Algorithms of Jane Street: how the invisible infrastructure of modern markets works

Jane Street does not publish its algorithms and does not disclose sensitive operational details. The available information comes from public technical presentations, official statements, and regulatory documents.

Even so, the general picture is clear: the firm operates one of the most sophisticated electronic market-making infrastructures in the world.

• Technological foundation: OCaml

Jane Street consistently states in its technical presentations that it uses OCaml for almost its entire software ecosystem: trading, infrastructure, research, and hardware components.

The language is appreciated for its strict type system and its ability to reduce errors in an environment where decisions are made in microseconds and involve billions of dollars.

The company has developed its own libraries, including HardCaml for hardware design and incremental computing systems.

• Market-making algorithms

These algorithms represent the core of the business.

They simultaneously place buy and sell orders on hundreds or thousands of instruments, trying to capture the spread between bid and ask as often as possible.

The systems constantly calculate a fair price based on order flows, correlations between instruments, liquidity, volatility, and execution risk.

The spread changes dynamically depending on market conditions.

• Statistical and microstructural arbitrage

Jane Street exploits temporary discrepancies between ETFs and their component stocks, ADRs and underlying shares, different exchanges, derivatives, and underlying assets.

The algorithms exploit transient correlations and order flows to detect very short-term opportunities.

Individual profits are very small. The scale is enormous.

• Machine learning and predictive models

According to the company's public technical presentations, Jane Street has been using machine learning since the 2000s.

The models estimate operational probabilities: the probability of order execution, the impact of an order on price, the probability of the presence of informed flow, changes in liquidity in real time.

The data collected daily includes huge volumes of order book snapshots and trades, used for continuous recalibration of the models.

• Execution and risk management

Smart Order Routing systems choose the optimal market for execution in microseconds.

Algorithms attempt to minimize market impact and quickly neutralize unwanted directional exposure through hedging.

At the same time, risk systems continuously monitor exposure and can automatically trigger stop mechanisms.

The entire process runs in microseconds, using colocated servers near exchanges and specialized hardware components.

• The India case

The interim order published by the Indian authority SEBI in July 2025 describes a mechanism called the Two-Patch Playbook or Expiry Day Trap.

According to the authority, the algorithms aggressively bought components of the Bank Nifty index in the morning while simultaneously building short positions in options. Subsequently, the selling influenced the settlement price.

Jane Street rejects the accusations and maintains that the strategies used are legitimate forms of market-making and hedging.

The case illustrates the constant tension between aggressive exploitation of market microstructure and the limits imposed by regulators.

• The real advantage

Jane Street's advantage does not lie only in speed.

The firm combines very large proprietary capital, integrated technological infrastructure, mathematical research, instant operational feedback, and the ability to adapt quickly.

It is the point where advanced mathematics, high-performance programming, and market microstructure converge into a single operational system.