Hello dear friends, investors and traders!
In light of the recent achievements of my UAE account, I would like to take this opportunity to tell you a little about myself, my trading philosophy and the principles I follow in trading. Communication between the Manager and investors is extremely important, because people who invest their money should at least understand who they trust. Whether this person is sane, whether he is sufficiently qualified in the field of trading and Finance.
A little about yourself.
My name is Vitaly, I am a co-founder of the DaVinci FX Group project. This project is primarily dedicated to automated trading on the Forex market. The main mission that we pursued when creating it is to tell novice traders about the aspects of automated trading in financial markets. Here you will find all the best and most relevant in the world of algotrading. Author’s free development of professional programmers of the DaVinci FX Group project will be highly appreciated not only by beginners, but also by more experienced traders. Currently, there is no high-quality tutorial guide in open sources that will guide a novice trader along the path leading to professional profitable Forex trading using trading algorithms. With our project, we want to change this!
Why the team, you may ask. The answer is simple - in a team and only in a team of qualified and purposeful associates, it is possible to implement serious projects in the Forex market. The market can be compared to the ocean raging during a storm, and sailing alone in a boat is much riskier than on a ship with a team of professionals, with a clear division of roles and functions. There is such a concept as a synergistic effect, when the sum of the actions of individual components during interaction and cohesion gives a much greater effect. So this concept is the best way to get an answer to this question.
I, as well as many who came to the market, started their trade with their hands. I got my first Forex trading experience in 2007, during which time there were of course UPS and downs. The main thing with experience came the awareness of what works in the market, and what does not. And the main thing that allowed us to reach another higher quality level in trading is the understanding that trading should be system - based and automated! This way and only this way, you can achieve a positive result over a long period of time in trading on the Forex market. Now I just do not even imagine how 3 years, up to 2011, I traded hands, it was a real test! Those who sell their hands will understand me. I have been a PAMM Manager for the last 5 years (one of the trading accounts - http://www.myfxbook.com/members/vetallnemo/nec-convertebantur/1959349), as well as a signal provider on the MQL site - https://www.mql5.com/ru/signals/mt4 As you can see, my trading account Firmum Fiducia Motus is in the top 3 signals for the reliability of investments. The road to this I can call a time of trial and error, the next trial and finding those systems that are potentially profitable for a long period of time. I emphasize that there are no potential risks, Grails and systems of infinite profit in the market, unless it is based on non-trading principles, such as arbitrage and so on.
All our PAMM accounts and signals are based on our own developments in the mql4 and mql5 programming languages. But before trading advisors get into the trading portfolio, they go through a long cycle: searching for an idea – writing an algorithm – testing for errors – optimization.
Trading philosophy, development and optimization of trading systems
For so many years in the market, a huge variety of trading strategies, tactics and approaches to the market have been tried. The first step was to filter out trading strategies based on order grid construction, hedging, and martingale principles. I can’t say that it can’t make a profit, we have a trading adviser created in the DaVinciFX Group team of programmers, which works successfully and makes a profit. But this approach to trading is not close to me. I only use stop trading systems that have a take profit and stop loss. The process of developing and optimizing a profitable trading system is roughly similar to searching for and cutting a diamond. At the same time, as you understand, millions of lines of written code just go to the dump of program code. I will say more, only 3-5% of expert advisors that are written at all are potentially profitable.
After the code is written, the adviser is refined and tested by a team of like-minded programmers and testers of the DaVinciFX Group. And only after all the procedures for improvements and bug fixes that inevitably manifest themselves, we proceed to optimization.
For optimization, we have specially developed an approach that allows us to determine the pool of the best parameters of the trading system using several passes of the genetic algorithm in the meta trader 4 tester. A prerequisite for optimizing the system is to pass the received set with the same indicators as during the optimization period, the backwardness and forwardness of the test. This is the only way to protect yourself from over-optimization, which is one of the main problems of novice traders using automated trading. At the same time, the period for which the system is optimized is at least 4 years, the forward and backward testing is 30% each. At the same time, sets are selected with the highest number of completed transactions, at least >1000, which allows the sample to be representative.
