Andrew Ang: Should AI Agents Run Your Asset Allocation? | #653
Andrew Ang discusses how AI agents can improve asset allocation through specialized expertise, productive dissent, and continuous learning, while emphasizing that taxes represent a larger drag on investment returns than most investors realize—with a century of data showing Uncle Sam's cut averaging 350 basis points annually and potentially rising significantly in the future.
Summary
Andrew Ang, founder of Tau Balance and former head of Factor Investing at BlackRock, explores two major themes: the application of AI agent swarms to investment management and the critical but often-overlooked impact of taxes on portfolio returns.
On AI agents and asset allocation, Ang explains that specialized agents can outperform traditional human investment committees through three mechanisms: scale (by having agents specialize in different asset classes and portfolio construction methods rather than one team doing everything), productive dissent (agents critique and vote on each other's proposals, surfacing valuable disagreements without the interpersonal friction that derails human teams), and continuous learning (agents improve their forecasting and reasoning over time). The agents work within strict governance frameworks defined by investment policy statements, and the system includes safeguards against collusion through perfect information transparency and clear scoring methodologies. Ang emphasizes that the role of humans shifts to oversight and exception-based decision-making rather than elimination, and that true innovation requires redesigning workflows around these technologies rather than simply substituting tasks at smaller scale.
On the factor investing front, Ang presents a diversified approach using four broad factor categories: momentum (measured through multiple signals including holdings-based metrics), quality, enhanced value (based on intangible capital and economic value added), and cyclical value (traditional measures like book-to-market). These factors vary over time and can be timed using three models—market similarity, factor momentum, and dispersion—though Ang acknowledges the difficulty of factor timing.
Regarding taxes, Ang's research spanning 1926-2025 reveals that the average federal tax drag on U.S. equity returns is approximately 350 basis points annually across the full century, though recent decades (1990s-2020s) have seen lower rates around 160-170 basis points due to preferential treatment of qualified dividends and capital gains. However, historically high tax regimes in the 1940s-1950s exceeded 500 basis points. Compounded over 30 years, even modest tax drags of 160-170 basis points eliminate approximately one-third of portfolio returns. State and local taxes compound this dramatically—New York City residents face a combined marginal rate of 55.576% when stacking federal, state, and city taxes. Ang argues this makes tax optimization vastly more important than investment alpha and demonstrates through examples how seemingly attractive yield products (like 10-12% private credit yielding 7% after taxes) mislead investors by obscuring tax treatment.
Ang identifies Warren Buffett as an exemplar of tax-aware investing through Berkshire Hathaway's no-dividend policy, which allows tax-deferred compounding. He contends that investors should maximize tax-advantaged accounts (IRAs, 401ks, Roths, trusts), strategically allocate different assets to different tax locations based on optimal post-tax treatment, and recognize that rising government debt (above 100% of GDP with 15% of the federal budget devoted to interest payments) makes tax increases highly likely. His company, Tau Balance, aims to solve the asset location problem by modeling every asset in every account across the entire portfolio holistically.
About this episode
My guest today is Andrew Ang, Founder of Tau Balance and professor at Columbia University. He previously ran factor investing at BlackRock. In today's episode, Andrew explains how a swarm of AI agents can reinvent the investment workflow rather than automating pieces of it. He shows how specialized agents, each running its own asset class or portfolio method, vote on one another's allocation calls. He also argues for diversification across signals, factors and time. To close, Andrew reveals how the quiet drag of taxes erases roughly 36% of a stock investor's returns over 30 years. (0:00) Starts (3:09) The Self-Driving Investment Portfolio with AI (8:08) Evaluating AI-Driven Investment Outputs (14:41) Multi-Level Diversification in Factor Investing (22:08) Factor Weighting, Timing, and Model Diversification (24:27) The Long-Term Drag of Taxes on Equity Returns (33:24) Rising Taxes: Strategies and Asset Location (37:53) Tax Alpha vs. Investment Alpha & Market Impact (42:08) Tax-Aware Bond Strategies ----- Sponsors: Upwork is the world's largest human and AI-powered freelance marketplace to hire top talent—trusted by businesses and professionals worldwide. ----- Follow Meb on X, LinkedIn and YouTube For detailed show notes, click here To learn more about our funds and follow us, subscribe to our mailing list or visit us at cambriainvestments.com ----- Follow The Idea Farm: X | LinkedIn | Instagram | TikTok ----- Interested in sponsoring the show? Email us at [email protected] ----- Past guests include Ed Thorp, Richard Thaler, Jeremy Grantham, Joel Greenblatt, Campbell Harvey, Ivy Zelman, Kathryn Kaminski, Jason Calacanis, Whitney Baker, Aswath Damodaran, Howard Marks, Tom Barton, and many more. ----- Meb's invested in some awesome startups that have passed along discounts to our listeners. Check them out here! ----- Editing and post-production work for this episode was provided by The Podcast Consultant (https://thepodcastconsultant.com).
