Discover “Honey” series of financial apps: HoneyMaster, HoneyCombs, HoneyBee, HoneyUpdater… Trading has never been so sweet before!
Discover “Honey” series of financial apps: HoneyMaster, HoneyCombs, HoneyBee, HoneyUpdater… Trading has never been so sweet before!
The main idea behind HCombs is that the only profitable and safe way of trading is deep diversification. For deep diversification we need hundreds of robots that utilize up to 100% of the capital(as margin) simultaneously.
HoneyCombs allows to manage thousands of robots. No other trading system on the market can provide this feature. Two years ago we decided: not to create just a few robots and miserably watch them lose all deposit, but to create a risk-free system based on a big population of robots, each making 1+ % of profit every year.
Seems small, but the fact that a decently configured by HoneyMaster system features 200+ robots(even 1000+) makes HoneyCombs able to multiply your deposit every year.
The ability to run hundreds of different algorithms also provides protection from big losses. We configure our robots in a way that they lose maximum 1% of capital per a single deal(spread is included, swaps may increase this value), or less. We took a safe approach: no aggressive and risky strategies, only careful money-management and deep diversification. With our software it is possible to keep drawdowns < 10-20%.
This application connects to the broker with the help of HoneyBee – Expert Advisor written in MQL5. It means that Honey trading system requires installed MetaTrader5. If you wish to use other kind of API – we will be glad to code a custom connector for you, write us an e-mail.
You can change robots’ behavior on the fly: edit amins, set risk level, turn them on/off. For each robot HoneyCombs gathers a detailed statistics: deals made, cumulative profit, DPD(deals per day) values. This allows attenuation of your trading strategy at any time with the full knowledge of its behavior.
To be always informed use “Watch” tool: it will always show current profit/losses level, allowing to keep track of trading events. This small window always stays on-top of other applications, supports transparency in order to be easily integrated in your every-day desktop layout.
Not only robots themselves are customizable: HoneyBee – MT5 connector – also has a lot of input parameters making our trading system even more adaptable. You can set it up with any capital currency, system locales, any MT5 broker.
To provide the best experience we made HoneyCombs as easy to set up as possible: do it once and forget about it. HCombs is already preconfigured for the most typical installations. The first startup and installation will take no more than a few minutes.
Our system automatically performs historical data updates with the feed from MT5 terminal. It carefully audits timestamps, possible data corruptions, never performing a trade if something went wrong. Out main goal is to create the most stable, safe system with minimum drawdowns and with as little risks as possible.
The real power of HoneyCombs comes from HoneyMaster – optimization engine. It is where all the magic happens…
To be able to optimize whole populations of robots we created HoneyMaster – optimization engine. It’s main purpose is to test a number of parameters that form a robot, filter the results, mutate them and start a new cycle with the most profitable population – genetically optimize trading system.
The idea behind HMaster is based on evolutionary principles: only the strongest survive natural selection. Since our trading algorithms take in hundreds of parameters(we call them “amins”) and each may be set in a wide range of values, we simply can not test all possible combinations(10^257+ of variations) even with the power of all supercomputers on Earth.
The workaround for this problem is to test a small number of combinations of input parameters in a genetic loop, while filtering every resulting population. To increase the productivity of this approach we created a number of tools that provide a deep insight into robot’s logic: you can always tell why a certain robot did or did not open a deal, what are the reasons behind its behavior. This allows to combine the power of CPU with the natural power of our brains.
Trading algorithms hard-coded in HoneyMaster are very adaptive: they keep track of average volatility, look for resemblant patterns in the past, determine whether the current wave is statistically finished based on the gathered vector pool, forecast the future wave. All these procedures depend on fractal waves concept. Elliot’s Wave Theory popularized by Robert Prechter has made quite an impact on our work. We didn’t code any of the described by them patterns into our system, instead we implemented the way for the computer to distinguish and categorize patterns itself.
Our pattern recognition system is able to distinguish different frequencies, recursively test patterns for resemblance, starting with the lowest frequency(the biggest wave length). It identifies the pattern as resemblant by comparing vector’s projections on both axes, time and price, understands levels and vector types. This approach enables trading both on high and low frequencies: you can configure robot to open deals a few hours long or a few days/weeks long. We do not support longer trades because there’s simply not enough statistical data to create a profitable system on such big timeframes, and we do not support microscalping: on timeframes lower than 30M there’s nothing but chaotic, unpredictable white noise.
Our optimization engine gathers information on every bar processed. To have a look into robot’s logic you can use “Quick Performance Analysis” tool. Every decision is based on something. You can identify the reasons behind any robot’s behavior. It helps in creation: this way it is possible to donate human logic to the algorithm, even if there’s too much of input parameters to be able to understand how the algorithm really works.
Waves concept and adaptive behavior form unique feature of our engine: robots created with HoneyMaster may work profitably with a number of instruments without any values changes. HMaster automatically scales absolute values in accordance with instrument’s average price, while other parameters’ values are relative: they represent ratios that do not need to be scaled.
This allows to increase analysis data scope: when you test the same robot with 18 instruments on 12 years of data you practically test it on 12*18 = 216 years of data! If the robot is profitable(at least does not lose any money) with the most instruments on 12 years of data, then the probability of its profitability in the future is much higher, compared to “one robot-one instrument” optimization approach.
To learn about all the features, how to install and use our software see documentation.