Created on 09/04/2024

High Frequency Trading or HFT has graduated from the sole domain of institutions with powerful technology to retail traders trying to emulate their success. The growth of personal computing power, improved trading platforms and high-speed Internet access have put HFT on a level playing field for anyone who wants to try it.

Is retail HFT trading with bots really possible? Let’s explore this exciting field and see how marketers can take advantage of it.

Although there is no single definition of high-frequency trading, most traders agree that it has the following common attributes:

1. A large number of trades. As the term “high frequency” suggests, HFT is characterized by a large number (ie frequency) of trades – often hundreds of trades (yes, hundreds!) in a single trading day or session.
2. Short retention periods. High-frequency trades are open for a very short time – seconds or even fractions of a second, and a few minutes at most.
3. Automatic algorithmic trading. Computer-based algorithms or rules automate the decision-making process and the execution of trade entry and exit. There is a practical element to this, as it is nearly impossible for people to efficiently enter and exit a trade in seconds or fractions of a second.

For this article, I will use the above three characteristics (large number of trades, short holding period and automated execution) to define high frequency trading. Within this framework, I can simplify the definition to: “automatic trading in small time frames”.

In this article I will focus on the retail side of high frequency trading. If you’re interested in the institutional side, the book Flash Boys by Michael Lewis (who also wrote The Big Short and Moneyball, both of which were made into movies) is a well-researched and interesting explanation.

There are two key differences between manual trading and HFT bot trading:

1. Manual trading means that the trader must physically enter and exit each trade. HFT bot trading automatically enters and exits trades. Manual trading can be time-consuming and require a lot of screen time, depending on the trading strategy.
2. Manual trading can be discretionary or systematic (rules-based). HFT bot trading is always systematic.

Discretionary trading means that the trader relies on his own judgment (or discretion) when trading. For example, I trade discretion when deciding whether price is forming a chart pattern. Discretionary trading has room for interpretation and is subjective – two traders looking for the same pattern may see the chart differently or enter at different levels. In contrast, systematic trading relies on objective rules with no room for interpretation. Two traders following the same systematic strategies should have identical entries and exits.

1. A complete trading bot must have three parts: rules for entry, stop-loss and take-profit.

An algorithm must cover all three areas to be a true bot. Otherwise there will be an element of manual input.

2. Common HFT strategies include:

• Market making: buying and selling on both sides of the market.
• Statistical Arbitrage: Using algorithms to identify mispricing or market inefficiencies.
• News-Based Trading: Using algorithms to analyze news and market data that can potentially move price and trade on it instantly.
• Momentum or technical trading: having an algorithm to identify technical patterns and trade them.

3. Swing and scalping strategies are most accessible to retail HFT traders.

Scalping is very similar to high frequency trading and if you feel you have the skills to program a profitable scalping strategy into a bot, that’s a great place to start. In contrast, using HFT for market making, statistical arbitrage and news-based trading requires technology such as optical connectivity to exchanges, which is not available to most retail traders.

4. After back-testing, forward-testing in real time.

The bias with back-tested strategies is that they may fit past data perfectly (known as “curve fitting”), but they don’t work in real time. To prevent this problem, pre-test the strategy in real time to make sure it produces results similar to backtesting.

5. Test the HFT bot with a small live account for realistic filling.

Demo accounts have perfect loading without slippage or liquidity problems. The results of high-frequency trading with real money are more sensitive to broker fulfillment than long-term trading. If I were a swing trader aiming for 100 pips of profit per trade, the difference of 0.5 pips of negative slippage per trade is insignificant. However, if I was trading an HFT strategy that records 50 trades per day, a negative 0.5 pip difference in execution means 25 pips less each day in profit. It could mean the difference between a winning and a losing strategy. Testing on a small live account will show if the HFT strategy works in real market conditions.

6. Direct Market Access accounts versus Market Maker accounts may produce different results when trading on smaller time frames.

Direct Market Access (or Non-Dealing Desk) accounts transmit orders directly to their liquidity providers. Market Maker accounts fill trades internally using the Dealing Desk. The price swings between the two account types differ significantly on small time frames. Many HFT traders prefer direct market access accounts (such as ECN and STP accounts) for several reasons:

• DMA accounts show every tiny fluctuation in the price movement of the underlying market, which is convenient for HFT bots trying to take advantage of small price movements.
• DMA accounts often have lower fees with variable spreads that can sometimes be 0 pips. Since HFT means placing a large number of trades, smaller spreads greatly aid profitability.

7. Consider buying a bot instead of building one yourself.

Bot trading has a vibrant community, with developers offering HFT trading bots for sale or rent. MetaTrader Expert Advisors (EA) are the most widely used bots for retail traders today. Last time I checked, the MetaTrader bot market had 1,700 bots available. The best high frequency trading platforms will have their own communities that offer bots from other traders.

8. Turn off the HFT bot for trading in market conditions that do not match the strategy.

For example, some automated strategies are not good around big news announcements due to slippage and large spreads. It is easy to turn off the HFT trading bot around these events or change markets.

Avg

1. HFT trading bots can automatically generate a large number of trades that would be difficult or impossible to do manually.
2. One of the biggest hurdles traders face is emotional discipline and trading according to their strategy. Automated trading removes the need for constant discipline.
3. High frequency trading can shorten losing streaks. For example, if I was a swing trader and suffered 4 losses in a row, it might take me weeks or months to break even. However, if I make dozens of trades a day, I could recoup my losses that day or that week.
4. Hundreds of HFT trading bots are available to buy or rent on MetaTrader’s marketplace and many more from other sources. The MetaTrader market is one of the reasons why automated Forex trading has become so popular.
5. I don’t need to be on the screen the whole time the trading bot is running because it automates the inputs and outputs. In comparison, manual trading of a large number of trades requires a physical presence on the screen to execute the trades, and may not be suitable depending on one’s lifestyle (eg if one has another job or one’s own time zone). When I moved to Canada from the UK, I realized that it was almost impossible to manually trade the London Open because it was 3am local time.

Cons

1. HFT aims to capture very small movements, and there is a fine line between profit and loss.
2. High frequency trading is highly dependent on brokerage spreads and slippage.
3. Many HFT trading bots I have seen have low reward/risk ratios, sometimes as low as $10 in risk for every dollar in profit. These types of bots rely on a high win rate to make money, but a small number of losses can significantly damage an account. (I want to emphasize that not all HFT trading bots have low reward/risk ratios.)

High frequency trading bots allow retail traders to access a trading style that can generate profits on a daily basis without having to sit in front of a screen all day. There has been tremendous development in this area, with hundreds of HFT trading bots available to buy or rent from sources such as MetaTrader. High-frequency trading is essentially automated scalping, which means taking small bites from the market multiple times a day in order to make a profit. Broker fill and spread must be excellent, as poor fill in a large number of trades can significantly damage profitability.