Veleta.AI backtesting and optimisation is the application of advanced analytics and AI to your trading strategies. The platform will allow you to develop from scratch or start from a standard trading strategy and customise your preferred approach to the asset and the market of your choice.
Veleta.AI allows for defining entry conditions and stay-in-the-trade conditions with a much richer set of possibilities than any other known platform, enabling any trader to quickly express what’s in their brain and back-test their intuition.
Flexible multi-granularity handling
Typically traders use several charts to decide when to enter or when to exit a trade. We have made possible expressing conditions in different granularities at the same time
(daily, 4-hourly, hourly… down to ticks).
Automatic channels & levels extraction
Tracing support and resistance lines, as well as chart channels belongs to the daily trading tasks of any experienced trader. Drawing these lines is not an exact science, non-deterministic and depends very much on traders’ experience.
In other words, there is no deterministic procedure or formula to get this task done. We have trained algorithms to simulate how a trader would draw channels and SR lines, calibrated them with multiple symbols, resolutions and period lengths and validated the result with experienced traders. The result is more than just the plotted levels and channels, rather a mathematical description of the lines, including slope, intercept and tolerance area, so that it can be automatically used in our back-testing and execution engines.
A.
Adjustable multi-granularity Support and Resistance levels extraction.
B.
Machine Learning optimized macro and micro channels extraction and pattern recognition (triangles, wedges, channels, head and shoulders, etc).
C.
Camarilla, Fibonacci, etc.
Premium Indicators
In addition to the usual trend and momentum indicators, present in any platform, we have defined new families of indicators opening up a raft of new possibilities to automate strategies.
A.
Current and recent past events based on candle patterns (e.g. in the last 5 candles one inverted hammer, a bull harami, etc.. up to 40).
B.
Distance to channels and SR lines (e.g. close approaching to the next support line).
C.
Slope extraction for trend and momentum indicators (e.g.: EMA5 slope > 10 or EMA5 slope < -10).
D.
Difference between angles (e.g.: Slope EMA 4 – Slope Mid Bollinger Band 2st).
Bayesian optimization
Usually, traders need to make a decision for the best ST/TP setup when they define a strategy. The final choice is the result of long rounds of manual attempts. Our Bayesian approach automatically finds the best optimal setup with the least iteration.
Self-tuning to safe-guard strategies
Traders usually have a rough idea about their strategies but fail at systematically understanding the influence of conditions based on all possible indicators to eliminate bad trades while preserving the good ones. The Veleta optimization engine mines additional conditions to the strategy definition, for both entry points and dynamic handling, using a long data history to make the results consistent.
Functionality & Methodology
-
Bayesian hyperparameter tuning for dynamic execution.
-
Assemble learning and super-learning based facetted optimization.
-
Enhanced tailor-made Support-Resistance and channels modelling.
-
In-trade prediction for dynamic optimal behavior based on Deep Neural Networks.
-
Optimization based on different risks profiles (from highly-conservative to highly-aggressive).
-
Customization of objective function (cumulative net profit, Sortino rate, Sharpe rate, etc).
-
Time Series multi-granular prediction techniques.
A.
The static handling with or without dynamic SL and TP: with entry points, SL, TP and number of candles. Advance trailing, consisting in adjusting stop losses and take profits dynamically as we get more information in the trade execution.
B.
The conditional exit: specifying the entry point and keeping the trade open while a condition is met.
C.
The Risk/Reward approach: adjusting the ST and TP according to a RR setting (eg.: 3R).
Adjustable execution modes
In traditional backtesting platforms, the user can set take-profits, stop loss, number of candles, etc. We have enabled a much richer set of handling possibilities, and all of them can be combined: