RIALTO.AI Prediction Trading Bot
2 things you need to know about prediction trading
As crypto markets are still relatively young, prices of currencies are not yet balanced. However, since the beginning of our trading activities, we have witnessed different states of markets: from euphoric growth to very volatile topping phase, and to the current bear market. During this period, we have engaged a team that has been using a variety of statistical techniques from predictive modeling, machine learning, and data mining that analyse current and historical data to find the edge on the market and profit from market anomalies.
Generally, machine learning involves feeding an algorithm data samples, usually derived from historical prices. The data samples consist of variables called predictors, as well as a target variable, which is the expected outcome. The algorithm learns to use the predictor variables to predict the target variable.
Machine learning offers the number of important advantages over traditional algorithmic procedures. The process can accelerate the search for effective algorithmic trading strategies by automating what is often a manual process. It also increases the number of markets an individual can monitor and respond to. Most importantly, they offer the ability to move from finding associations based on historical data to identifying and adapting to trends as they develop.
There are two things you need to know about the prediction trading:
- It automates processes others are performing manually thus excluding human biases from decision making.
- It allows quick adaptation to changes in real time.
These factors ensure us a competitive advantage, more opportunities, and consequently profits.
RIALTO.AI prediction trading bot structure and testing period performance results
Prediction trading bot is using historical data to set trading rules upon it and profitability of strategies depends on the recognition of future opportunities and the elimination of past errors. Prior to the setting of rules, we had to go through a lengthy process of historical data analysis that gave us an insight into the crypto market behaviour.
However, the difference between the prediction trading bot and AI bot is that technically, prediction trading bot presents a stage in the development of AI bot. The rules of the deployed backtested strategies in prediction trading bot are based on human observation of markets and are being quantified. While in the case of the AI bot, the data is given and machine learning methods are used upon which the model makes self-learned decisions when and whether to make a trade.
Our prediction trading bot runs five strategies which include two binary strategies with a constant market exposure either long or short and strategies based on momentum and mean reversion which can profit in a rising market.
Currently, our strategies are using four different window lengths, and we plan to test additional shorter and longer ones in the future.
Since the beginning of our trading on December 1st, 2017, our trading bot has performed 1.19% cumulative return with a maximum drawdown of 18%. It is important to notice that until the end of May, we have been testing long-only strategies.
Future plans
In the second half of August, we will begin one of the more demanding and time-consuming stages of the development, the evaluation of the machine learning models. Nevertheless, we will continue with development, back-testing, and deployment of new strategies. Insights into the further performance results, you will be able to get in the future quarterly performance reports.
The first official trading period for the prediction trading bot starts on Monday, July 9, at 00:00. Until then, the users will have time to activate the bot in their accounts.
To stay updated, do not forget to subscribe to our newsletter and follow our social media channels!
Newsletter | Twitter | Medium | Facebook | Linkedin| Telegram
