New$10 Sign Up Bonus & $30 Free Trial Reward
Start challenge

Algorithmic trading

Also called algo trading, automated trading

Trading in which a computer program decides order details such as timing, price or size with little or no human input. It ranges from execution algorithms that slice large orders to fully automated strategies.

Reading
2 min
Sources
3RegulationRegulatorInternational body
Updated
On this page· 6 sections
  1. How it works
  2. Example
  3. Watch out
  4. At TRONEXO
  5. Sources
  6. Related terms

What should be bought, at what price, in what size, on which venue and at what moment? These questions have long been central to trading[2]. In modern equity and bond markets, software answers many of them, reading market data and deciding when, where and how to trade[2].

That is algorithmic trading. How much the program decides varies: some algorithms only work out how to fill an order a person has already chosen, while others decide whether to trade at all[3].

How it works

In EU law, trading is algorithmic when a computer algorithm automatically sets individual order parameters, such as whether to send the order, its timing, price or quantity, or how to manage it after submission, with limited or no human intervention[1]. A system that only routes orders to a venue, or only confirms and processes trades, does not count[1].

Two broad families sit under that label:

  • Execution algorithms carry out a decision already made. They split a large “parent order” into smaller “child orders” and execute them over time; the simplest, time-sliced kind spreads them evenly[3].
  • Strategy algorithms decide whether to trade. The BIS names market-making algorithms and opportunistic ones, which principal trading firms and hedge funds commonly use to generate profit[3].

The BIS estimated in October 2020 that FX execution algorithms handled 10–20% of global spot FX trading, about $200–400 billion a day[3].

In the EU, a firm that trades algorithmically must have risk controls, including trading thresholds and limits, that prevent erroneous orders and keep its systems from contributing to a disorderly market[1]. It must also notify the regulators of its home country and of the venues where it trades this way[1].

Example

Say an asset manager decides to buy 48,000 shares between 12:00 and 16:00 and hands the order to a time-sliced algorithm.

  1. One child order every 5 minutes for 4 hours gives 4 × 12 = 48 child orders.
  2. Each child order is 48,000 ÷ 48 = 1,000 shares.
  3. The price drifts up. The first 24 fills average $50.00 and the last 24 average $50.20, so the order averages $50.10 and costs 48,000 × $50.10 = $2,404,800.

The manager chose the stock, the size and the window. The algorithm set the size and timing of each child order.

Watch out

Errors travel fast. In 2012 a US broker-dealer’s order routing system, after a systems update, sent millions of erroneous orders into the market over 45 minutes and cost the firm more than $460 million[2].

Algorithmic trading is wider than high-frequency trading. The sole purpose of an execution algorithm, the BIS writes, is to execute a trade as optimally as possible[3].

At TRONEXO

On the simulated Challenge accounts and TRONEXO Accounts, in every phase, each order must be placed manually and personally by the trader. Trading robots, algorithms, scripts, Expert Advisors, API connections, auto-clickers and macros that place or manage trades without manual action are prohibited. See Funding models & rules and the FAQ.

All terms in Tronexopedia

Your challenge starts here

Start your challengeFree TrialNew
Registration bonus

Get $10 free in your Wallet when you sign up

Register and use your $10 Wallet credit to reduce the price of your first challenge.

The $10 bonus is credited to your personal Wallet and can only be used to reduce the price of a challenge purchase. Credits cannot be withdrawn or used for payouts.

Start now & get $10