Hedging your PETs – Hedge Fund Market Optimization via Privacy Preserving Technologies

Globally, hedge funds control approximately $5 trillion in assets, with tens of thousands of firms competing in different market sectors. A defining feature of hedge funds is the importance of data to their business. In this post, we will explore how hedge funds are increasingly using Artificial Intelligence (AI) to support data-driven decision making – and why hedge funds that adopt privacy-preserving technologies will gain a market edge. This is a companion to our previous post about how hedge funds are utilizing Generative AI.

About Hedge Funds

Hedge funds generally offer attractive risk-adjusted returns. As a limited partnership of private investors with relatively few regulatory controls, hedge funds usually serve wealthier clients or those with the means to take higher risk investments. Fund managers are well versed in analytics and aim to identify financial opportunities by deeply analyzing companies, sectors and macroeconomic trends.

Today’s hedge funds access huge amounts of data that they use to manage their portfolios and mitigate risk. A hedge fund’s competitive advantage often derives from cross analyzing data to build proprietary models that guide investment strategies and approaches to risk management.

Hedge funds typically have a variety of roles, each contributing to the fund’s overall operation. In different ways, data is central for each of these roles.

• Portfolio Manager: Responsible for developing and implementing data-driven investment strategies and managing the overall portfolio.
• Analyst: Provides recommendations to portfolio managers based on thorough research and analysis of financial data, market trends, and company performance.
• Risk Manager: Based on data analysis, identifies and assesses potential risks associated with the fund’s investments and implements risk management strategies.
• Quantitative Analyst (Quant): Applies mathematical and statistical models to analyze financial data, then develops algorithms and models to identify profitable trading opportunities.

The Rise of Alternative Data

Hedge funds have relied for decades on data sources such as company filings, press releases and analyst reports. In the never-ending search for new insights that can yield a competitive advantage, hedge funds are increasingly looking to alternative data. Alternative data, with a market estimated at $4.7 billion, includes social media, consumer transactions (such as credit and debit cards), Internet of Things (IoT), geolocation data and more.

According to a 2023 survey by Lowenstein Sandler of investment advisers, 48% of hedge fund respondents are currently using alternative data and an additional 33% expect to do so in the next 6-12 months. Of course, the challenge is to extract and analyze relevant data to support decision making.

The Transformative Power of AI

AI and Machine Learning (ML) are transforming the way hedge funds operate. By analyzing vast amounts of data, ML can enable hedge funds to extract value from traditional and non-traditional data alike. The Lowenstein Sandler survey indicates that 72% of hedge funds use AI in conjunction with alternative data. Hedge funds use AI to generate trading ideas, optimize portfolios, streamline data analytics, automate trade execution and more.

Many sophisticated hedge funds have been using ML at scale for over a decade. This is grounded in real-world results. Research by Statista shows that in the long run, hedge funds utilizing AI/ML perform better than other hedge funds. In fact, in recent years a class of hedge funds has emerged that operates based completely on AI and ML. However, the larger trend is for established hedge funds to integrate AI into their businesses. According to a recent BarclayHedge’s Hedge Fund Sentiment Survey, the investment managers leveraging AI range from AI-driven specialists to large quant-driven shops to traditional fundamental investors seeking an edge.

The Tightening of Regulations

Hedge funds may have more latitude than other types of investment firms, but they are increasingly subject to international and local data protection regulations. For example, to the extent that hedge funds use personal data of EU individuals, they are subject to the GDPR. 

Ever since the financial crisis of 2008, greater transparency has been demanded of hedge funds – with regulations and reporting requirements often varying by client type and jurisdiction. Just this year, the U.S. Securities and Exchange Commission (SEC) and Commodity Futures Trading Commission (CFTC) jointly approved new regulations that require hedge funds to confidentially disclose more information about their investment strategies on a quarterly basis.

In this environment, data security is more important than ever. Imagine a hedge fund with a data science team whose members need to collaborate and build on each other’s work in a remote environment. A breach of data protection rules will now have very serious consequences. Hedge funds must ensure the security of their internal systems and procedures, as well as those of service providers with access to sensitive data. 

Perhaps not surprisingly, in the Lowenstein Sandler survey the most cited concern for hedge funds in using alternative data was increasing regulatory scrutiny, with data security/breach issues not far behind. Data and AI are mission critical for hedge funds, but that comes with risks.

Privacy-Preserving AI for Hedge Funds

In the evolving landscape of hedge funds, leveraging data with AI must go hand-in-hand with prioritizing security and privacy. The funds that embrace innovative technology in service of privacy and security – as they have in service of insights and optimization – will come out ahead. 

Despite a widespread perception that ensuring data privacy means compromising on data utility, privacy-enhancing technologies (PETs) support both, For example, Inpher’s XOR platform uses PETs such as Secure Multiparty Computation (MPC) to provide data-in-use encryption that enables building highly accurate ML models without compromising data privacy. The platform supports multiple ML use cases, from collaborative model development to secure federated learning and cross-border data compliance. With XOR, data and models are secret shared across parties during computation, safeguarding against quantum attacks. The result is the ability to leverage data while complying with global privacy and data regulations such as CCPA, GDPR, and more.

To learn more about Inpher’s privacy-preserving solutions, visit our website.