Busting 50 Years of Investing Myths | Rob Arnott

Episode Summary

Executive Summary: Rob Arnott argues that investing advances by busting entrenched myths: many popular finance beliefs are shaped by incentives, data mining, and career risk rather than evidence. He defends fundamental indexing, critiques cap-weighting, smart beta, static equity risk-premium assumptions, and buybacks-as-dividends, while warning that passive flows have made markets more concentrated and less efficient.

Main Topics: Myth-busting as the engine of progress in finance (Priority: 5/5): Arnott frames investing as an area where new ideas repeatedly challenge entrenched beliefs, but warns that new myths can replace old ones. He emphasizes falsifiability, curiosity, and the danger of career incentives preserving bad ideas. Data-first vs. theory-first vs. Bayesian methods (Priority: 5/5): He distinguishes data mining from hypothesis testing and argues that much factor research is overfit. Bayesian thinking should test ideas without endlessly tuning them to historical winners. Fundamental indexing and the logic of reweighting away from price (Priority: 5/5): Arnott explains fundamental indexing as weighting by economic footprint rather than market price, reducing exposure to overpriced stocks and exploiting rebalancing alpha. Smart beta’s evolution and dilution (Priority: 4/5): He says smart beta originally meant rules-based strategies that break the link with price, but the term has become so broad that it is now nearly meaningless. Passive indexing, concentration, and market efficiency (Priority: 5/5): Arnott argues that cap-weighted passive flows increase concentration, distort prices, and create predictable valuation gaps between index members and non-members. Debunking long-term return myths and static risk-premium assumptions (Priority: 5/5): He challenges the idea that long horizons guarantee attractive equity risk premia, arguing that expected returns depend on current yields, growth, and valuation levels. Buybacks vs. dividends (Priority: 4/5): He rejects the notion that buybacks are automatically a stealth dividend, arguing that they mainly change ownership structure and only help shareholders if they improve future growth or are paired with sound capital allocation.

Key Arguments: A lot of factor and academic finance is data mining disguised as discovery; without falsifiability, apparent patterns may not persist. Factors can look successful because they have genuine merit or because capital crowded into them and revalued them upward; source of returns matters. Fundamental indexing works because it rebalances against price and avoids allocating more to stocks simply because they became expensive. Smart beta originally referred to mechanistic strategies that break the link with price, but the label became so overused that it lost meaning. Passive cap-weighted indexing can make markets less efficient by concentrating ownership in a small set of large stocks and forcing index funds to buy high and sell low around reconstitutions. Historical equity excess returns do not imply a fixed future equity risk premium; current yields and valuations matter more than long-run averages. Long-horizon return forecasts are often easier than short-term forecasts because income, growth, and valuation are the main components of returns. Buybacks are not equivalent to dividends for buy-and-hold investors; they only add value if they lead to better future growth or capital allocation. A static assumption that stocks always deserve a 5% premium over bonds can lead institutions to take excessive equity risk at exactly the wrong time. The market’s current concentration in the largest U.S. stocks is historically extreme and, in Arnott’s view, reflects a valuation mismatch rather than pure fundamentals.

Data Points: Top 10 S&P 500 weight: just under 40% - Arnott says the index is more concentrated than at any point in history. Top 10 businesses weight in fundamental index: a little over 20% - He contrasts this with the economic footprint of the largest publicly traded businesses. Low-volatility inflows: billions of dollars a month - He cites rapid asset inflows into low-volatility strategies around 2016. Single low-volatility ETF inflow pace: almost $1 billion a month - Used as an example of crowding and valuation expansion. Fundamental index historical outperformance estimate: about 2.5% per year - He says sales/book-value weighting would have added this over the prior 30 years in early tests. FTSE RAFI live track record vs Russell Value: beat in 16 of 19 years by 2.5% per annum - Arnott cites live performance from 2006 to 2024 with about 2% tracking error. FTSE RAFI worst drawdown: a little over 2% - He highlights the strategy’s relatively low drawdown versus value indexes. S&P 500 index fund ownership: about 25% of each constituent’s market cap - He uses this to explain the scale of forced trading around index changes. Tesla stock move between announcement and inclusion: up 47% - He uses Tesla as a vivid example of index inclusion pressure and crowding. Market-on-close trading in Tesla inclusion: about 20% of Tesla’s market cap - He describes the huge block traded at the closing price on inclusion day. Additions vs deletions performance window: about 15% gap between announcement and effective date - He says additions beat deletions by roughly this amount over the reconstitution window. Deleted stocks subsequent performance: outperform by 7% per annum in first couple years; 28% over five years - He argues deletions often rebound strongly after forced selling. Delete-and-hold compounding result: 73x money from 1991 to 2023 - He describes a historical strategy of buying deletions and rotating annually. S&P 500 buy-and-hold comparison: 25x to 30x money - Used as a benchmark against the deletion strategy. RACWI live performance: +92 bps per annum with 50 bps tracking error - He says the cap-weighted fundamental-selection version has beaten the S&P 500 live since Sept. 2021. S&P 500 concentration myth: never this concentrated, even in the 18th and 19th centuries - He emphasizes the unprecedented scale of concentration today. Historical stock vs bond excess return: about 4.5% to 4.75% per annum over a century - He discusses how people extrapolate from Ibbotson data. Historical real growth in earnings and dividends: 2% - Used in his critique of assuming a fixed equity premium. Yield on stocks at 1950 start: 8% dividend yield - He uses this to explain why postwar stock returns were strong. Yield on stocks at 1999 end: 1% - Used to show how valuation changes drove returns. 10-year TIPS real yield: about 4% in the late 1990s; about 2.25% today - He compares stocks to inflation-linked bonds when discussing the equity risk premium. Current estimated U.S. stock return: about 3% - From Asset Allocation Interactive, his 10-year forecast for U.S. stocks. Current estimated 10-year government bond return: almost 5% - He contrasts this with bonds to argue equities are less attractive today. Shiller P/E peak: 38 or 39 times recently; 44 at dot-com peak - He uses valuation to argue current U.S. stocks resemble prior expensive regimes.

Pivotal Quotes: "science advances one obituary at a time" — Rob Arnott: Explaining why finance progress often comes from busting old myths. "The value side of the U.S. stock market is cheap relative to the market. Historically, just about as cheap as it can get." — Rob Arnott: Describing the current opportunity set outside expensive U.S. growth stocks. "20 years is not a long horizon." — Rob Arnott: Arguing that long-term investors can still make poor relative-return choices even over decades.

Implications: Listeners should be skeptical of popular finance narratives, especially static return assumptions and simplistic factor stories. The episode suggests future alpha may come from valuation-aware, fundamental, and non-cap-weighted approaches rather than crowded mega-cap passive exposure.