TIP688: Long-Term Market Cycles w/ Rob Arnott

Episode Summary

Executive Summary: Rob Arnott explains why cap-weighted indexing embeds overvaluation, how Fundamental Index/RAFI reweights companies by economic footprint to harvest rebalancing alpha, and why today’s equity risk premium is thin to negative. He argues long-term return forecasts are more reliable than short-term ones, sees value near a cycle low, and highlights NIXT as a way to exploit post-deletion mean reversion and index-flow distortions.

Main Topics: Fundamental Index / RAFI design (Priority: 5/5): Arnott describes RAFI as an economy-weighted approach that uses company fundamentals like sales, profits, cash flow, book value, dividends, or employees instead of market cap, aiming to mirror the macroeconomy rather than market popularity. Source of outperformance: rebalancing and value tilt (Priority: 5/5): He argues most of Fundamental Index’s alpha comes from disciplined rebalancing against price moves not supported by fundamentals, with value tilt as a byproduct rather than the primary engine. Equity risk premium and long-term return forecasting (Priority: 5/5): Arnott distinguishes historical excess returns from forward-looking equity risk premium and says current stock returns relative to bonds/cash look modest, even negative after likely mean reversion. Value cycle and market regime (Priority: 4/5): He believes value is near a cycle bottom, with relative valuations extremely stretched versus growth, and expects value to outperform over the medium to long term as mean reversion plays out. NIXT and post-deletion performance (Priority: 4/5): He explains the NIXT strategy, which owns stocks removed from major indexes, based on research showing deletions tend to outperform after being dumped due to forced selling and cheap valuations. Index construction is not fully passive (Priority: 4/5): Arnott challenges the notion that cap-weighted indexes are purely passive, arguing index additions/deletions trigger active trading, crowding, and price pressure that create inefficiencies. NVIDIA, AI narratives, and margin compression risk (Priority: 3/5): He uses NVIDIA as an example of a great business priced for extraordinary growth, warning that future competition and margin compression may cap earnings even if revenue rises.

Key Arguments: Cap-weighted indexes overweight overvalued stocks and underweight undervalued ones because weights rise with price, not business fundamentals. Fundamental weighting reduces pricing distortion by tying portfolio weights to fundamental size, creating a persistent rebalancing alpha. The value tilt in RAFI is a consequence of using fundamentals, but the main return driver is buying low and trimming high as fundamentals and prices diverge. Long-term forecasting is feasible because returns largely come from yield, growth, and valuation change; short-term forecasting is noisy because valuation and growth can swing widely. Current U.S. equity valuations imply low forward returns; historical outperformance over the past decade is not a reliable guide for the next decade. Value may be near a relative bottom because its valuation discount versus growth has become extremely large. Stocks removed from major indexes often outperform after deletion because they are sold mechanically, become cheap, and then mean revert. Index additions often experience price pressure and froth because funds must buy quickly, creating a buy-high/sell-low effect. NVIDIA may continue growing revenues, but competitive response and margin compression could prevent earnings from compounding as much as the market expects. Index investing is still useful, but investors should understand that index reconstitution and market-cap weighting create embedded active bets.

Data Points: Assets in strategies developed by Research Affiliates: $156 billion - Arnott says more than this amount is invested in strategies developed by his firm. Fundamental Index outperformance vs S&P 500: ~2% per year - He says the strategy has exceeded the return of the S&P 500 over the past 20 years by about 2% annually. Magnitude of Nvidia weighting in S&P 500: 6.25% - Used to illustrate how cap weighting allocates more capital to the largest names. Value vs growth valuation spread in 2007: 3x - Arnott says growth stocks were three times as expensive as value in 2007. Value vs growth valuation spread in summer 2020: 9x - He says the valuation gap widened to nine times by mid-2020. Value underperformance vs market: 38% - He cites Russell Value underperforming the market by 38% during the stretch discussed. RAFI vs Russell Value in U.S.: 14 of 17 years - He says RAFI beat the Russell Value Index in 14 of the last 17 years. RAFI vs MSCI ACWI Value globally: 15 of 17 years - He says the global RAFI index outperformed in 15 of 17 years. Cash yield cited: ~5% - Used in his forward equity risk premium arithmetic. Stock dividend yield cited: ~1.25% - He describes this as near the second-lowest in history. Consensus long-term earnings/dividend growth: 18% - He criticizes this expectation as unrealistic given GDP growth conditions. Reasonable nominal growth assumption: ~5% - Arnott’s estimate for long-run earnings/dividend growth. Implied equity return estimate: ~6.25% - He derives this from 1.25% yield plus 5% growth. Forward U.S. stock return forecast: ~3.5% - Research Affiliates’ forecast after allowing for possible valuation mean reversion. Historic benchmark for valuation: 38x Shiller PE - He notes current U.S. market valuation around this level versus the long-term norm. Historic norm / new normal Shiller PE: 18-20x historically; ~23x new normal - Used to estimate valuation drag from mean reversion. Stocks removed from top indexes outperforming: 28% over 5 years - Research on deletions from the largest 500 and 1,000 names showed this average outperformance. NIXT outperformance of winners vs losers: ~18% when it wins; ~5% when it loses - He describes a favorable asymmetry in the deleted-stock strategy. Index addition/deletion performance gap: ~16% - He says additions and deletions diverge by about 16% relative to one another in later research. Index ownership of S&P 500 constituents: ~25% of a company - He estimates S&P 500 index funds own around a quarter of each constituent’s market cap. Tesla price move ahead of inclusion: >40% - He cites the stock rising sharply between rumor/decision and effective inclusion. Tesla run-up from March to inclusion: ~250% - He notes the stock had already surged dramatically on inclusion expectations. Contested-election post-election rally window: 4-6 weeks - He says markets tend to rally in the weeks after contentious elections regardless of outcome. NVIDIA revenue/profit scale: $100 billion annual sales; $60 billion annual profits - Used to emphasize the company’s current scale and profitability. NVIDIA margin assumption discussed: 53% to 30% profit margin - He models future compression in margins as competition rises. NVIDIA chip price cited: $40,000 per chip - He uses this to question whether pricing can remain elevated.

Pivotal Quotes: "The big alpha comes from rebalancing, from concentrating against the market's constantly changing opinions." — Rob Arnott: Explaining why Fundamental Index outperforms beyond simple value exposure. "Reports of value's death have been greatly exaggerated." — Rob Arnott: Referencing his paper arguing that value’s underperformance was driven by cheapening, not failure of the style. "I don't have a problem with 70 times earnings. I have a problem with 40 times sales." — Rob Arnott: Discussing NVIDIA valuation and why revenue multiples can be harder to justify than earnings multiples.

Implications: Investors should question cap-weighted “passivity,” temper return expectations for expensive U.S. assets, and consider value, small cap, international, and deletion-based strategies as potential sources of better long-term returns.