The short version
- Five papers — published between 1952 and 2008 — quietly determine how well-constructed long-term portfolios actually behave. Most retail content references them by name without explaining what they actually say.
- Markowitz, Fama-French, Jegadeesh-Titman, Sharpe, and Daryanani together cover diversification, factor exposure, momentum, the math of fees, and how to rebalance without overtrading. That is a near-complete operating manual.
- Bottom line: if you can summarize each paper in two sentences, you have a sturdier framework than the majority of investors holding the same ETFs you do.
The 10-year Treasury sits at 4.39% and the VIX at 16.99 (FRED, asof 2026-05-01) — a regime where neither the bond bid nor the volatility premium is doing much heavy lifting on its own. In quiet markets, structural decisions matter more than tactical ones. Which is a useful moment to revisit the academic work that defines what "structural" actually means for a long-horizon investor.
Five papers do most of the load-bearing. Each one resolves a question that, unanswered, leaves a portfolio quietly miscalibrated. The point of this piece is not to summarize the math — the originals are public — but to clarify which decision each paper actually informs, and where the literature genuinely has limits.
Why these five, and not others
Quantitative finance has thousands of papers. The five below get top billing because each one introduces a primitive that the rest of the literature builds on. Skip Markowitz and modern portfolio construction stops making sense. Skip Fama-French and the entire smart-beta industry collapses into marketing. Skip Sharpe (1991) and you misunderstand why fees compound the way they do.
The selection deliberately leans toward papers a buy-and-hold investor can actually act on. There is interesting work on volatility surfaces, options pricing, and high-frequency microstructure — none of it changes how a 30-year ETF allocation should be set up. Readers thinking about how diversification translates into single-country versus global allocations may also find Single-Country ETFs vs Global Diversification useful as a companion application.
1. Markowitz (1952) — Portfolio Selection
Harry Markowitz's "Portfolio Selection" (Journal of Finance, March 1952) introduced the idea that an investor should care about the joint behavior of holdings, not the holdings individually. The mathematical core — minimize variance subject to a target expected return — defined modern portfolio construction.
What the paper actually tells the long-term investor: if two assets have the same expected return but imperfect correlation, the combination has lower variance than either one alone. That is the only true free lunch in investing. The corollary that gets less attention: when correlations rise toward 1 (as they often do in stress), the diversification benefit shrinks — exactly when it is needed most.
Mean-variance optimization itself is fragile in practice. Small errors in expected-return estimates produce wildly concentrated portfolios. This is why most institutional implementations use shrinkage estimators or simpler heuristics (equal risk contribution, fixed weights). The intuition is unimpeachable; the literal optimization is not.
2. Fama & French (1993) — Common risk factors in returns on stocks and bonds
Eugene Fama and Kenneth French's three-factor model (Journal of Financial Economics, 1993) extended the single-factor CAPM by showing that size (small-cap minus large-cap, "SMB") and value (high book-to-market minus low, "HML") explain a substantial share of the variation in stock returns that beta alone leaves unexplained.
The reader-facing implication: a portfolio's long-term return is largely determined by its factor exposures, not by stock-picking skill. A "growth fund" with a momentum tilt will behave very differently from a "value fund" with a quality screen, even if both hold 200 large-caps. This is what the smart-beta industry sells, with mixed execution quality.
The honest caveat the academic literature itself raises: the original SMB and HML premia have shrunk meaningfully in the post-2003 sample. Whether this is decay (the factor was an arbitrage opportunity that publication closed) or regime (one period being unusual) is genuinely unsettled. Either way, the framework — decompose returns into factor exposures — is more valuable than any specific factor's expected premium.
3. Jegadeesh & Titman (1993) — Returns to Buying Winners and Selling Losers
Narasimhan Jegadeesh and Sheridan Titman's paper (Journal of Finance, 1993) documented that stocks with strong recent returns over 3-12 month windows tend to outperform in the following 3-12 months. Momentum has since become arguably the most replicated anomaly in equity markets, surviving across decades, geographies, and asset classes.
