High-frequency trading has become an essential part of financial markets, representing 50% of the trading volume in the US equity markets. Due to its extreme importance to our current markets, it is very important for investors to understand how high-frequency trading works.
What is high-frequency trading?
High-frequency trading (HFT) is a type of systematic trading in which a high number of orders are processed in fractions of a second. The goal is to make numerous transactions with small margins. High-frequency trading also involves very short holding periods. While some traders will hold positions for years, and HFT portfolio ends the trading day without any position.
Therefore, it uses advanced algorithms, and automated trading platforms to select and execute trades. High-frequency trading is mainly used by investment banks, institutional investors, and hedge funds. HFT is a subset of algorithm trading.
High-frequency trading systems enable traders to execute millions of orders and monitor a wide range of markets and exchanges. HFT firms will usually have their operations close, or in the same building as the exchanges, in order to access the data feed faster. This allows them to have information slightly before the rest of the market.
High-frequency trading is characterized by:
- Predefined trading strategy, and decisions
- A large number of orders
- Automated management of orders, and the portfolio
- Profit from buying and selling, with very low margins on each trade
- Proprietary trading algorithms
- Completely done by computers
- Not holding any positions at the end of the trading day
- Fast order cancellation
- Proprietary trading algorithms
- Very short holding periods
- Trading very liquid assets
- Co-location and low latency are crucial
- Used by large hedge-funds, broker-dealers, and proprietary firms
How does high-frequency trading work?
High-frequency trading is dependent on processing a large number of orders in a short period of time. Computers are faster than humans and are capable of processing a lot more information than we can in a very short span of time. HFT firms develop algorithms according to a certain strategy in order to buy and sell multiple highly liquid securities.
This is why high-frequency trading firms are dependent on algorithms that process orders much faster. Thus, it requires getting in and out of trades, almost instantly. Instead of taking a few seconds to process a transaction, high-frequency trading systems take milliseconds or even microseconds in order to buy and sell a certain financial instrument. They profit from minor fluctuations in the bid-ask spread, while at the same time providing liquidity to the markets.
What is the importance of data speed for high-frequency trading?
High-frequency trading firms rely heavily on the fastest real-time market data. This means that they are usually located close to the exchanges, to have low latency. In order for their algorithms to process all of the information, they need to have access to real-time data from the exchanges at the fastest speed possible.
Getting the data from the exchanges can make a large difference for high-frequency trading firms. As it gives them an advantage over the competitors. Knowing the current price, the bid, and the ask, faster than your competitors allows you to place orders before them.
Therefore, high-frequency trading firms compete for the fastest data. They try to have their locations as close as possible to the exchanges, in order to get the data first. HFT firms often use private proprietary technology, in order to determine the best trades and in order to process large amounts of data.
What are the types of high-frequency trading firms?
High-frequency trading firms are roughly classified into three kinds:
- Independent proprietary firm
- Broker-dealer firm
- Hedge-funds
The independent proprietary firm is the most frequent and largest type of high-frequency trading firm. Proprietary trading is done using the firm's own capital, these firms usually do not receive external capital. Therefore, the profits are shared among the shareholders.
Some high-frequency trading firms are divisions of broker-dealer firms. Many traditional broker-dealer businesses have a division known as proprietary trading desks where HFT is performed. This section is independent of the work the firm does for regular, external customers.
Finally, some high-frequency trading firms function as hedge funds. Their primary goal is to benefit from inefficiencies in price across securities and other asset classes through arbitrage.
High-frequency trading approaches
There are different strategies that usually fall onto 2 different approaches:
- Passive
- Active
Passive
Passive high-frequency trading strategies are focused on providing liquidity. This strategy includes creating both a buy and a sell order for both the bid and ask. The goal of this strategy is not to invest or speculate, but to simply skim the spread, between the bid and ask.
Additionally, this is extremely beneficial for both institutional and retail investors. As it provides liquidity to the market and increases order execution.
Active
Active high-frequency trading strategies involve anticipating orders from other market participants and profiting from them. This strategy includes for example monitoring index funds, and ETFs, and buying or selling securities based on the portfolio rebalancing of these funds.
However, there are other active strategies that include anticipating momentum in a given security. This type of HFT strategy is speculative in nature and relies on both fundamentals and technical indicators. Thus, the goal is to try to place orders before other investors, and profit from them in a very short time.
Advantages of high-frequency trading
There are two main advantages of high-frequency trading:
- Narrower bid-ask spreads
- Increased market liquidity
- Improves market efficiency
- Reduced transaction costs
High-frequency trading has increased market liquidity, which in turn has contributed positively to narrowing or even eliminating bid-ask spreads.
