Investors need to strengthen their quant approach to benefit from the data revolution

The Future of Asset Management

Quantitative models have become extremely important to the asset management industry. Since markets are increasingly difficult to predict, quantitative driven approaches offer consistent and rational methods to deal with ever-growing data sets and generate performance for investors.

Asset allocators are also becoming more at ease with quantitative strategies, from index replication to smart beta approaches. This is not a revolution but rather an evolution of how investment analysis and decisions can be enhanced through the use of algorithms. Recent surveys have shown that US investors have allocated more capital to systematic strategies than to traditional discretionary solutions. We expect this trend to grow further in Europe into the coming years.

The power of Big Data and Machine Learning

Besides, the exponential growth of data, the advancement in methods of analysis and the emergence of sophisticated computing solutions have led to the machine learning revolution. This has the potential to profoundly change the investment industry and will give an edge to investment managers who are willing to learn and adopt new technologies.

At Napoleon Capital, we believe that analysis of big data forms an important opportunity for systematic investing. By assessing big data using machine learning and artificial intelligence, we look for improving our models and investment signals across asset classes with predictive forecasting of long-term price movements. Our algorithms will become more complex and assess thousands of variables in real time to better inform investment decisions and enhance the diversification and risk/return potential of portfolios.

Napoleon Capital, investors' best ally to leverage on thedata revolution

Open Competition

We have been implementing quantitative models for over 10 years, including in Tier 1 asset management companies. But, in order to generate sustainable performance, we believe in open competition so that our own internal research is continuously challenged by external analysts and researchers, submitting their own trading algorithms. We have partenered with Ecole Polytechnique, which is one of our shareholders, to bring its research power to our project.

Our investment solutions are based on top performing algorithms respecting a set of criteria and generating strong performance, selected through a Darwinian process, regularly rebalanced to ensure best-in-class models at all time.

Partnership with BNPP AM, a Tier-1 asset management company

While we are fully focused to match our clients’ needs with our algorithmic approach, we rely on a Tier-1 asset management company, BNPP AM, to secure the asset management per say. This partnership allows to logically benefit from the global expertise of BNPP Group (custody, fund administration), answering to the most drastic due diligence in terms of fund security.

Investment philosophy : multi-layer best-in-class approach to deliver long lasting performing strategies

Raw strategies
Blend of Blends
High performing
Quant strategies
Raw strategies
In house library of a various range of uncorrelated strategies on a wide variety of liquid assets

Strategy composition (Blend or Blend of blends) relies on open competition, fuelled by external contributors

Strategies composition based
on continuous competition
AI Engine
Optimized investment bricks on a single asset

Darwinian approach with regular rebalancing to ensure best-in-class approach

Decorrelated Blends
AI Engine
Blend of Blends
Multi-assets solutions Customized

Mandates & dedicated funds manage by BNPP AM (partnership)

Raw Strategies

  • Low frequency quantitative strategies on single underlying.
  • Most liquid instruments across global markets, including equities, fixed income, FX and commodities.
  • Internally developped or externally outsourced via competitive challenges.

We maintain an ever growing library of quantitative strategies on single assets. These strategies are long-short or long-only using low to ultra low frequency trading(daily, weekly or monthly). They are developped internally or outsourced via trading challenges.

Research and Development

Our R&D is split into two different pillars: raw quantitative strategies generation and automatic blend building function.

For what concerns the generation of raw quantitative strategies, we leverage on a library of quantitative strategies developped over the years by our R&D team headed by Stéphane Ifrah. We regulary organize data and quant challenges to source alternative raw strategies feeding the library, notably leveraging our link with the Ecole Polytechnique, France's most prestigious graduate school.

We have also developped staking algorithms to build blends from raw strategies and blends of blends from blends. We have been selected for the 2019 Data Challenge organized by the Collège de France and the Ecole Normale Supérieure to optimize these algorithms, with AI techniques.

This website uses cookies to improve your browsing experience, enhance the security of our users, and help us analyse the use of our websites. By clicking or navigating our website, you consent to our use of cookies for this purpose. For more information about our cookies and how to stay in control, visit our Data Privacy Policy