AI Signal Models
1. KnightML
Introduction:
KnightML offers adaptable trading frequencies and exposures, allowing it to align with traders' preferred styles and risk tolerance levels. This flexibility enables traders to adjust their trading frequency and exposure to suit their unique trading systems.
Signal Interpretation:
Long Signal: When KnightML issues a "long" signal, it indicates an opportunity for traders to open long positions. However, traders are not required to commit their entire position immediately. Future signals may either reinforce or adjust the initial signal’s implication.
Short Signal: When a "short" signal is issued, traders have the option to fully close existing long positions and may choose to open multiple short positions based on the signal's guidance.
This structure ensures that traders can make informed decisions tailored to their strategic approach and market outlook.
Example
Following the initial "long" signal, traders have the flexibility to open positions in line with their risk preferences. Subsequently, traders may encounter additional "long" signals. In live trading scenarios, they can selectively execute trades based on their trading systems, considering one or both of the subsequent signals.
Upon the emergence of a "short" signal, traders have the option to fully close their existing positions and execute new short trades.
2. ViperAI
Introduction
ViperAI is the flagship deep learning-based AI signal/strategy, designed to deliver superior risk-adjusted returns by predicting directional momentum. ViperAI operates in a long/short market-neutral framework and is expected to be profitable in any market condition.
As a complete signal system, ViperAI can be used via Telegram notifications by linking your account with the Trading Signal Bot. Users will receive Long, Short, and Close Position notifications.
Here’s how position entries and exits work in ViperAI:
Entry: When a long/short signal is generated.
Exit: When a close signal is generated.
There are two exit mechanisms:
When the model generates a close signal due to the strengthening of the opposite signal.
When a 2% stop-loss is triggered for BTC and ETH pairs.
A position will close when either 1) or 2) occurs, whichever happens first. In both cases, a "Close" notification will be sent via Telegram.
It is recommended to set a 2% stop-loss for BTC and ETH after every long/short entry. In many cases, an early close signal will trigger before reaching the stop-loss point. Once other trading pairs are added to ViperAI, different stop-loss points will be assigned depending on the pair.
Example
Here is the trading data for the system from 1/1 to 7/31 for BTC and ETH Perpetual Futures, based on an average of 70% exposure/position size.
The signals and strategies for each trading pair available in ViperAI are generated by independent AI models, meaning they do not depend on each other and can be used in combination. Currently, the strategies for BTC and ETH operate at the default frequency. In the future, ViperAI will offer multiple trading frequency options for the same trading pair (some models reduce the number of trades by half while still delivering quality returns) to cater to different trader profiles.
The equity curve below shows the data for ViperAI BTC and ETH pairs used together at 60% exposure each from 1/1 to 7/31:
3. AI Directional Risk
Introduction
The AI Directional Risk model is an advanced tool designed for market regime segmentation, dividing the market into two distinct zones to guide directional risk exposure. The model aims to optimize trading decisions by recommending position adjustments based on current market conditions.
This model is effective only for trend-following strategies. Traders using trend-following techniques can benefit by aligning their positions with the directional insights provided by the model, thereby enhancing their ability to capture sustained market movements and optimize their trading performance.
4. AI MeanRev
Introduction
The AI MeanRev model excels at capturing short-term trends and market volatility, with exceptional performance in range-bound and oscillating markets. With an average trade holding period of about a day, this model is particularly versatile in seizing minor trend opportunities for additional profits. AI MeanRev also frequently identifies trade opportunities at price action turning points, such as local minima and maxima.
Entry into a position is initiated when a long or short signal is generated. Exit from the position occurs upon receiving a close signal. There are two exit mechanisms in place:
A close signal generated by the model due to a strengthening of the opposite signal.
A 2% stop-loss being triggered.
A position will close whenever either of these conditions is met, whichever happens first. In all cases, a "Close" notification will be sent via Telegram.
5. OptimaShort
Introduction
OptimaShort is a prime example of the versatility and robustness of the core infrastructure platform. Custom-tailored for optimal performance in bearish market conditions and built on the strong foundation of ViperAI, OptimaShort has undergone core technology updates, rigorous backtesting, and parameter optimization to effectively detect and capitalize on sharp market downturns.
Positions are entered based on long or short signals and exited upon receiving close signals. There are two primary exit mechanisms:
The model generates a close signal when an opposite signal strengthens significantly.
A 2% stop-loss is triggered for BTC and ETH pairs.
The first condition met will trigger the exit. "Close" notifications will be promptly sent via Telegram, ensuring you are always informed of your position status.
Last updated