This was verified by increasing HFT costs, which caused bid-ask spreads to rise. One research looked at how Canadian bid-ask spreads altered after the government imposed HFT costs. It discovered that market-wide bid-ask spreads grew by 13%, while retail spreads climbed by 9%.
Therefore, it is undeniable that high-frequency trading has a net positive effect on the market, and it actually benefits all of the market participants. By constantly improving market efficiency across the board.
Who are the largest high-frequency trading firms?
The largest high-frequency trading firms include:
- Tower Research Capital
- Hudson River Trading
- Citadel Securities
- XR Trading
- GTS
- Optiver
- Flow Traders
- IMC Financial
- Tradebot Systems Inc.
- Virtu Financial
- XTX Markets
- Jump Trading
- Quantlab Financial
- Two Sigma Securities
- DRW
- Allston Trading LLC
- Five Rings Capital LLC
- Jane Street
- Geneva Trading
Is high-frequency trading legal?
Albeit several ethical and professional concerns, HFT is still legal in most jurisdictions around the world. Regulators offer varying definitions of high-frequency trading.
There are certain rules that keep a keen eye on the misuse of HFT. However, there are no set rules that define HFT as an illegal activity as yet.
For instance, the financial industry regulatory authority (FINRA) in the US requires HFT platforms to follow compulsory risk management to avoid misuse of HFT such as spoofing and fictitious quoting.
Similarly, the ESMA in the EU requires HFT firms to keep their time records of trading data for up to five years to boost transparency and avoid misconduct.
What is wrong with high-frequency trading?
Critics of HFT argue that it offers an unfair advantage to institutional traders against retail investors. Since HFT requires expensive tools and sophisticated skills, retail traders would lack access to such technology.
Institutional investors can execute hundreds and thousands of trades within seconds as they rely on tiny profit margins. Virtually, that would be impossible for retail traders to replicate or compete.
Also, critics argue that it offers false liquidity to the market. By the time retail investors would look to buy or sell a specific security, HFT traders would have completed their trades.
Can you do high-frequency trading from home?
Theoretically, anyone can perform high-frequency trading from home, office, or through a broker.
However, to be able to compete directly with other HFT firms you require sophisticated technology, special skills, and high capital investment. You will also need to learn about HFT.
As a retail HFT investor from home, you’ll need:
- Fast computers and the internet
- Co-location set up to keep your HFT system as close to exchange serves as possible
- Access to real-time data and fast analysis
- Programming language skills such as C++, Python, R, Java, etc.
- Technological requirements such as Xilinx field-programmable gate arrays (FPGAs)
Can HFT lose money?
Like any other trading strategy, HFT can lose you money as well. An erroneous algorithm can execute faulty orders resulting in huge losses within seconds. Similarly, faulty algorithms can execute several unwanted trades such as stop-loss orders for no reason.
Even without errors, algorithm trading can execute stop-loss orders that a human trader wouldn’t. Another big risk comes from intensified volatility due to increased liquidity in the market. As HFT pulls (and pushes) demand for certain securities, their demand-supply curve behaves indifferently.
Thus, retail investors as well as HFT traders can experience increased volatility and risks of losing investments.
Is HFT profitable?
High-frequency trading chases small profit margins that the naked eye cannot detect. However, when multiplied by hundreds of thousands of trades in a single trading day, it brings exceptional profits for investors.
As with most trading strategies, there is no upside limit to the profit potential with HFT. However, HFT traders usually take a fraction of a penny per share that is then multiplied by the number of executed trades. Therefore you can see how profitable it can be.
How do I become an HFT programmer?
You can become a successful HFT programmer by learning both soft and hard skills for HFT systems. Here is a roadmap for you to become an HFT programmer:
- Formal education in computer sciences, MIT, etc.
- Technical and industrial expertise through working on computing, big data, and analytical projects
- Gaining financial exchange knowledge
- Learning programming languages that build HFT systems such as Java, C ++, R, and Python.
- Excellent knowledge of fast processors and hardware such as Xilinx field-programmable gate arrays (FPGAs), GPUs, etc.
- Skills for speed execution such as kernel optimization like Linux Kernal optimization
Conclusion
High-frequency trading has biome an essential part of the financial markets. Liquidity is essential for every investor, both the large institutions and retail. Despite the controversy surrounding HFT, there are a lot of benefits that clearly outweigh the possible negative consequences like flash crashes.
Featured image source: Unsplash