Fundamental base for testing
You can devote an entire article or blog to this, but I will limit myself to the most important. A mandatory condition for checking the trading algorithm in the strategy tester is the presence of tick quotes of the best quality. In this case, we are best helped by the Tick Data Suite software package, which allows us to download quotes from the best brokers (including Darwinex) and perform testing and optimization on them. At the same time, fine-tune the trading conditions, thereby simulating trading on the real market as much as possible.
Main tools of Tick Data Suite:
- real spread;
- simulation of the slippage of orders;
- selecting the ability to download quotes from different brokers with 99% quality, etc.
After the finished sets are obtained, they are also checked using the excellent program Quant Analyzer, which allows you to comprehensively evaluate the results of the tests on the history.
Advantages Of Quant Analyzer:
- the Monte Carlo method;
- portfolio constructor (correlation Principles)
- detailed analysis of statistical regularities of the obtained data, etc.
We also apply cross validation methods to all of the above. This is quite a time-consuming process, and you can also talk about it for a long time. The information is available online, you can read it, it is very useful.
Forward testing of created expert advisors
After a long and laborious process of creating a ready-made potentially profitable expert Advisor, it is installed on a specially opened real trading account with a broker, for its verification by means of the so-called forward testing. As a result, minimal risks are established, and the trading code is checked in the harsh conditions of the real market. Why harsh, because whatever the testing conditions were in the terminal, they do not reflect what can happen during real trading.
First of all this:
- the change in the spread;
- order execution;
- specifics of the broker’s work with liquidity providers.
After the trading advisor has proved its worth on the forward test, it is allowed to join the portfolio of trading advisors and sets working on a real large trading account.
Example of back tests of our development on quotes with 99% quality in the TDS software package - https://www.myfxbook.com/strategies/zodiaq-v22-eurusd-h12010-2019mm/167813
Darwinex is a high-tech and promising broker. This company has a great future with this approach and development vector. There are actually favorable and comfortable conditions for trading, including for authorized trading systems. I hope the company’s management will continue to develop this project in the same traditions as now.
Darwin account UAE
This trading account contains our best developments that have proven their effectiveness in more than one trading year. Trading is carried out primarily by a portfolio of advisors, while all developments are multi-currency, so they have sets for different financial instruments. In this way, risk diversification and smoothness of direct balance changes are achieved.
- Trading system breakdown levels - the EA’s Strategy embodied in the algorithm is based on determining levels of support and resistance, as well as identifying the mood and direction of the market, with the aim of assessing the strength of the momentum and the likelihood of its continuation. Trading from levels is clearly formalized and supported by fundamental factors. Due to the fact that pending orders are used to enter a position, and the average profit / loss transaction is 38-33 points and the average duration of the transaction is 200 minutes, an absolutely high percentage of compliance with the results of testing and the adviser’s work on a real trading account is achieved.
We will analyze the results of the settings back-test for the EUR / USD currency pair included in the adviser’s set portfolio using the powerful software product QuantAnalyzer. During the testing period from 2010 to 2019, the trading algorithm generated 9504 points of profit, while the maximum drawdown was 347 points. As a result, we will calculate the recovery factor, a key statistical indicator for evaluating the efficiency and effectiveness of the trading algorithm settings, which is determined by the ratio of absolute profit to the maximum drawdown for the entire test period, the Recovery Factor is 27.4, which indicates a high recovery rate after a drawdown.
Special attention should be paid to the number of transactions for the entire period of the backtest, their number was 4351, which in turn indicates the high statistical significance of the test. At the same time, the percentage of wins was 82%, this indicator is achieved by using a preventive filter for converting a position to breakeven.
The trading strategy embedded in the algorithm is based on the classic system of trading from levels and as can be seen from the presented chart, the greatest trading activity is observed closer to the middle of the week, when significant price levels and clusters of pending orders of market participants have already been formed.
