Rob Arnott: Dissecting Smart Beta, Investing in Disruption, and Momentum (EP.157)

Episode Summary

Executive Summary: Rob Arnott argues that cap-weighted indexing systematically overweights overpriced stocks and underweights undervalued ones, while fundamental indexing breaks the link with price to harvest rebalancing alpha. He distinguishes this from factor investing, critiques backtest-driven strategy selection, explains why value’s drawdowns are normal, and says momentum is best used as a trade filter rather than a standalone high-turnover strategy.

Main Topics: Cap-weighting vs. fundamental indexing: Arnott explains that cap-weighted indexes are low-cost and market-tracking, but they mechanically overweight overpriced securities and underweight undervalued ones. Fundamental indexing instead weights by business footprint metrics like sales, cash flow, book value, and dividends to create contrarian rebalancing. Smart beta vs. factor investing: He argues the original concept of smart beta was any strategy that breaks the link with price, not simply tilting toward academic factors. Fundamental indexing has a value tilt and some size/quality effects, but its key source of edge is rebalancing, not pure factor exposure. Backtests, data mining, and strategy selection: Arnott warns that backtests can be engineered to look great and often reward cherry-picked, already-successful factors. He argues investors should put very little weight on historical simulations when deciding whether to adopt a quantitative strategy. Value investing, intangibles, and drawdowns: He says value’s recent underperformance is extreme but not unprecedented, and that book value is distorted by missing internally developed intangibles. Adjusting for intangibles improves value measures, and long drawdowns are normal for factors. Bubbles, disruptive companies, and the 'big market delusion': Arnott defines bubbles as prices requiring implausible growth assumptions and notes that disruptors can be wonderful businesses yet terrible investments when expectations are already extreme. He emphasizes that dominant innovators often become overpriced and later get disrupted themselves. Momentum and factor momentum: He says stock momentum is powerful in theory but difficult in practice due to turnover, trading costs, and rollover risk. Momentum is better used as a filter to avoid trading into free-falling or runaway names. He also finds factor momentum is stronger than sector or stock momentum.

Key Arguments: Cap-weighted indexing is efficient and cheap, but it systematically buys more of what has become expensive and less of what is cheap, creating a structural performance drag. Fundamental indexing reduces that drag by weighting firms by economic footprint rather than market price, forcing disciplined contrarian rebalancing. The main source of return for fundamental indexing is rebalancing alpha, not a simple factor bet, even though it produces a persistent value tilt. Smart beta originally meant strategies that break the link with price; it is broader than classic factor tilts like value, size, quality, or momentum. Backtests are highly vulnerable to data mining and hindsight bias, so investors should not rely on them heavily when evaluating a strategy. Factor drawdowns can be large and prolonged; value’s recent pain is serious, but long factor drawdowns are normal rather than evidence of failure. Book value alone is flawed because it ignores internally developed intangibles; adjusting for intangibles makes value more powerful. Value and profitability can complement each other, but that does not mean intangibles should be ignored in value definitions. Bubbles can be identified in real time when prices require implausible DCF assumptions, but bubble stocks can remain detached from fundamentals for a long time. Disruptive companies often create enormous social value without creating proportional shareholder returns because expectations get priced too aggressively. Momentum is best used carefully and selectively, especially as a trade-blocking signal rather than a high-turnover standalone strategy. Factor momentum appears to explain much of stock and sector momentum, suggesting broader style regime shifts drive much of the effect. Dynamic factor allocation can add value, but only with modest sizing and disciplined guardrails rather than aggressive market timing.

Data Points: Published journal articles: More than 130 - Rob Arnott’s body of published work cited by the hosts Value outperformance vs. cap-weighted value indexes: ~1.5% to 2% annually - Live performance of RAFI strategies over 17 years, depending on geography Emerging markets RAFI outperformance: Above 2% annually - Relative to cap-weighted value indexes in emerging markets Small-cap RAFI outperformance: Above 2% annually - Relative to cap-weighted value indexes in small companies Backtested alpha for equal weighting of eight factors: About 3% - In the 'Alice in Factorland' analysis over 50 years Alpha when selecting best trailing 10-year factors: Cuts alpha in half - Choosing only the three best recent factors lowered returns Alpha when selecting worst trailing 10-year factors: About 6% - Choosing the three worst recent factors doubled alpha versus the equal-weight case Price-to-book/value drawdown: 13.5 years - Fama-French value’s relative performance drawdown since its 2007 peak Price/earnings-based value drawdown: 7 years - Relative performance peak in 2014 Price-to-sales-based value drawdown: 4.5 years - Relative performance peak in 2017 RAFI composite value drawdown: 3.5 years - Fundamental weight vs cap weight value proxy peaked in early 2018 Intangible-adjusted value improvement: 10 to 15 bps - Including intangibles modestly improved RAFI backtests DFA intangibles analysis: About 2x more powerful - Price-to-intangibles-adjusted book value versus ordinary book value Value vs. growth wealth effect using intangibles: 4x wealth after 60 years - Fama-French value versus growth Value vs. growth with intangibles adjustment: 8x wealth after 60 years - Using price-to-intangibles-adjusted book value Value vs. growth spread in COVID aftermath: 12.5 to 1 - Price-to-book spread between growth and value at an extreme point EV sector sales share in 2020: ~2% - Electric vehicles’ share of auto industry sales in the big market delusion example EV sector market cap share in 2020: ~45% - EV makers’ share of auto industry market cap despite tiny revenue share Conventional auto makers’ revenue share: 98% - Compared with EV makers in 2020 Conventional auto makers’ market cap share: 55% - Compared with EV makers in 2020 Tesla assumed growth example: 50% annualized for 10 years - DCF illustration for bubble analysis Tesla size-growth implication: 55x today’s size - If 50% annualized growth persisted for 10 years Amazon historical growth example: 26% annually - Used as a comparison for how extreme Tesla assumptions are Tesla valuation example: $430/share - DCF estimate using optimistic assumptions and BAA bond yields Momentum factor history: Negative return since 1999 - Standard momentum (12-month excluding latest month) cited as not working net since 1999 Momentum crash examples: 2002 and 2009 - Highlighted as major momentum drawdown periods Momentum alpha wiped out: 15 years - 2009 crash erased cumulative alpha from prior momentum success Mutual fund factor momentum finding: Negative correlation - Higher momentum exposure associated with lower returns in the study Dynamic factor allocation range: 5% to 35% - Typical minimum and maximum weights in the five-factor dynamic portfolio Portfolio concentration ceiling: About 3 of 5 factors - Most concentrated possible allocation under their dynamic rules

Pivotal Quotes: "Cap weighting's biggest Achilles heel is that any stock that's overpriced ... is above its fair value weight in the portfolio." — Rob Arnott: Explaining why market-cap weighting structurally overweights expensive stocks and underweights cheap ones "The beauty of this is not getting the weight right... It's having a steady anchor to contra trade against the market's shifting expectations." — Rob Arnott: Describing the purpose of fundamental indexing weights based on sales, cash flow, book value, and dividends "If you're using backtests to choose your strategies, you are self-selecting to be in the ones that have outperformed their own expertise expectations and are poised to disappoint." — Rob Arnott: Warning against choosing quantitative strategies based on historical performance

Implications: Listeners should treat backtests skeptically, expect long factor droughts, and think in terms of valuation, rebalancing, and discipline rather than chasing recent winners. For practitioners, the strongest takeaway is to use factors and momentum modestly, with guardrails, and to avoid confusing great businesses with great investments at any price.