> ## Documentation Index > Fetch the complete documentation index at: https://systematica.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # Systematica > BlockForce Capital Hero Light Hero Dark ## Setting up Install the project. General use cases. Extend Systematica. Code format & principles. Opportunity and limitations. Run tests. ## Investment Philosophy Our interdisciplinary analysis is rigorously driven by economics, econometrics, and econophysics. We recognize that regime changes significantly impact portfolio risk, causing systematic shifts in Beta. We capitalize on diverse investment horizons and liquidity factors to exploit these opportunities. We identify and leverage the underlying dynamics that govern market behavior, regardless of specific market conditions. Our strategies exploit mispricing in volatile markets, incorporating both rational and behavioral finance within adaptive and fractal market structures, while actively managing risk across diverse time horizons. ## Market Hypothesis **Noisy Market Hypothesis**: * Prices do not always reflect the true value of an asset at a given time. (speculators, momentum traders, insiders, or institutions buying or selling for reasons unrelated to the company's fundamentals). * These short- or long-term movements are called “noise” and can obscure the true value of an asset. **Adaptive Market Hypothesis**: * The adaptive market hypothesis (AMH) combines EMH with behavioral finance. * Investors are rational but may overreact during periods of high market volatility, exacerbating price movements. * In AMH, investors are driven by self-interest, make mistakes, and tend to adapt and learn from their errors. **Fractal Market Hypothesis**: * The fractal market hypothesis (FMH) is an extension of EMH and a theory of uncertainty related to different investment horizons. * Increased uncertainty (change) over future horizons leads to sudden volatility. * In reality, market prices follow fractal properties over time, evolving based on available information. * Market Randomness, Investor Time Horizons and varying liquidity across different market segments influence the market over an entire economic cycle. ## Directional vs. Non-Directional Strategies **Statistical Arbitrage**: Risk arbitrage based on the differential between discount and premium per unit of risk. * **Underlying Asset**: Crypto * **Positioning**: Long/Short * **Risks**: Market Neutral (Systemic), Liquidity, Model, Idiosyncratic * **Objectives**: Identify undervalued/overvalued securities, analyze risk factors. * **Advantages**: Low volatility, minimal market correlation, historically stable performance. * **Disadvantages**: Less attractive during bull markets, high transaction costs when rebalancing. **Relative Value Arbitrage**: Detecting temporary pricing disparities between securities that are mathematically and historically cointegrated. * **Underlying Asset**: Crypto/Derivatives * **Positioning**: Long/Short * **Risks**: Market Neutral (Systemic), Liquidity, Model, Idiosyncratic * **Objectives**: Combine multiple strategies: Direct pairs, Indirect pairs * **Advantages**: Purely non-directional strategy, capitalizes on market inefficiencies. * **Disadvantages**: Struggle to compete with specialists, exposure to management and model risks. **Volatility Arbitrage**: Treating volatility as an asset and predicting its increases or decreases using combined positions in call and put options. * **Underlying Asset**: Crypto * **Positioning**: Long/Short volatility * **Risks**: Market Neutral (Systemic), Model * **Objectives**: Use combined positions in call and put options, internal option pricing model * **Advantages**: Flexibility, controlled risk. * **Disadvantages**: Complexity at the portfolio level, limited profit potential when shorting volatility. ## Combining Diverse Alpha Sources * **Asymmetric Distributions Model**: Captures market asymmetries and tail risks in return distributions, adapting quickly to inflection points * **Volatility Models**: Leverages historical/implied volatility and trading volume to predict price movements * **Price Actions**: Analyses indicators and capitalises on persistence of asset performance trends * **Non-Linear Mean Reversion Model**: Models complex dynamics where prices revert to long-term averages in sophisticated patterns * **Momentum & Sentiment**: Combining multiple time horizon, momentum and volatility. * **Volume & Sentiment**: Combining volume-based signals with price movement. Highlights volume surges and volume-price trends to generate an indicator * **And More...** ## Portfolio Construction **Portfolio Selection**: * Integration of computer science, power engineering, and quantitative analysis. * Multi-objective optimization leverages the Pareto front to identify efficient trade-offs between risk and return. * BlockForce Capital’s workflow efficiently reproduces trials from a large pool of cross-validated tests and parallelizes their execution in multiple ways. Strategies can be combined in the same portfolio by stacking their respective arrays along columns, enabling joint simulation and analysis. * Example: Hull Point selects solutions based on the convex hull of the Pareto front or Hypervolume Contribution. **Optimization**: * Walk-forward analysis is best to prevent overfitting historical data. * The API connects with each Suggest or Pruning algorithm. Each worker represents a different trial. * The suggest and pruning algorithms update the storage simultaneously. Each worker makes decisions based on the current state of storage. * The report and return methods make calls directly to the storage. ## Meta-Modeling The process starts by defining the current market regime. A decision-making entity uses different alpha sources to evaluate actions based on state representation, environment and reward. The Agent makes a choice which enhances trading accuracy by correlating indicators and identifying new information. External conditions influence agent decisions by adapting to market changes. The outcome feedback is measured using a binary reward function to encourage maximized returns. ## Dynamic Risk Management Our approach integrates cutting-edge research with financial engineering to identify the most relevant factors and manage portfolio construction efficiently. By constantly innovating and adapting to market regimes, we avoid overcrowded trades whilst maintaining optimal risk-return profiles. **Risk Identification & Monitoring**: * Core risk factors analysis * Continuous monitoring via proprietary non-linear systems * Dynamic exposure adjustments based on market signals **Portfolio Construction**: * Alpha generation through cross-sectional return prediction * Maximise residual returns whilst maintaining beta neutrality * Position sizing optimisation based on evolving conditions **Risk Metrics & Tools**: * Maximum drawdown (MDD) monitoring * Conditional Value at Risk (CVaR) calculations * Real-time stress testing and scenario analysis