Yesterday we published instructions for newbies in cryptotrading, and today we will talk about automated trading, or robotrading.

Robotrading is the process of using a set of instructions for placing bids to buy or sell shares, money, or cryptocurrency. The instruction set is based on various market indicators, including price, time, volume, and others. When traders are getting tired and do not have time to trade manually, they think about moving to robotrading. But the threshold to enter this area is quite high, especially for a person who has no special financial or technical education. We asked Arthur Shponko, the leader of the automation strategy in the department of trade services and applications at FINAM, to tell us about how robotrading works, what its benefits are, and how to earn in automated trading.

About robotrading hobby

Since my childhood, I have watched a TV program called "Umniki i Umnitsy" ("Smartiepants" show) and wanted to do something related to international affairs. That whole time I thought I'd be a diplomat, but when I was in the ninth grade, my dad said that I would never make a good politician or a diplomat and suggested drawing my attention to to the economics. In the end, I joined Higher School of Economics (HSE), Faculty of World Economy and there I chose my major: international currency, finance, and credit relations. My supervisor, who introduced me to robotrading, impressed me a lot. His influence probably determined my occupation.

About the transition to robotrading and free time

If the private investor focuses on profits, understands the difference between manual and automated trading, and has some programming skills, they should at least think about switching to automatic rails. The effectiveness and scope of work are increasing: people just perform more and better. By trading automatically you can cover more exchanges, use more tools and strategies. But it doesn't mean that you don't need to monitor automatic operations like manual ones. Do not forget about monitoring, it's always important to keep your eyes open.

I believe that any trade, whether manual or algorithmic, requires effort and full involvement in the process. If you want to earn, you have to spend a lot of time on trading. Ideally, you should have a full working day, when you cannot just write code and go relax. Of course, you will not be attached just to one place, you will be able to travel around the world but only if you are always online. But it's definitely not about lying in a hammock under a palm tree for days.

About fiat money and cryptocurrency

In programming terms, there is no difference between what you are trading, fiat money or cryptocurrency. The main thing is to create a good mechanic trading system. I even admit that there are algorithms that can be used effectively whether with conventional currencies, cryptocurrencies, or stock shares. It's not important what tool is applied for entry. What does matter is whether or not the rules implied in the algorithm describe behavior of the specifically selected tool. So here is the question of the initial trade idea formulation and strategy. In any case, when an algorithm is developing, the behavior of a particular trading asset is taken into account.

About benefits, risks and foolish mistakes

The greatest advantage of automated trading over manual trading is a lack of emotions. People often make decisions based on feelings and emotions and it does not always have a positive effect on the results, especially in critical moments. In contrast, the algorithm simply performs its work without any emotions or excitement; it does not take impulsive decisions and trades on strictly specified strategy. And if the algorithm is started, the trader could not intervene and change something.

Of course, there are risks, and they are usually market risks. The algorithm trading idea may stop earning at some point when any unpredictable events and crises occur on the market. Most robots cannot trace the signal. You will have to track these cases and change your robots in time or disable them. There is also a small risk that the exchange will fail you. In 7 years of my experience in trading, I've seen probably only five cases.

Moreover, programmers could sometimes make silly but serious mistakes. I have a very good and simple example. A simple algorithm has the cost and the price. Of course, in order to make income, your price should be higher than your cost. If you confuse them, your robot will be unprofitable. Often newbies face this problem. As a result, a very simple mistake at any stage leads to enormous losses. One Russian bank lost a considerable amount of money for the same reason.

About partnership and business

I had projects that I led alone, exclusively for self-interest, and I had projects when I was invited to be an independent tester. When you choose your format, your decision should be based on the goals that you are pursuing. The teamwork quickly solves logical and mathematical problems, and you constantly share your experiences and observations.

But there is usually a clear division of labor in teamwork, so if you want to cover and execute all processes yourself, try to trade alone. Moreover, each your know-how invention within the company will be publicly available to all your colleagues or even to the entire society. Maybe you should consider the algorithm propagation rules in advance.