Key Insights
- Ang argues that AI agents can harness productive dissent more effectively than human teams because agents lack ego investment and interpersonal politics, allowing genuine disagreement to surface without HR involvement or talent departures.
- The research demonstrates that a seemingly modest 160-170 basis point annual tax drag compounds to eliminate approximately one-third of a portfolio's value over 30 years, yet most investors focus 95-99% of their attention on investment alpha rather than tax alpha.
- Ang claims that the current tax environment (1990s-2020s) with tax rates around 160-170 basis points represents historically anomalous lows compared to the century-long average of 350 basis points, suggesting taxes are more likely to rise than fall.
- Private credit yielding 10-12% annually requires investors to account for income tax treatment, effectively reducing post-tax returns to approximately 7% in equivalent equity terms, revealing why 14-15% yields would be needed to match 10% equity returns after taxes.
- Ang proposes that optimal portfolio holdings must differ by tax location because the same equity holding generates different after-tax cash flows depending on whether it sits in a taxable account, Roth IRA, traditional IRA, or trust structure.
- The speaker argues that AI agents should be designed with perfect information transparency between agents and clear, disclosed scoring methodologies to prevent collusion and gaming, contrasting with human committee dynamics where information asymmetries enable political maneuvering.
- Ang contends that factor timing, while extremely difficult, is not impossible and should be attempted using multiple models (market similarity, factor momentum, dispersion) rather than relying on a single timing method.
- The research shows that federal debt-to-GDP ratios now match World War II levels above 100%, with interest payments consuming 15% of the federal budget (exceeding defense spending), creating structural pressure for future tax increases across income, dividend, and capital gains rates.
Topics
Transcript
It sounds small, but you compound that over 30 years and that results in 36% of your portfolio that just disappears. Is there ever a chance when the agents are like kind of whispering behind your back where they're like, hey, by the way, if you vote for me, I'll vote for you. And if you look at this contrasted to like human teams, you know, when you have disagreements, sometimes in extreme cases, HR gets involved or a really talented member leaves the firm. It's sort of hard to harness this sort of disagreement or differences of opinion, but you can really do it easily with agents because they're agents. They kind of don't care. Ends of this, the…
Full transcript available for MurmurCast members
Sign Up to AccessMore from The Meb Faber Show - Better Investing
Cambria Fund Profile: Shareholder Yield Suite
MedFavor from Cambria Investment Management explains the concept of shareholder yield—the total cash companies return to shareholders through dividends and buybacks combined—and introduces five ETFs designed around this strategy. The approach addresses a blind spot in traditional dividend-focused investing by capturing the complete picture of shareholder returns across U.S., foreign developed, and emerging markets.
Dave Iben: The World Is on Sale While America Trades at a Premium | #652
David Iben, co-chief investing officer at Copernic Global Investors, discusses his value investing philosophy focused on finding undervalued assets globally, particularly in commodities, emerging markets, and underappreciated sectors. He argues the U.S. market is at historically expensive valuations similar to 1972, while opportunities exist in overlooked international markets like South Korea, Brazil, and Kazakhstan where quality companies trade at significant discounts.
Roger Ibbotson: Why Isn’t Everyone Rich When the Math Is This Simple? | #651
Roger Ibbotson discusses why exponential wealth from 100+ years of stock market data hasn't made everyone rich, explaining that most people consume rather than reinvest, face taxes and fees, and must endure inevitable bear markets lasting years or decades. He presents historical returns data, bond yield trends, and forecasts through 2050 while addressing topics like market timing, valuation metrics, and the shift from dividends to buybacks.
FT’s Robin Wigglesworth: Why Bonds, Not Stocks, Rule Everything Around Us | #650
Robin Wigglesworth, editor of FT Alphaville, discusses his book 'A Fabulous Debt,' which traces 1,000 years of bond market history and argues that bonds—not stocks—are the true foundation of the financial system and geopolitical power. He explores how bonds have shaped major historical events, from financing the Napoleonic Wars to toppling Liz Truss, and examines modern challenges including leverage, ETFs, and private credit.
Why Gold Stopped Being a Commodity (AllianceBernstein’s Inigo Fraser-Jenkins) | #649
Inigo Fraser-Jenkins from AllianceBernstein argues for US equity exceptionalism while cautioning against dollar exceptionalism, and makes the contrarian case that gold is no longer a commodity but rather money in the current environment. He contends that bonds will no longer serve their historical diversifying role, and AI productivity gains will likely only maintain current growth rates rather than accelerate them.