PodcastThe Monetary Matters NetworkVictor Haghani

Who Killed The Random Walk? | Victor Haghani on Momentum, Passive Investing, and LTCM

The conversation centers on a model where different investor types — value investors, static allocators, and return‑chasing extrapolators — create persistent volatility, momentum, and boom‑bust dynamics that simple random‑walk models miss. Corporate flows (buybacks, retirement contributions, issuance) and AI‑driven capex are highlighted as mechanical demand drivers that can amplify or reverse market moves. Practically, the guests favor rule‑based, low‑cost strategies combining value and binary momentum, run modestly defensive with a non‑U.S. overweight, while warning that consensus assumptions imply only about 6% long‑term U.S. equity returns and a narrow equity risk premium.

Market VolatilityInvestment StrategyEquities

Preview · 15 more points in the app

Read the Full Summary

Key Takeaways

Corporate AI capital expenditure is a central determinant of S&P 500 earnings over the next few years: continued AI capex with strong returns can sustain aggregate earnings, whereas a capex pullback or poor returns (including markdowns/depreciation) would materially weaken reported earnings — and much of the recent earnings momentum is concentrated in hyperscalers and semiconductor names tied to that AI spending, making earnings strength potentially fragile and sector‑concentrated.

Corporate buybacks (running over $1 trillion per year) together with steady retirement‑plan contributions have mechanically lifted market levels, and the authors' simulated market framework shows this inelastic demand — and potential removal of $1–2 trillion of marginal demand if buybacks fall or issuance rises — can meaningfully move prices.

The episode distinguishes passive index investing from passive asset allocation: market‑cap indexing is defensible, but mechanically fixed asset‑allocation rules (large static allocators) amplify booms, busts, and persistent momentum because they act as predictable, inelastic buyers or sellers within a heterogeneous‑agent market.

Binary, rule‑based momentum (e.g., jump exposure from 50% to 75% when momentum is positive and cut to 25% when negative) functions as a practical, low‑cost risk proxy and historically outperforms slow, continuous return‑chasing, which tends to creep into rallies and exit declines slowly and thereby exacerbates tail risk.

Combining long‑term value measures (cash flows/earnings) with momentum is recommended as the most robust practical framework: the managers model value‑plus‑momentum strategies and implement them in low‑fee, transparent vehicles while keeping a modestly defensive posture (net underweight equities, ~10% overweight fixed income).

TECHNOLOGY

AI Capex — Speakers argue corporate AI capital expenditures will be a central determinant of S&P 500 performance over the next couple of years. Continued AI capex with strong returns can support aggregate earnings; a pullback or poor returns (including markdowns and depreciation) could make reported earnings materially worse. The debate sits alongside disagreement over whether earnings still drive stock performance, with semiconductor examples used to show earnings‑beating firms have driven the biggest gains.

“I think that the fate of the S&P 500 over the next two years it really does depend on is the spending on AI capex going to continue and what…”—

ECONOMY

LTCM Lessons — Victor emphasizes LTCM lessons: real 'skin in the game' matters and running a standalone leveraged capital pool creates concentrated financing risk. He argues relative‑value strategies have fatter tails than delta‑one trades and that diversified pod shops with enforced tight stops and mixed strategies manage crisis risk better.

“I think the biggest lessons are about personal risktaking. The first big lesson has to do with skin in the game.”—

Momentum Rule — Momentum is modeled as a binary, sign‑based rule — for example, increase exposure from 50% to 75% when momentum is positive and cut to 25% when negative. Because signals can flip quickly, momentum can act as a low‑cost, practical risk proxy; Elm uses this transparent rule to adjust equity exposure in ETFs and SMAs.

“A momentum strategy is that when momentum is positive, when today's price is higher than last year's price or the moving average price, when momentum is positive, I overweight my…”—

Return Chasers — Return‑chasing is modeled as a slow, continuous exposure adjustment based on past returns, creeping from 50% to 75% in rallies and exiting slowly in declines. The speakers find binary momentum implementations historically outperform continuous return‑chasing, which lacks fundamental anchoring and can exacerbate tail risk.

“And what the heck? How can that how can momentum be there?…”—

MARKETS

Heterogeneous Agents — The paper models three investor types — long‑horizon value investors, large static allocators that mechanically maintain target mixes, and extrapolators who form expectations from recent returns — and argues this heterogeneity explains persistent volatility, momentum, and booms and busts a random‑walk model cannot. The framework can also quantify price impact (for example, how $100bn of selling moves prices) and shows momentum's prevalence depends on relative capital among investor types.

“Finally, I mean, well, we've modeled a bunch of other investor types, too. But the third type, which is really the key one here, are extrapolators or return chasers…”—

Price Impact — Their simulated market quantifies how large flows (retail or institutional selling, buybacks) move prices by measuring how modeled investor types absorb supply. The results show inelastic demand is a key mechanism for sizable price moves and underpin estimates of how big corporate or retirement flows mechanically affect market levels.