For a buy-and-hold investor, the relevance is indirect. Momentum is real but expensive to harvest — it requires high turnover, which means transaction costs and short-term capital gains in taxable accounts. Most retail "momentum ETFs" capture only a fraction of the academic premium because of these frictions. Carhart (1997) added momentum as a fourth factor, and most modern factor models include it.
The non-obvious takeaway: momentum's existence is a useful counterweight to a strict efficient-markets view. If markets price information instantly, persistent momentum should not exist. That it does — for over thirty years across dozens of out-of-sample tests — argues for a more behavioral view of price formation than the textbook EMH suggests.
The academic premium for a factor is the gross number; the version a retail investor actually receives is net of expense ratio, bid-ask spread, tracking error, and tax friction. The gap between the two can swallow the entire premium.
4. Sharpe (1991) — The Arithmetic of Active Management
William Sharpe's two-page Financial Analysts Journal article from 1991 contains the cleanest argument in finance. Before fees, the average actively managed dollar must, by accounting identity, earn the market return — because the active and passive segments together hold the entire market. After fees, the average actively managed dollar must underperform passive by the difference in expense ratios.
This is not an empirical claim. It is arithmetic. The paper takes about ten minutes to read and decides the active-versus-passive debate for the median investor.
The downstream implication is the case for low-cost broad index funds, which is now so widely accepted that it is easy to forget how recently it was contested. The same arithmetic also explains why high-fee active products in efficient asset classes (US large-cap equity, intermediate Treasuries) have systematically failed in long-term outcomes data, while a small minority of active managers in less efficient corners (small-cap value, frontier markets) sometimes justify their cost. The arithmetic does not prohibit alpha — it just guarantees that the average dollar chasing it pays a tax for the attempt.
5. Daryanani (2008) — Opportunistic Rebalancing
Gobind Daryanani's "Opportunistic Rebalancing" (Journal of Financial Planning, January 2008) is the practitioner paper on the list. Daryanani tested rebalancing approaches against a simple buy-and-hold portfolio across multiple decades and showed that monitoring drift on each asset and rebalancing when allocations breach a tolerance band — typically around ±20% relative drift — added measurable utility versus calendar rebalancing.
Vanguard's 2024 update broadly confirmed this: bands rather than calendar dates, applied to a small number of asset classes, capture most of the available benefit at low turnover. The numbers most commonly cited are ±15% for core holdings and ±25% for satellite positions, checked weekly or monthly.
The reason this paper makes the top five: every other paper assumes a portfolio that is somehow maintained at its target weights. Daryanani is the only one that actually addresses how to do that without overtrading — which, in practice, is where many otherwise-well-constructed portfolios bleed return.
What this body of literature can and can't tell you
Each of these papers describes the average behavior of a large group of securities over a long sample. None of them tells you what the next five years will look like, what the right allocation is for a specific household, or whether a given factor's premium will repeat in your investing lifetime.
What they do, collectively, is establish the priors a thoughtful long-term investor should hold: that diversification is the dominant free lunch, that factor exposures explain more than stock selection, that fees compound mercilessly against the average active dollar, that momentum is real but expensive, and that mechanical rebalancing rules outperform improvised ones.
The literature on tactical timing — including the moving-average switching rules popularized by Faber (2007) and others — shows backtested gains, but the live-versus-backtest performance gap is uncomfortably large for any strategy that requires getting in and out of the market. A buy-and-hold investor reading this list is well-served by the first four papers and by Daryanani; the timing literature deserves more skepticism than its backtests suggest. For a closer look at how a downturn actually compounds — or does not — over a 10-year window, see If the Nasdaq 100 Corrects 20%, Will Long-Term Investors Lose Money?.
Where each paper actually shows up in a portfolio decision
| Paper | Decision it informs |
|---|---|
| Markowitz (1952) | How many asset classes; how correlations behave in stress |
| Fama-French (1993) | Whether to add a small-cap or value tilt to a market-cap core |
| Jegadeesh-Titman (1993) | Whether a momentum sleeve justifies its turnover and tax cost |
| Sharpe (1991) | Whether an active fund's fee is defensible vs the passive equivalent |
| Daryanani (2008) | How wide to set rebalancing bands; how often to check |
Frequently asked questions
Do I need to read the actual papers, or are summaries enough?
