The Coming Revolution in Pension Investment Technology
Artificial general intelligence (AGI) has become a focal point on Wall Street, reminiscent of a modern gold rush. Many are already leveraging large language models for data collation, report drafting, and executing complex inferential analytics that would traditionally require days of human effort. “Sandboxes” are erupting, designed to process diverse data forms swiftly. The stock market fervently embraces companies with credible AGI or quantum computing promise, and industry leaders are contemplating “superintelligent” systems capable of surpassing human intelligence in the coming years.
The trajectory for this innovation is clear, driven by market forces rather than mystical foresight. This technological evolution replaces human labor as the main source of value and profits. The critical unknown is whether these systems will match or outstrip human abilities, transforming from cost-efficient research tools to decision-making platforms. Institutional investors, particularly pension trustees, stand to benefit at scale from these developments after initial boutique successes serving affluent clients.
AGI systems are poised to first enhance, then potentially replace, human analysts, portfolio managers, and consultants. Office workers engaged in routine tasks might find their roles at risk within a few years as these technologies advance beyond the “garbage-in/garbage-out” phase. It’s likely that institutional portfolio design and management will undergo significant transformations.
Historically, public pension portfolio management has experienced three revolutions. Initially, the prudent-person rule and modern portfolio theory, alongside capital market theory and Fama’s efficient markets thesis, liberated pension funds from lower-return bond portfolios. The second revolution saw index fund management rise to prominence as large pension funds realized the futility of maneuvering vast portfolios to surpass market averages in increasingly efficient markets. The third wave introduced alternative assets, with endowment fund strategies proving the efficacy of private equity investments. This shift accepted less liquidity for better long-term returns, leading to growth in real estate, hedge funds, commodities, and private credit as asset classes.
Looking forward, we can anticipate a few large firms developing powerful investment-tech platforms employing AGI and eventually AAI (artificial applied intelligence) across various tasks and workflows. Financial industry use cases are emerging, marked by trends like industry consolidation and the acquisition of private market data stores by top investment management houses. Relaxed federal deregulation and antitrust enforcement could empower big players with even more market clout. While boutiques might offer charm and innovation, size remains crucial when capital, expertise, and technology interact with pension fund assets.
The next phase of this AGI revolution is particularly intriguing for private and alternative assets. AGI and AAI have the potential to replace the information advantages of non-public partnerships such as hedge funds and private equity. Proprietary insights may not remain exclusive forever, with AAI tools potentially narrowing performance gaps enjoyed by privileged managers.
Despite this, the wealthiest and savviest individuals will likely find ways to secure exclusive deals. However, the question arises: why rely on middlemen fund managers when direct access to pension fund capital via smart AAI-driven networks could lower costs for entrepreneurs, CEOs, borrowers, and developers? Such platforms could democratize access to institutional capital, reducing the load of hefty management and performance fees currently accepted by middlemen.
To unlock this, pension funds must navigate public-sector freedom of information (FOI) laws that hinder competitive advantage. The information edge currently enjoyed by portfolio intermediaries will erode as efficient systems process data, model scenarios, assess risks, and design portfolios without heavy reliance on high-fee managers. The inception of this shift is evident in the private credit market, with leading investment consulting firms assembling their private equity portfolios.
Pension consulting companies, protected from FOI laws, are well-positioned as industry disruptors. A consulting industry motivated by flat-fee retainers could displace high-fee managers, bringing investment fees closer to eliminating lopsided information advantages. Public pension funds need to anticipate these trends, staging the industry towards a secure, collaborative platform for confidential information exchange to reduce costs.
Pension funds could start by creating an AI-supported proprietary database of portfolio holdings and performance, gradually incorporating secured private data. This database, potentially managed by a public benefit company, could generate revenue through license fees, benefiting participating pension systems. An industry consortium could emerge, functioning like an outsourced deputy chief investment officer, managing alternative investments at reduced costs.
Initially contracting AAI leaders, the consortium could later integrate these functions internally, attracting forward-thinking professionals to compete with high-fee traditional models. This shift would bring lower-cost capital for entrepreneurs and potentially halve pension fund fees, resulting in better returns from alternative asset classes. The consortium could also carry an independent audit function to ensure AI systems’ integrity and accuracy, addressing potential gaps in private sector transparency.
A private asset marketplace, shielded from FOI laws, is taking form. A two-way database maintained by a consortium can provide confidential access to consultants and pension funds directly. Without this intermediary, plans remain dependent on costly middlemen providing FOI buffers. This initiative presents an exciting opportunity for industry visionaries and pioneering pension attorneys, even if AGI and AAI do not fully meet human intelligence expectations.