As far as business is concerned, it seems to me that the person who understands finance and programming should be the head of such business enterprise. If you do not understand the robotrading and just want to make money by bringing a team of programmers and traders under your authority, you will be definitely deceived by your staff members. In this business, control is very important as in any other.

About skills

Any specific skills will be very useful. By saying any specific skills, I primarily mean good knowledge of mathematics and logic. It's more about clear logic and understanding of how to write a formal trading rule rather than about mathematics because you don't need any algebraic transformations at the early stages. There is a clear distinction between the feeling of how to trade and the ability to describe the trade logic in a formal set of events and actions. At this step, we need mathematical knowledge. Also, programming skills and command of basic program languages (for example, "the C") are essential.

I have been interested in the markets since the second year of the college. I am that man in a thousand who has continued to evolve in my specialty after graduation. I have been doing the same for ten years; I've been improving my skills in robotrading. Even when I was a student I had classes, such as finance, financial mathematics, and physics, that helped to understand how the market works, the quotes movement, and the laws that may or may not determine market movements and forces.

About the pitfalls

Sometimes you seem to come up with a profitable strategy, stunning and working trading idea. You weigh it up in any mathematical modeling program, for instance, Mathcad. Then, you look at the data, design a data model, and make required calculations. After that, you include all the commissions and surcharges, and everything works fine. (By the way, by no means, don't kid yourself, include all the payments, it's vital). But then you realize that technical implementation of your idea needs more money than you can earn from this idea.

Once upon a time, one private investor made a magnificent trading robot, named Panda. It generated an 8,000% return over the course of three months with almost no initial risk. Panda was discussed and hyped a lot, but it really worked well. But the thing was that the robot made this money from the initial capital in 50 thousand rubles, while the monthly support for the entire technical infrastructure required to sustain the project cost around a half million rubles.  Thus, it was initially an unprofitable project, because it couldn’t even cover all costs for electricity, powerful servers and computers carrying out calculations for the project Especially, one should remember that the robot earnings are not always high, but the money for maintenance is to be paid steadily. So, it is important to calculate not only how to earn, but also how much you have to spend in order to earn.

About fast money

It is clear that robots can be configured differently, it depends on the volume of trade, what leverages and loans are used when trading, whether you invest money on your own or on someone’s behalf, etc. There are a lot of nuances. But in general, you can earn little by little every year with almost no risks by trading for ten years at 10% per year, or you may try to trade one year at 100%.

But there is a law in the stock market that you won't be able to cheat — the higher the potential return, the greater the potential risk. It's as strong as the rule of supply and demand. So, if you count on big money, be ready for big losses. This rule cannot be avoided and should be taken into account when it comes to return on investment (ROI) or algorithmic trading profitability.

About platforms for newbies

There are some platforms that allow anyone without programming skills to collect full mechanical systems and launch them in trading. For example, TSLab is the most famous platform, but there are many others. It is a good option for a newbie to try and see how it all works. It's not necessary to use them all the time. If a person has programming skills, they can write algorithms after delving into the subject. It is important to understand that the creator will never just give away their idea that will bring money to the investor. The idea is to be sold through social trading, like Percentage Allocation Management Module (PAMM) when inexperienced investors have the chance to choose a successful trader and to trade on their strategy without any access to their code. So, they won't know what's inside the black box but will enjoy the results of the trade.

Therefore, the platform like TSLab is enough to realize some software opportunities and to understand the simplest strategies.

About books and sources

There are practically no Russian sources about robotrading. Basically, all of the books are in English, and they are quite famous. Perhaps the only thing I can advise is to find research papers and lectures of Alexander Gorchakov, the legendary trader. He was one of the first who began to formalize market behavior. But his works contain higher mathematics market analysis with integrals, probability theory, chaos theory, and so on. This information would be more useful for improving skills and deeper understanding of the process. One of my favorite books is "Options, Futures, and Other Derivatives" John Hull. It is for anyone who is interested in the derivatives market.