Markowitz, Sharpe (1991), and Daryanani are short and accessible — under 20 pages each, with minimal math beyond basic algebra. Fama-French (1993) and Jegadeesh-Titman (1993) are more technical but readable with some patience. For a long-term investor, the originals are worth the few hours; they vaccinate against secondhand misreadings.
Is the Fama-French value premium dead?
Honestly unsettled. The premium has been substantially smaller in the 2003-2024 sample than in 1927-2002. Whether that reflects publication-induced decay, a regime shift toward intangible-heavy companies that book value mismeasures, or just one rough patch in a noisy long-run series remains an open question. Reasonable researchers disagree, which is the most honest summary available.
Why isn't Black-Scholes on this list?
Black-Scholes (1973) is foundational for options pricing and derivatives — a different problem set. For a long-term investor holding equity and bond ETFs, it informs almost no decision. If a covered-call ETF is ever considered, Black-Scholes becomes relevant; until then, it is not.
What about behavioral finance — Kahneman, Thaler?
Worth reading separately. Kahneman & Tversky's prospect theory (1979) and Thaler's work on mental accounting are arguably more useful for not sabotaging yourself than any of the five papers above. The behavioral literature explains why investors behave irrationally; the quant papers describe what disciplined behavior looks like. Both halves matter.
Does any of this apply to a Korean-resident investor holding US ETFs?
Yes — the underlying mathematics is currency-agnostic. The implementation friction differs: KRW-USD currency exposure, withholding tax on US dividends, and KRW-listed equivalents (TIGER, KODEX) sometimes track different indices than their US counterparts. The framework holds; the execution details require local adjustment.
Key takeaways
- Five papers — Markowitz (1952), Fama-French (1993), Jegadeesh-Titman (1993), Sharpe (1991), Daryanani (2008) — cover the structural decisions a long-term portfolio actually requires.
- The papers establish priors, not predictions. They describe average behavior over decades, not what will happen next quarter.
- Sharpe's arithmetic alone — that the average active dollar must underperform passive by the fee difference — settles most debates about expense ratios.
- Momentum is real but expensive; factor premia have decayed since publication; mean-variance optimization is fragile in practice. The literature has known limits, and acknowledging them is part of using it well.
- Daryanani is the only paper that addresses maintenance — how to keep a portfolio at its targets without overtrading. The other four assume this problem is solved.
Editor's read
If forced to rank these for actionability rather than influence, the editor places Sharpe (1991) and Daryanani (2008) first — not because they are the most cited, but because they directly govern decisions a retail investor faces every year. Markowitz and Fama-French are the conceptual scaffolding; without them the others do not cohere. Jegadeesh-Titman is mostly cautionary in this audience: the premium is real, the implementation is hard, and most retail momentum products dilute the academic effect into something close to a closet index with extra turnover. The boring conclusion the literature actually supports — broad low-cost exposure, modest factor tilts where conviction is high, mechanical rebalancing — is the one worth holding.
The editor does not hold any specific positions referenced in this article (the piece concerns academic papers rather than tickers).
Methodology
Paper citations are the published versions in their original journals: Markowitz, Journal of Finance (March 1952); Fama-French, Journal of Financial Economics (1993); Jegadeesh-Titman, Journal of Finance (1993); Sharpe, Financial Analysts Journal (January-February 1991); Daryanani, Journal of Financial Planning (January 2008). Long-run factor premium estimates are drawn from the Kenneth R. French Data Library (mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html). Rebalancing band conventions are drawn from Daryanani (2008) and Vanguard's 2024 rebalancing research note. Macro snapshot data (10-year Treasury, VIX) is from FRED, asof 2026-05-01. A companion in-house portfolio review dashboard implements the Daryanani-style band logic referenced in section 5.
This article is for educational purposes and does not constitute personalized financial advice. See the full Disclaimer.