Algorithmic trading brings unique strengths of computer systems to traders that traditional trading does not have. Taking advantage of the algorithmic potential can dramatically change investment performance.

The Law of Large Number

In probability and statistics, the law of the large numbers states that if the same experiment is repeated independently a large number of times, the average result must be close to the expected value. Alternatively, even with random and unknown events, a sufficiently large number of attempts will lead to a stable average result.

The law of large numbers can be illustrated by rolling the dice. The probability of the occurrence of all sides 1, 2, 3, 4, 5, and 6 is the same. Therefore the expected value of a dice roll event is:

(1 + 2 + 3 + 4 + 5 + 6)÷6 = 3.5

If one only rolls the dice a few times, the average value of the results may be far off from this expected value. For example, roll the dice 3 times and the results are 6, 4, and 5 respectively. The average value is thus 5. However, according to the law of large numbers, if one rolls the dice many thousands of times, the average result will come closer to 3.5.

In trading, the winning or losing results of a handful of transactions can occur randomly and unpredictably. According to the law of large numbers, the results from a small number of trading orders are not sufficient to judge its effectiveness. However, if there’s a reasonable trading strategy with a winning expectation over 50% across a large number of trades, then on average it will generate profits in the long run even though there might be periods of consistent losses or consistent wins. Moreover, these results can be used as tests to evaluate a trading strategy’s performance.

In practice, there’s a big issue when applying the law of large numbers to real-life trading. It’s often difficult for investors to ensure all transactions are sufficiently consistent according to a predefined strategy, due to the following factors:

• Effects of psychology and emotions;
• Risks such as mistakes in placing orders, or slow reaction time that leads to missing buy or sell opportunities;
• Unable to place orders according to the predefined strategy due to irregular price monitoring.

Because of these issues, the results obtained can be far from the initial expectation despite an effective trading strategy and a sufficiently large number of trades. More importantly, it’s impossible to assess whether this result is due to a low-performing strategy or poor execution. An algorithmic trading system can solve these problems. It ensures the application of the law of large numbers, minimizing the impact of undesirable factors. Investors only need to focus on building, evaluating, adjusting, and improving their strategies to create stable long-term profits.

Ensure Investment Consistency and Eliminate Emotional Interference

In stock trading, besides a clear strategy, emotional management is a key factor of investment results. Investors are often swayed by emotions and make decisions contrary to their investment principles or strategies.

For example, before the 2008 global financial crisis, the market signaled that a crisis was imminent. Yet a lot of investors ignored these signs because they were caught up with the “frenzy” up-trend momentum in the market. Few thought a crisis was likely to happen.

Algorithmic trading solves this problem by ensuring all trades follow a predetermined set of rules and strategies. It eliminates the emotional influence of trading altogether. A simple example is the 7% stop-loss principle, common among Vietnamese investors to avoid big losses. However, very few investors actually carry out this trading principle. They often keep holding on to falling stocks even when the price has fallen too low. A trading algorithm will help investors avoid this. A stop-loss strategy does not guarantee good results, but it does ensure consistent investment principles.

There are random unpredictable events every day. They are out of control yet have a direct impact on the market and stock prices. These events require investors to make a decision to buy, sell or keep holding. Making uninformed decisions gives most investors ambiguity, stress, anxiety, and insecurity.

The algorithmic trading system is not a magic wand that can tell the future and ensure all trading decisions are correct. However, it is an automated system that quickly reacts to random events that can happen without human intervention.

Investors only need to set up periodic reviews, monthly or quarterly, to evaluate an algorithm’s performance. They can make necessary adjustments and upgrades to optimize profits. This mechanism boosts investors’ confidence and reduces stress because there’s no need to be hands-on every day.

Minimize Market Impact When Trading Large Volumes

A high-volume transaction from institutional investors, hedge funds, or major shareholders can dramatically change the market equilibrium. To avoid this, investors usually split into several small-volume trades.

For example, if you would like to buy 5 million shares of FPT, investors can divide it into 1000 orders, each at 5000 shares per order every 5 minutes. After one hour of this execution, the investors assess the impact of the orders on the stock price. If the share price doesn’t change much in the unfavorable direction, keep following the same strategy and check in on the price every hour until reaching the target of 5 million shares.

However, there’s a time disadvantage to this strategy. In this example, even under favorable market conditions, it still requires at least 83 hours or 21 trading sessions to complete.

An algorithmic trading system solves this problem by simultaneously calculating the volume and making the stock purchase. It also checks the market impact immediately to significantly reduce the number of buy orders and the time required.

Save Time

A complete algorithmic trading will ensure full execution of the investment process without any investor intervention. It is the best benefit as investors can spend their time more efficiently instead of closely monitoring live boards all day long to place orders.