“And you can ask a bunch of different questions like how inelastic is demand as a function of price?”—

Supply Drivers — Corporate buybacks (running over $1tn/year), net retirement‑plan contributions, and changes in corporate supply (IPOs, secondaries, insider sales) act like mechanical buyers or sellers. Buybacks and 401(k)/IRA contributions have lifted markets, while increased issuance or reduced buybacks could remove $1–2tn of marginal demand and weigh on prices; the precise impact depends on the mix of indexers, closet‑indexers, and fundamental investors.

“We have more leverage, you know, through margin loan. You know, you can see retail investors are so bullish on stocks…”—

FORECASTS

Low Equity Returns — Multiple institutions' capital‑markets assumptions and the managers converge around roughly a 6% long‑term expected return for U.S. equities over a 10‑year horizon. Given current 30‑year Treasury yields (~5%), this implies a narrow equity risk premium of about 1%, a situation the guest calls 'voluntary financial repression' that will make wealth accumulation harder.

“When you look at the capital market assumptions of 20 investment management bank investment bank brokerage Vanguard Black Rockck Goldman everybody right the they're saying that the expected long-term return…”—

Valuation Snapshot — Blended operating earnings for the S&P have recently shown about 24% growth and are expected to continue near that rate over the next 12 months, implying a forward P/E near 20. That level is higher than average but below bubble extremes, supporting a cautious stance that expects some mean reversion in elevated earnings growth.

“So, so Victor, the operating earnings of the S&P have been growing about 24% talking about blended operating earnings and they're expected to grow by 24% over the next 12…”—

Guest Introduction

Host introduces Victor Haghani, his books, and Long-Term Capital Management background as the conversation begins on market dynamics and the 'random walk' idea.

Investor Types Modeled

The model includes value investors, static allocators, and extrapolators, with extrapolators shown as the critical driver of trends and booms.

Simulated Market Dynamics

Simulation setup: shocks, trading frictions, and trading behavior produce stochastic volatility, momentum, and boom–bust patterns similar to real markets.

Challenging Random Walk

Discussion frames the paper as a challenge to the classical random-walk model, citing momentum and other persistent anomalies as evidence.

Why Momentum Persists

They explore why momentum remains profitable: behavioral reasons, relative capital sizes, and the uncomfortable logic of 'buy high, sell higher.'

Momentum Versus Fundamentals

Panel contrasts momentum's persistence with fundamental momentum in earnings growth and competition eroding long-term advantages.

Passive Investors Defined

Haghani distinguishes passive stock indexing from passive asset allocation, arguing the latter (static 60/40 allocations) is problematic for markets.

Impact Of Indexing

They debate whether market-cap indexing distorts prices and conclude passive asset allocation, not indexing, likely drives macro-level distortions.

Market Concentration Effects

Discussion on why most stocks appear reasonably priced while a small subset attracts retail frenzy and large valuation disconnects.

Consensus Return Forecasts

Haghani and host examine consensus long-term U.S. equity return assumptions (~6%) and implications for the equity risk premium.

Drivers Of Elevated US Prices

They attribute elevated U.S. stock prices to extrapolators, heavy buybacks, low issuance, margin leverage, and static allocators rather than pure fundamentals.

Earnings Growth And CAPE

Conversation on cyclically adjusted earnings, why short-term high growth shouldn't be extrapolated indefinitely, and the role of buybacks and retained earnings.

AI CapEx & Earnings Sustainability

They debate whether current AI-driven capex and hyperscaler spending can sustain earnings momentum or is likely to mean-revert over time.

Flows, Buybacks, Issuance

Haghani emphasizes supply–demand mechanics (buybacks, IPOs, secondary sales, 401(k) inflows) as key market-level price drivers in the model.

Micro To Macro Transmission

They explore how corporate actions (targeted buybacks) can lift the entire market via indexing, closet indexers, and reallocation effects.

Defining Extrapolators

Haghani defines extrapolators (return chasers) as agents updating expected returns based on recent history, creating reinforcement and trend dynamics.

Return Chasers Vs. Momentum

Detailed comparison: momentum seen as a binary, rule-based strategy that can front-run extrapolator behavior, while return-chasing is smoother and often underperforms.

Value + Momentum Approach

They endorse combining long-term value signals with binary momentum as a robust, research-backed investment approach used at Elm.

Five-Star Screening

Haghani describes Elm's five-star screen for risky assets: clear risk-premium story, estimable premium, diversified systematic exposure, liquidity, and tax/fee efficiency.

Dynamic Asset Allocation

Elm's rules-based dynamic allocation adjusts exposures using valuation-driven expected returns and momentum (as a risk proxy), at very low fees.

Current Tilts & Rationale

Current positioning: modest underweight to U.S. equities, overweight to non-U.S., and a slight fixed-income overweight driven by valuation and momentum signals.

Related Episodes

Important Disclaimers

Not Investment Advice: For education only — not financial, tax or legal advice.

AI-Generated: Made by AI from source transcripts and may contain errors.

No Guarantees: Past performance does not predict future results; investing carries risk.

Our Commentary · Request Removal