top of page
robot

ALGORITHMIC TRADE

What is an algorithm?

It is expressed as a logical finite set of operations, whose beginning and end are determined to solve a certain problem. There are two approaches to problem solving, algorithmic and heuristic. Algorithmic approach is expressed by choosing the most suitable method for the solution in terms of content and specifying the steps to be applied. For the solution, the most appropriate method is selected in the light of the available data, and what needs to be done is revealed step by step. Algorithms can be defined on computers by means of a programming language. The word algorithm comes from al-Khwarezmi, who was born in the city of Hive, today's Turkmenistan. The first algorithm is mentioned in Al Khwarizmi's "Hisab al-algebra and al-mukabala" book. Algorithm EL is the Latin pronunciation of Harezmi. (1)

We see that many types of algorithms, which are encountered in computers, which now occupy an indispensable place in our lives, are becoming prominent in the investment arena. Technological developments can also provide advantages in financial markets thanks to algorithms.

Algorithmic began to be heard often in recent trade name in Turkey, while the West has long been used in trading as a decision criterion, is an emerging technology with the increased functionality with each passing day trade method. In the complexity of the market where hundreds of data flows and affects each other at every moment, algorithms draw a roadmap, enabling investors to exhibit a higher trading performance. You can get rid of the chaos in the market thanks to an algorithm that you can test its accuracy beforehand, know its weaknesses, good sides and bad sides, determine stop levels and stop methods.

rb5.jpg

How Are Algorithms Written?

A clear buying and selling rule is put forward with algorithms written using indicators that are written with the help of instant or cumulative data coming from the price of the relevant financial instrument to be traded. As a result of the realization of the trading rules, the orders are transmitted to the market with the robot software used in order transmission. With algorithms, manual operations on the human side are eliminated, and computers now work as a trader. Nowadays, we see that many brokerage houses and data terminals are taking steps to improve their infrastructure for algorithmic trading, fast order sending with high speed connections.

It is possible to generate automatic purchase and sale orders by filtering in very short time intervals thanks to the ability to digitize your own investment strategies with the help of mathematical models by using financial data. Thanks to the improved algorithms, financial statements, news, macro data can be interpreted, trading strategies can be created between levels by calculating the possible effects of the news, measuring market momentum and trading volume, and in the light of these strategies, automatic buy-sell orders can be sent by robots.

World of Speed, Effects of High Frequency Processing

Algorithmic trading is only part of the transactions made by robots. In addition, there are also transactions translated into Turkish as high frequency transactions, which are defined as high frequency trading. It is estimated that 60-70% of the transaction volume realized in global markets today is realized through high frequency transactions. (2) It is worth noting that this transaction volume fluctuates according to the changing liquidity conditions in global markets. With high-frequency transactions, transactions can be carried out within milliseconds by obtaining low profits with small price steps in high volumes. The holding period of the positions taken varies according to the content of the algon, news, data flow, transaction volume and high volatility.

The high frequency transactions using the algorithms and robots mentioned above differ from the algorithms used by individual investors in the market at two points. Location and bus provided. Speed is prominent in high frequency operations. For this reason, the level, volume and price information of the instrument to be traded can be obtained directly from the stock exchange. Orders are transmitted directly to the stock market without any intermediaries. Transactions are carried out in very low price ranges, in tiers, but in large quantities. For example, with the Borsa Istanbul Collocation Service, market participants have the privilege of keeping their systems at the same location with the Exchange systems. Colocation customers who move their systems to the Borsa Istanbul Primary Data Center can directly access the Stock Exchange in the fastest and safest way. Various market participants such as Borsa İstanbul members, HFT / Algo companies, service provider institutions and data broadcasters prefer colocation service to access exchange systems at the highest speed. Total access latency (software + network latency) based on May 2015 data by colocation; It is less than 1 millisecond for the Equity Market and less than 1.2 millisecond for VIOP. Considering that a millisecond is one thousandth of a second, the speed achieved is of great importance for investors.

In order to ensure equality between all customers who will benefit from the colocation service, the cable lengths between the cabinets and market systems are the same, equality is ensured among customers. Cable length may sound like a funny expression, but the length here can cost nanoseconds when considered in terms of speed. A nanosecond is one billionth of a second.

Since the Borsa İstanbul colocation area can be used as an alternative access point, this may mean saving from the current data center operating costs and connection costs. It stands out as the benefits of colocation in technical infrastructure facilities for lowering operational risks and high frequency algo operations. The information that the opportunity to access the order base market data published in ITCH format, which includes the information of all orders entered, will be given only to customers who benefit from the colocation service, is stated in the "Borsa Istanbul Collocation Service" brochure. (3)

robot

You can see the performance of the buy-sell rule in previous periods. Historical data can be stored in data terminals. In this way, you have the opportunity to see how your portfolio is shaped if you stick to the algorithm you have written within a specified period. You can find out how much the buy-sell criterion you have written has provided you in the past, where you made mistakes, missing points, etc. you can see

Algorithms offer the opportunity to intertwine different trading strategies. The trend tracking stop offers the convenience of following micro trends within a certain main trend according to market volatility. Since computers decide the trading transactions with the algorithm, many transactions take place in a short time. In micro trends, these transactions can be realized with low profits with the appropriate trading volume of the market, but due to the high number of them, they can stay in a trend and get a higher return than the return from trend tracking. It offers advantages especially for investors with low transaction costs. It saves you the trouble of constantly monitoring the market. While the market is monitored, the price, which acts according to the instant data, affects the buying and selling decisions on the people and causes the wrong decisions to be made. After these price movements, positions can be changed frequently, forgetting the main trend direction. In addition to causing high transaction costs, this situation can cause capital losses. Thanks to algorithms, these problems can be eliminated as they act within a certain trading strategy.

