiv.
Client query logs are strictly confidential. We are legally bound against trading on client signals or aggregating user data for third-party sentiment products.