Algo Trading Basics

What is Algo Trading?

In capital markets,
Algo trading means using a software enabled with taking decisions to trade under pre-defined constraints.
Software may provide in-built algos or also called strategies which take market feed and indicators as inputs. Algo works under certain constraints to achieve or maximize an objective. Some software also gives its user the flexibility to write own algos.
For example, we take example of a simple two-leg algorithm.
Constraints:-
Buy Instrument A in a quantity
Hedge it against Instrument B by selling Instrument B in b quantity
such that price difference is p
Software will run an algo which shall process the incoming feed of Instruments A and B.
Whenever price difference of Best Ask Price of A and Best Buy Price of B is p
Algo will send a corresponding buy order for A and sell order for B. Thus if trade happens, price difference of p can be achieved.
Trader can run the reverse algo of selling A and buying B at price difference P.
The difference of p and P is the profit earned by trader.
Exchange and setup charges when excluded will give the net profit of the trader.

The example given above is the basis of Cash-Future, Future-Future algorithms.

Experts in trading should only use algo trading systems. Though there are some algos (eg. IOC algos) which are risk-free when run under some experts supervision.

What are different algo types?

• Arbitrage Algos
  Cash vs Future
  Future vs Future or Calendar Spread
  Conversion Reversal
  Butterfly strategy
  Box strategy
  2Leg, 3Leg, 4 Leg Ratio Spread
• Execution Algos
  VWAP
  TWAP
  SOR
• Quant Based Algos
  Mathematical model based
• Charting algos
  Indicator Based

Components of Algo Trading Software

• GUI Trader Application
• User Management system
• Risk Management system (RMS)
• Market Data Handler
• Order Management System (OMS)
• Exchange Adaptors
• Algo Logic

 

What are the factors that govern performance in algo trading?

Hardware

• Server
• Ethernet card
• Switch
• FPGA

Network

• Distance to the exchange servers (Colocation plays a big part)
• Topology and rules
• Cabling
• Internet

Exchange-side latency

• Market Data (Format, Protocol used to transmit)
• Turnaround time
• Matching time
• Supported order rate
• Queueing mechanism
• Post trade risk management

Software latency (Following three are also called Tick-To-Trade)

• Feed processing
• Computation logic
• Order sending
• Algo logic
• Trading skills