In the market where there is a continuous data flow, it is often not possible to filter these data through a logical filter and make a decision. When the ambition to win and the drive to avoid loss are added, it means that the hours of torture have begun for a trader who has entered the position until that position is closed in profit or loss. Thanks to the algorithms, the option that suits you can be selected from among the unlimited options.

Silahlı Robotlar

Algo Robots and Their Problems.

With algo robots and high-frequency transactions, which we have mentioned above, there are also negative aspects of order sending like a knife. It may open the door to malicious traders, especially those who want to manipulate the market. In addition, the lack of technical knowledge of individual investors should not be overlooked. In addition to the technical infrastructure and support for the individual investor, the algorithm it uses must have a projection that protects the investor in the first place under every market condition. Due to the fact that there are more than one investor using the same algorithm, everybody expects the same buy or sell rule, so transactions may not take place because the price does not move. Even if there is no transaction in a certain trend, a price movement can be created by inflating the demand or supply in terms of the price direction by sending orders as if the transaction is made in that direction. The market can be negatively affected by order cancellations. While a high depth appears in the treated board, a shallow board can be created by canceling orders. After insider trading, faster order submission and high-volume transactions can be done in the market. This is a situation that damages equality in the market and justice.

Excessive optimization in studies on historical data, indicator fitting error in the trend of the price can cause hunting while going hunting. Mistakes made in tests that cover certain periods and are used to find the optimum statistical result can lead to a trade disaster by causing the error of fitting the indicator to the price. Especially in changing macroeconomic conditions, algorithms created with only technical analysis decision criteria may cause negative results in the future.

Lack of sufficient market depth. In order for algorithms to work properly, technical indicators calculated with the help of six data taken from the price must be able to objectively define the price movement and stable. Such a structure is possible if the market in which the relevant financial instrument is traded can be liquid. The realization of the orders sent by the algorithm is healthy if there are enough buyers and sellers in the market.

Lack of sufficient infrastructure. Internet outages, computer capacities that calculate the algorithm, speed of sending orders may be external factors that can affect you negatively. If there is not enough infrastructure, the written algorithm will not matter.

Limiting the periods of system tests in the process of creating an algo. The mistake of accepting healthy signals only in trending markets or only in horizontal markets as if it is the reality of the general market can cause portfolio meltdown when market conditions change. In the event that algorithms that detect this in a clear trend in a certain direction come into play, the decline or the rise is supported by harder price movements. In this case, the formation of volatile movements, triggering of movements in the financial markets, exaggerated decreases or exits may occur. Price movements triggered by high-frequency transactions in derivatives markets can bring along sharp declines or exits.

In the world where computers are in our pocket, they have their share of the computer environment in financial markets. Now computers also entered our wallet. With its positive and negative sides, there is now an algo robot reality in the financial markets. Will the effects on humanity be positive or negative as the computers that we transfer the decision process to with a click of a button are heading towards a new beginning? It is too early to answer this question. But there is an unchanging truth that whether you use algorithms, robots, or do manual action. In any case, there is only one truth that you are alone with, if there is no personal discipline, it is all empty. Individual investors who send robot orders by putting trading rules into algorithms, or those who send orders using high-frequency transactions, will only continue to create trading volume in the market if they do not have the patience to comply with these systems.

RESOURCES

1 Introduction to programming with “C” Asst. Assoc. Dr. Hasan H. BALIK Elazig 2003

2 Fast Guys Michael Lewis 2016 Public Portfolio Publications

3 www.borsaistanbul.com/urunler-ve-piyasalar/teknoloji-servisleri/colocasyon/resources

4 Your Money and Your Brain Jason Zweig 2011

5 Investor Psychology Dr. Serpil Döm 2003 Change Publications

AUTHOR:

Ali Erkan TANACIOĞLU

“Investment information, comments and recommendations contained herein are not within the scope of investment consultancy.

Investment consultancy service; It is offered within the framework of an investment consultancy agreement to be signed between brokerage houses, portfolio management companies, banks that do not accept deposits and the customer.

The ratings contained herein are based on comments and personal opinions. These views may not be suitable for your financial situation and risk and return preferences.

Therefore, making an investment decision based solely on the information contained herein may not produce results in line with your expectations. "

Benefits of Using Algo Robot

Until a normal investor makes a decision for trading and takes this decision into action, the algorithms perform their transactions quickly within milliseconds. Even if the appropriate timing in the direction of the trend is caught, the transactions that take place within the volatility created by the robots make the investors nervous and the trader fails to show his patience while losing, and makes the mistake of leaving the position early.

“Financial decision making is not just about money,” says psychologist Daniel Kahneman of Princeton University. It is also concerned with spiritual motives such as avoiding regret or being proud. ”(4)

Psychological and cognitive biases, herd behavior, mental accounting and investor sensitivity (5) based errors encountered by traders and defined as investor anomalies in behavioral finance can be prevented by algorithmic operations. I sold it, it turned out. I bought it and fell. It dropped to the stop level but returns from here again. Early to get into position or early to get out of position. As if this share is more expense, judgments based solely on feelings, investors' fears and empty hopes can be eliminated to a large extent.

In case of data pollution in the market, it speeds up your choice, provided that you define in advance which data you need to pay attention to. In markets that are constantly bombarded with new news data, a news algorithm can even be written about which news can be effective and to what extent. In addition, according to technical analysis criteria, even the increases in trading volume, the slopes of the emerging trends, the weakening of an existing trend and the breaking of the support-resistance levels can now be written using the algorithm, and orders can be sent to the market within these rules.

bottom of page