LeibnizGroup

The Leibniz DISLOCATION program consists of a range of machine learning and AI driven focused investment strategies. The strategies are diversified, with about 1000 live positions on average. Strategies can have up to 2000 orders per day and consume 1- and 5-minute bars. The portfolio is built via a combination of closed-form, stochastically robust, and meta-heuristic optimization techniques applied at various stages of the portfolio formation process, such as synthetic asset formation, industry group, and overall portfolio optimization.

The Leibniz EMOTION program is a fully systematic investment offering with track records dating back to 2011 achieved at some of the most prominent hedge fund firms in the industry. EMOTION has proven to be resilient against drawdowns. The strategy exploits both investor underreaction in the medium-term and investor overreaction in the short-term. The strategy is low frequency and capacity constrained.

The Systematic Intelligence Program (SI) is a fully systematic and globally diversified investment strategy. The strategy deploys assets globally by investing in 80 different equity, currency, commodity, interest rate and volatility markets via highly liquid instruments to maximize diversification and limit volatility.

Average holding periods range from 3-40 days per sub-strategy. Risk Exposures are adjusted daily and are reduced within 1-2 days in trend reversals. SI consists of 3 Sub-Strategies which are dynamically weighted, allowing complete de-allocation from an asset class and avoiding over fitting and a pre-established bias.

The SWARM Program (SWARM) is a fully systematic investment strategy based entirely on machine learning approaches. The strategy screens 400 assets globally and then builds positions via 25+ FX spot pairs on average. Various sub-strategies diversify trading approaches and contribute their inputs to their respective regularized neural networks.

The execution of the strategy is systematic, and all facets of the models, risk management and trade allocation are fully automated. Average holding periods approx. 1 day with the shortest average holding periods for pairs being less than 4 hours.

The capacity constraint REGIMENT investment program is a fully automated volatility harvesting strategy which aims to profit from idiosyncrasies in medium, short and ultra-short volatility environments via directional and mean-reverting approaches in FX markets. REGIMENT aims to algorithmically emulate the inverse behavior of loss-making FX market participants and employs numerous disciplines as well as quantitative analysis, order routing algorithms, low latency tech and liquidity management to capture its trading edge. REGIMENT consists of a diverse range of trading and hedging strategies which detect market-, correlation- and volatility-regimes, to adjust their trading and interconnections accordingly. The master risk layer executes 1000s of concurrent computations per second.

Leibniz COMMO is a pure commodities focused investment strategy. The algorithm detects trade-able patterns across 19 commodities markets and trades them in the most suitable sub-strategy based on daily observations.

The strategy makes use of entry and exit algorithms to detect regimes and employs an active drawdown limiting system. The strategy rests on the GRAB strategy infrastructure which was developed in 2013 and ongoing research has lead to further improvement implementations.

GRAB utilises a swing and trend focused trading methodology with a focus on elliptical cycles in financial markets. The algorithm detects tradeable patterns across 30 markets and trades them in the most suitable sub-strategy based on daily observations. The strategy makes use of entry and exit algorithms to detect regimes and employs an active drawdown limiting system.

The GRAB strategy was developed in 2013 and ongoing research has lead to further improvement implementations as recent as March 2019, leading to GRAB+

The EVEREST Program is a fully systematic, diversified trading strategy which was designed to achieve superior risk-adjusted returns regardless of overall market direction. It is live since March 2016 and has achieved a double digit annualized net return with a controlled level of volatility.

EVEREST is active in foreign exchange and precious metals markets employing a blend of complimentary market timing methodologies in order to diversify trading risk.

The BLUESTAR investment program is run by a former BlueCrest Executive and employs a diverse and growing set of low frequency multi-model, multi-indicator and multi-timescale sub-strategies. BLUESTAR strategies exploit market effects with systematic macro, fundamental and technical approaches in Futures and Spot FX markets. The strategies systematically interpret and position around short- & long-term interest rates as well as commodities markets. Short-term liquidity provisions, medium-term technical factors and longer-term fundamental effects define how the strategy adapts to changing market conditions continuously. The foundations of some of the strategies go back more than 20 years.

AUXENTA is a systematic all-weather strategy with the ability to sustain shocks similar to auxentic materials. It consists of complementary medium frequency sub-strategies which digest data every millisecond.

The strategy does not solely rely on trend following components but rather reacts to short term market behavior via its adaptive hierarchical machine learning algorithms and risk models.

AUXENTA development started in 2016 with the initial strategy going live in June 2019.

ROCKET is a systematic high-to-medium-frequency algorithmic program that trades US futures on the main market indices. It tracks the instant changes of the available liquidity and reaction to these changes, to take advantage of either an excess of liquidity supplied by buyers or liquidity excess, supplied by sellers.

ROCKET focuses on market microstructure and liquidity changes caused by either recurring or unexpected liquidity events and utilizes holding periods ranging from 1 second to tens of minutes. The mathematical foundation of the ROCKET infrastructure consists of a series of specially developed data transformation mathematical algorithms, that take into account the non-ergodic character of the traded instrument dynamics (“memory”).

Leibniz P-REACTION rigorously applies a data-driven framework using machine learning methods and natural language processing to predict how investor behavior impacts asset prices and leads to short-term mispricing surrounding earnings related news and events in a medium-frequency approach. P-REACTION digests and pre-acts as as well as reacts to many thousand events via its news-driven strategies in a fully systematic manner. Many years of development work and attention to detail now allow the strategy a reliable approach and fast reaction time to market moving news by exploiting behavioral biases and market inefficiencies.