Vol Street LogoVol Street Logo
Request a pilot

Deterministic options backtests on your vendor REST data

Connect Amberdata or IVolatility REST. Hydrate daily windows within vendor limits, replay bit for bit, and export a signed report. Runs in your environment.

Built for systematic funds and research shops that:

  • Access options data via REST APIs or provider databases
  • Need robust backtesting without the 3-6 month build
  • Want infrastructure that runs in their own environment
  • Need deterministic, auditable results for compliance

First valid backtest in 1 week • Replay drift under 1 bp • Runs inside your VPC

Schedule a quick call

15 minutes. I'll show you the product and answer your questions.

REST is one day at a time

Most vendors throttle NBBO options data to 1 day per request. If you run backtests straight from REST you end up making hundreds to thousands of calls, waiting on limits, and you still cannot replay results reliably.

Typical 5 year study requires roughly 1,250 to 1,800 daily windows per symbol

(depends on market vs calendar days per vendor).

Unsafe parallel pulls trigger throttles and retries

Safe parallelism still needs a queue that respects per-key rate limits.

Vendor corrections and late prints change history

Without versioned merges, your replays drift.

What Vol Street does

  • • Hydrates daily REST windows into a versioned store that matches your license and limits.
  • • Merges vendor corrections, emits a human-readable diff, and pins timezone and image.
  • • Runs deterministic backtests with a run hash, seed, and signed exports.

Hydration Queue

01-23
02-23
03-23
04-23
05-23
06-23
07-23
08-23
09-23
10-23
11-23
12-23
01-24
02-24
03-24
04-24
05-24
06-24
07-24
08-24
09-24
10-24
11-24
12-24
0
Windows fetched
4
Parallel jobs
0
Corrections applied

Two ways to run

Vol Street is not a data vendor

Your Workflow

1

Connect provider database or choose research store hydration

2

Define strategy in Python using templates

3

Run in your environment and export results

Quickstart on your data

volstreet-engine v2.1.3 • dataset: SPY 2018-2024 • adapter: ivolatility
Completed • 5.9s
Loading chart data...
Run Results (sample)
Sharpe
0.00
CAGR
0.0%
Max DD
0.0%
Trades
0
DateTypeStrike (LP/SP/SC/LC)DTEPnL

• Bring your license. Vol Street is not a data vendor.

• All computation runs in your environment.

• Offline activation available.

12-line quickstart, deterministic outputs, export to Parquet.

Features help you explore. Strategy needs a backtest.

Endpoint metrics are useful for ideas. They cannot decide strikes, manage positions, or respect capital. Options strategies are path dependent, so you need chain snapshots and lifecycle rules.

No chain state at entry

Strike choice depends on the whole chain and tradable strikes, not a single delta column.

No lifecycle

Rolls, early exits, assignment, and events change the path and the risk.

No execution or capital

NBBO fills, size, and margin rules shape real PnL.

What Vol Street adds: Selects strikes from the chain using your policy, simulates rolls and exits, enforces liquidity and margin, merges vendor corrections, and exports deterministic results.

Data and deployment

You bring your license and keep the data.

Providers

  • Amberdata
  • iVolatility
  • Additional providers on request

Where it runs

  • Your workstation
  • VPC (AWS, GCP, on-premise)
  • Optional offline activation
  • Hydrator outputs: S3 Parquet, Postgres, GCS

What teams get in week one

Connect a provider database or build a local research store with the hydrator, run a representative EOD strategy, export Parquet and a run manifest for review.

A mid-size fund with IVolatility connected Vol Street to its provider database and ran portfolio EOD backtests on SPY and QQQ.

  • Time to first run: 5 business days
  • Typical run: weekly OTM strategy, 2018–2024
  • Outputs: Parquet + run manifest for audit
  • Result use: analyst notebooks and PM review deck

Numbers shown are representative. Your timeline depends on provider, scope, and environment.

Deterministic by default

  • Strategy definitions and parameters are versioned
  • Inputs and adapter versions captured per run
  • Optional seeds for any randomness
  • Run manifest export for review and CI

Pricing for institutional buyers

Perpetual license with optional support. Pilot credited to license.

Pilot

2 to 3 weeks, fixed scope

  • • 1 vendor, 1 symbol set, 1 to 2 policies
  • • Exports: manifest, notebook, parquet, CSV, PDF
From $15k
one time

Note: 100% of pilot fee credited to production license within 60 days

Production license

Perpetual right to run the compiled SDK in your environment

  • • Includes implementation and handoff
  • • Node-limited license file, no telemetry
From $45k
one time

Scope note: single vendor and desk. Multi-vendor or multi-desk priced by scope

Support and updates

Vendor schema compatibility, bug fixes, minor features, security patches

  • • Replay drift assistance with SLA
12-18%
yearly of license price

Optional but recommended

Runs in your VPC • Replay drift under 1 bp • Signed artifacts and manifest • Respectful of vendor limits and corrections

Term subscription available on request. Buyout option converts current version to perpetual.

Common Questions

Do we need our own database?

No. Connect a provider database or build a local store with the built-in hydrator.

Where does it run?

On your workstation or in your VPC. Offline activation available.

How do you handle rate limits?

Respectful batching, backoff, and completeness checks when hydrating from REST.

What formats do you output?

Pandas DataFrames, Parquet, CSV. Trade logs, Greeks time-series—all exportable.

Ready to see it?

Book a 15-minute call. I'll walk you through a live backtest, show you how hydration handles rate limits, and answer questions about your setup.

Schedule a call

Built by Tyler Prahm

Currently working with pilot customers to validate market fit. First 5 pilots get discounted pricing and heavy input on the roadmap.