
Dataset Name: axovision_market
Group: trading_signal
Vendor: Axovision
Data Starts at: 2013-01-04 00:00:00
Asset Class: Equity
Data Update Frequency: day
Axovision US Market
AXOVISION provides daily updated AI trading Signals for the S&P 500
Thoughtfully designed AI Signals covering daily movements of the US equity market (trained on the S&P 500).
You can use them as overlay for your existing strategies - increasing sharpe ratio and decreasing drawdowns - or to implement new optimized strategies.
The AI signals data from AXOVISION can be used directly in trading strategies by using compiled data or analysts can take larger raw data and feed them into their own models.
The insights that AXOVISION AI signals can provide to portfolio management strategies are diverse and well suited for discretionary and quantitative managers.
US Market Signal (Closing Auctions)
- Daily signal, sent before closing auction
- Build robust strategies with low vola, low beta and reduced drawdowns
- Well suited for market timing
Data Input
- Price & Volume Data
- Market & Intermarket Data
- Sentiment Data
- ...
High Conviction
Daily signal, sent before the closing auction- Sophisticated signal, enriched with novel data sources, suitable for implementing leveraged strategies
- Well suited for high return opportunities with equity market near vola
High Conviction+
Daily signal, sent before the closing auction- Sophisticated signal, suitable for implementing leveraged strategies
- Well suited for high return opportunities with equity market near vola
AXOVISION DATA
Investment Philosophy
Innovative products at the nexus of Big Data, Machine Learning & Cloud Computing.
Increase Alpha : Outperformance signals offering substantial advantages to optimise your portfolio
Plug & Play : Directly convert signals into trades, no further calculations needed
AXOVISION provides daily updated AI trading Signals for the US S&P500
- Hypothesis free analysis
- Dynamic adjustments through constant re-training
- Parallel analysis of a multitude of input factors
- Recognition of highly complex fuzzy patterns
- Three different signal types (Signal, High Conviction, High Conviction+)
Raw Data Available
Within our product suite, AXOVISION provides different types of data for their AI signals.
Customers can choose between lightweight compiled signals data and larger raw data.
The raw data includes all scores directly from our model's outcome, while the compiled data is already structured to clients needs.
Methodology
Technology : The analyses performed by AXOVISION is based on several large ensemble model architectures which incorporate machine learning and reinforcement learning methodologies. The architecture is composed of different levels, which can feed their outcome into the next higher-level layer. We provide compiled outcomes and larger raw outcomes of our analysis. Higher-level layers can observe phases of weakness in lower-level layers and take action.
Every layer is optimized according to risk and return; this way we encourage risk adjusted portfolio management strategies by applying the signals outcome. Years of research showed that ensemble models can push performance of optimization tasks within the finance domain. By incorporating tree-based algorithms with deep neural networks and deep reinforcement learning agents, we achieve less risk of overfitting and profit from different capabilities of these models.
Selection Level : This layer analyses the lowest level, directly applied on single assets. Its outcome represents a future valuation for the given assets. Especially on this level ensemble models achieve higher performance.
The price development of stocks is driven by the interaction of various influencing factors. With our machine learning approach, we enable an emotion- and hypothesis-free analysis of a multitude of heterogeneous data sets. This allows the influencing factors to be systematically identified and attractive risk-adjusted returns to be generated.
Allocation Level : Within this layer a higher-level market analysis is applied, that controls direction of selection levels. Among other models, we extensively utilize reinforcement learning models on this higher-level analysis.
To ensure an optimal portfolio, further market data and their interrelationships is included in the portfolio allocation. The intermarket analysis enables the models to use correlations between markets to control the portfolio positioning and investment ratio.
About Axovision
AXOVISION is a quantitative asset management company, mainly focused on machine learning methodologies. We have several years of experience applying AI technology within the portfolio management domain. With our signal as a service data products, we provide highly beneficial market analysis from our machine learning models, that can generate alpha for other portfolio manager. Our signal data involves various analysis of different sources from traditional and alternative data: market analysis, intermarket analysis, sentiment analysis.Our DNA
State of the Art Solutions - Our mission is to harness the potential of AI on financial markets in order to offer our customers modern investment opportunities.
Pioneer work - Questioning the status quo, proactive research and the willigness to break new ground are firmly anchored in our DNA.
Strong Partnerships - We maintain our network to the science and finance domain in order to actively promote innovation in the financial sector with our partners.
Knowledge Transfer - Our research-oriented approach enables us to transfer our knowledge about artificial intelligence into the industry.