Conference Presentation, Panel
The Future of Finance
Milken InstituteIzabella Kaminska, Geoffrey Duncombe, Antony Jenkins, Brian Oliver, Omar Selim, Anthony Todd
- Future Industry Leadership: Omar Selin predicts that the largest financial institutions in 10 years will likely be Big Tech companies (e.g., Google, Apple, Facebook) rather than traditional banks like JPMorgan or Barclays, driven by data and technology capabilities.
- Three Disruptive Forces: Selin identifies the future of finance as being shaped by:
- A unique macroeconomic environment of rising interest rates and negative mark-to-market values in fixed income, forcing capital flows toward equities.
- The ability to process 90% of the world's data (less than one year old) using AI to correlate non-financial ESG data with stock performance.
- The "lifestyle feeling" of finance, where investors, particularly Millennials, demand alignment between their capital and personal values.
- S-Ray Tool Specifications: Arabesque Partners' "S-Ray" tool utilizes machine learning to analyze over 200 ESG metrics across 50,000 news sources in 15 languages, generating scores in three dimensions:
- UN Global Compact Score (human rights, labor law, environment, anti-corruption).
- Sector-specific ESG Score (financial materiality).
- Business Involvement Score (screening for specific ethical concerns).
- Investment Thesis: Selin asserts that sustainable finance generates alpha and acts as a "different engine" for the industry, similar to the shift to electric vehicles, rather than sacrificing performance for ethics.
- Computing Power Democratization: Anthony Todd notes that the convergence of affordable high-performance computing, AI, and Distributed Ledger Technology (DLT) enables the radical re-architecting of financial services by removing physical components and reducing costs.
- Predicted Market Contraction: Todd predicted two years ago that 50% of financial services personnel and 50% of bank branches would close due to technology; he now views this prediction as conservative given the accelerating rate of change.
- Elimination of Intermediaries: Todd argues that DLT and blockchain can eliminate central counterparties and the associated friction and costs currently embedded in payment and trading systems (e.g., CHAPS fees).
- Cryptocurrency Skepticism and Outlook: Selin acknowledges Bitcoin's high energy consumption but dismisses it as a solvable problem, comparing the current state of crypto to the internet in 1990; however, he highlights the lack of regulation as a more profound issue than energy usage.
- Risk of Overfitting: Jeffrey Duncan warns that the combination of massive data volumes and machine learning introduces the risk of "overfitting," where models find spurious correlations in historical noise rather than structural relationships.
- Systemic Risk Scenarios: Duncan outlines three forms of model risk:
- Models based on historical patterns that no longer hold.
- Models based on purely spurious correlations.
- Models based on correlations traded by too many market participants, creating a false sense of safety before a potential "disorderly regime change."
- Mitigation Strategies: Two Sigma addresses these risks by:
- Hiring talent with deep domain expertise and the ability to contextualize data (e.g., understanding when a model might fail).
- Imposing human constraints and intuition into algorithmic models.
- Retaining significant headcount alongside increasing compute power to ensure human oversight.
- Market Structure Evolution: Brian Oliver (Citadel Securities) notes that since 2014, regulatory changes have lowered barriers to entry, allowing new entrants to compete on liquidity quality rather than balance sheet or research.
- Hybrid Service Model: While electronic market-making has expanded, Oliver confirms a continued role for "high-touch" services in fixed income to facilitate large risk transfers and structured transactions that cannot be fully automated.
- Shift in Client Demand: Anthony Todd observes that institutional investors are moving away from traditional correlated returns (driven by the last 35 years of declining rates) and now demand:
- Performance uncorrelated with stocks and bonds.
- Steady, reliable return streams in a multi-strategy context.
- Liquidity Mismatch Risks: Todd identifies two primary risks in the modern quantitative sector:
- The proliferation of low-cost, static, rules-based ETFs and risk-premium strategies that lack active research and are prone to performance erosion.
- Liquidity mismatches where ETFs offer daily liquidity but hold underlying illiquid assets, creating a risk of "locked doors" during market stress.
- Data Ownership and GDPR: The panel highlights the critical battle over data ownership, noting that while tech companies view data as an asset, financial institutions often view it as a liability; new regulations like GDPR will force a redefinition of who owns data and how it is monetized.
- Credit Scoring Implications: Duncan argues that advanced data analysis will reduce false positives in credit scoring, potentially lowering costs for good credit risks while restricting access for high-risk individuals, leading to greater societal stratification.
- Unintended Consequences: Panelists acknowledge that increased financial transparency and efficiency can lead to negative second-order effects, including the potential blacklisting of certain communities or the facilitation of crime in less regulated digital spaces.
- Inclusivity of ESG: Selin defends the relevance of sustainable finance for all wealth levels, arguing that transparency allows individuals to align their values with their money regardless of the amount, and that ESG performance correlates with better financial outcomes, not just ethical ones.
- Productivity Trends: Brian Oliver cites internal data from fixed income markets showing that technology has significantly increased productivity, allowing smaller teams to deliver higher-quality liquidity services compared to traditional larger incumbent organizations.
- Universal Basic Income (UBI): Anthony Todd suggests that as automation transforms the labor market, policy levers including education, infrastructure, and potentially UBI may become necessary to manage inequality and social mobility.