Algorithmic trading uses computer programs for sending trade orders with algorithms deciding on aspects such as timing, prices, and quantity of the order with lack or even without human intervention. (Related post: Statistical Arbitrage)
This kind of trading is widely used by buy-side (investor-driven) institutions and traders in general. Large amounts of trades are divided into several smaller to manage market impacts and risk.
Reasons for using algorithmic trading are many, but here are the most relevant:
- Identifying and reacting to opportunities faster.
- Avoiding own pitfalls in trading psychology which could harm long-term performance through reducing emotions and increasing discipline;
- Minimizing human errors and tracking more markets simultaneously among others.
However, this class of systematic trading has its inherent drawbacks:
- One of the biggest downsides is the technical insolvency and resources required.
- Lack of control that comes with automation.
- Most of the programs were backtested in environments without high volume managed by algorithms, so, underestimated risks could take a prime role.
If you want to get a deeper insight, I recommend you review this course offered by Quantra: Algo-Trading