Now we turn to the most significant method of evaluating the trading system, the Monte Carlo analysis, which gives us an excellent idea of how reliable our strategy is and how vulnerable it is to changes in market conditions. The historical results of the strategy give us an idea of how the strategy behaved in the past and to some extent was encouraging in the future. However, the market may change in the future, and therefore it is necessary to measure how vulnerable the strategy is to changes such as market cycles, sequence of transactions, etc. We always use Monte Carlo analysis for this purpose. This helps us get a real sense of what we can expect from the strategy. As you can see from the diagram below, the basic principle of Monte Carlo analysis is that many simulations are performed, each time with small changes. The more simulations we run, the more reliable results we get. This analysis allows you to trust the expert Advisor’s back tests with a high probability and indicates high predicative qualities of the trading algorithm.
Now let’s look at a little bit of the theory why this kind of strategy, working on a breakout level with the potential to continue moving, works in financial markets. First, we need to answer the question: do the movements in financial markets differ from absolutely random ones - because if the market is random, then there is absolutely no sense in predicting the system’s performance on it. However, it has long been known that the distribution of increments in financial markets is a distribution with “heavy tails”. The figure shows the densities of two distributions (standard distributions are taken), and the “heavy tails"are highlighted in blue. The arithmetic mean of the sample elements” very well " evaluates the mathematical expectation: it is itself a random variable, with a distribution “very similar” to the normal one with the center in the estimated expectation and a variance that decreases with increasing sample size.
Now note the difference from the normal distribution in the example of the S&P500 index.
How can I decipher this graph? The difference between the distribution of increments of the S&P500, as well as many other markets, from the normal one is that it has more pronounced pointed peaks and “heavy tails”. In this case, the degree of" sharpness “shows how much the increments rotate around the zero point, and the” weight of the tails " - how often and how much force there are strong directional movements. That is, the market is usually in two States-calm and explosive, and less often in transition. It is the “heavy tails” that are traded by momentum systems, including systems that work on breaking levels.
- Countertrend trading strategy – a countertrend multi-currency trading system with a differential approach to trading timing. It works mainly in the American-Asian trading session. The logic of the adviser is extremely simple, but it is absolutely effective and clear. The entry is carried out under low volatile market conditions, when the price takes the “flat” movement, an order is opened at certain triggers, such as exit the current price of the financial instrument abroad built price channel, as well as a number of checks and filters that allow you to achieve the most accurate as the inputs, and the output of the trading position.
As you can see from the results of the test run on the example of the EUR AUD pair, the trading algorithm allows you to generate profits with a high profit factor of 2.33. At the same time, the recovery factor, the most significant indicator for evaluating test results, is equal to 24, which shows how much profit exceeds the maximum drawdown, thus showing the potential of the system for recovery.
Based on the results of the resulting equity change graph, we will build a trend line, as you can see, the Deposit change line precisely wraps around the trend line, while there have been virtually no periods of stagnation in the trading model embedded in the algorithm over the past years. This indicates, first of all, the high predicative properties of this adviser.
Let’s run our mandatory method for checking the ready-made results of trading algorithm settings, the Monte Carlo method, simulating a set of trading conditions, thereby allowing you to exclude the so-called fit for a specific market. As you know, the accuracy of this mathematical method is directly proportional to the number of iterations of calculations under changing conditions. To do this, set the number of simulations in Query Analyzer to 400.
Please note that all trading systems used on the account are stop, for this reason, there may be periodic periods of drawdowns, which you need to wait out and make top-UPS of funds, just during periods of stagnation and drawdown. This way, you can recover faster and earn more profit. These are the main principles of investing!
I also forgot to mention that an important component of success in automated trading is a stable and high-performance VPS with low latency to Darwinex servers. From all that we tried, the best ping was from Chocoping, especially their recommended ping to Darwinex servers below 2 ms.