高估值时代的复利危机，美股会迎来新的「失落十年」吗？

Original title: When the Decade Goes Missing

Original author: AdvisorAnalyst Editorial Team

Original translation: Peggy, BlockBeats

Editor's note: The faith in holding stocks for the long term is often built on a sufficiently long time scale: as long as the cycle is extended, the market will eventually reward patience. But for real investors, time is not an abstract variable. Retirement, cash flow, redemption pressure, and emotional fluctuations can all turn "long-term average returns" into a promise that cannot always be fulfilled.

Based on 155 years of U.S. stock market history, this article reviews three periods of prolonged real return stagnation: 1929–1954, 1966–1982, and 2000–2013. It points out that the so-called "lost decade" is not a historical anomaly but a recurring structural phase of equity markets. These periods collectively account for about 35% of market history since 1871, and they bring not just a delay in wealth growth but permanent damage to the compounding path.

The article further warns that several current valuation metrics of the U.S. stock market are at historical highs: CAPE is near the 99th percentile since 1881, and the Buffett Indicator, Tobin's Q, and the equity risk premium all point to a similarly fragile environment. At the same time, the author refutes the traditional "missing the best trading days" argument, pointing out that most of the best single-day gains actually occur during bear markets and crisis phases, often adjacent to the worst trading days. For investment advisors and long-term investors, the issue is not about predicting when the next crisis will arrive, but whether they can identify risks in advance through signals like valuation and market breadth, and protect compounding from being damaged before a long period of low returns arrives.

Below is the original text:

The traditional argument for stock investing is built on long-term average returns. But it does not fully consider what happens when a client's wealth accumulation phase happens to fall within the wrong 16 years.

Ryan Gorman, CFA, CMT, Shawn Keel, CFA, CMT, and Vincent Randazzo, CMT, portfolio managers at Tamarisk Capital Management and Quoin Capital Analytics, published a research paper through the CMT Association that deserves a place on every investment advisor's desk: "Navigating the Lost Decade: Protecting Long-Term Compounding in Secular Bear Markets." Based on 155 years of data from Robert Shiller’s Yale University database, the article presents a judgment that is both empirically robust and strategically urgent: the so-called "lost decade" is not an anomaly but a structural characteristic of the stock market. The current market environment shares similarities with the eve of these historical phases and deserves serious attention.

The Historical Record Has Already Given a Clear Answer

The authors identify three distinct phases in the U.S. stock market during which buy-and-hold investors received virtually no real returns. From 1929 to 1954, it took the market 25 years to return to its previous real peak. During the stagflationary period from 1966 to 1982, the annualized real return was approximately -1.77% over 16 years. From 2000 to 2013, spanning the bursting of the internet bubble and the global financial crisis, the annualized real return was about 0.05%, with a maximum drawdown of 52%. These three phases combined account for 54 years of market history, approximately 35% of the entire period since 1871.

The authors state directly: "Lost decades do not need to be triggered by exactly the same catalysts. They can appear in different economic cycles and institutional environments, but they bring the same experience to investors: prolonged drawdowns, damaged compounding, and negative behavioral reactions that often persist long after the market eventually recovers."

International precedents further reinforce this judgment. Japan's Nikkei 225 hit a peak of 39,000 in December 1989 and did not reclaim this level until 2024, a period of 35 years. Europe's Euro Stoxx 50 peaked in March 2000 and only returned to that high by the end of 2025. The authors caution that the pattern of the U.S. market always eventually recovering "should not be viewed as an immutable law."

The Mathematical Mechanism That Makes Losses Permanent

This is also where the paper's analytical contribution goes beyond historical review. The authors demonstrate that lost decades do not merely delay wealth accumulation; they cause permanent damage. Assuming two portfolios both target a long-term average return of 7%, but one experiences a 13-year period of zero returns in the middle of its investment journey, the difference in final terminal value will be significant. Path B can only achieve 80% of the terminal value of Path A. This gap is permanent and cannot be compensated for even if normal returns resume afterward.

The mathematical conditions required for recovery further amplify the problem. A 50% drawdown requires a 100% gain to break even. If the annualized return is only 3%—consistent with the level of returns available in historically high-valuation environments—then breaking even would take 23.4 years. The authors clearly state: "This is the hidden cost of the lost decade: it brings not only the low returns of that period itself but also a permanent impairment of the compounding path."

Valuation Context: The 99th Percentile

The valuation section of the paper presents a finding that investment advisors should not easily overlook. The current CAPE (Cyclically Adjusted Price-to-Earnings ratio) stands at 39.9, placing it at the 99th percentile of all historical observations since 1881. Historically, only one instance has exceeded the current level: the peak of 44.2 in March 2000. The historical average for CAPE is 17.7.

The authors are cautious in their wording—CAPE is not a timing tool—but its directional signal is very clear. When CAPE is in the lowest historical quintile, the average real return over the following 10 years is 10.7%, with no instances of negative returns. When CAPE is in the highest quintile, the average real return over the following 10 years is only 3.6%, with 24% of the observations showing negative returns. The Buffett Indicator (total market capitalization to GDP) is currently near 190%, higher than the peaks of 2000 and 2007. Tobin's Q and the equity risk premium convey the same signal.

"When CAPE, market cap/GDP, Tobin's Q, and the equity risk premium all indicate elevated valuations, historical records suggest that the market's margin for error is narrowing."

Deconstructing the 'Missing the Best Trading Days' Argument

The most practically valuable part of the paper is its direct response to the industry's most common rhetoric against tactical management. The authors examined the 20 best trading days for the S&P 500 between 1988 and 2025 and found that 18 of them, or 90%, occurred when the index was below its 200-day moving average. 42% of the best trading days occurred in traditionally defined bear markets.

This means: "The best trading days are not randomly distributed between bull and bear markets. They tend to cluster in crisis periods when prices are depressed." Furthermore, these best trading days during crises are often interspersed with the worst trading days. In October 2008, the market's largest single-day gain (+11.6%) occurred just days after its largest single-day loss. The two cannot be easily separated. The authors note: "Investors cannot capture only the best trading days during these periods without also experiencing the worst ones."

Market Breadth Framework: What to Watch

The final part of the paper proposes a systematic framework for identifying market states, based on market breadth—observing the degree of participation among different securities rather than relying solely on the average performance of a market-cap-weighted index. The core insight is that structural deterioration "often manifests first in market breadth before appearing in the market-cap-weighted price index."

Prior to the 1973–1974 bear market, the advance-decline line had already diverged from the S&P 500 in early 1973. In 1999, market breadth consistently deteriorated, preceding the 2000 technology stock crash. The authors argue that market breadth can provide "earlier warnings than indicators based purely on price trends." When combined with the valuation backdrop, this framework becomes even more explanatory: "High valuations set the environment... while deteriorating market breadth provides behavioral evidence."

Key Takeaways for Investment Advisors

The paper's conclusion is well-suited for inclusion in conversations with clients: "The question is not about choosing between optimism and pessimism, but about choosing between complacency and preparation."

Specifically, investment advisors should understand four points from this research. First, sequence-of-returns risk is not a theoretical concept. For 35% of U.S. market history, we have been in a "lost decade," and if a client retires during such a phase, they face not a temporary delay but permanent damage to compounding. Second, a CAPE at the 99th percentile does not predict a specific point in time, but it does define a more fragile market environment. Valuation and market breadth are not competing signals but complementary ones. Third, the "missing the best trading days" objection does not hold up to empirical scrutiny because these best days tend to cluster in the same periods as the worst days; systematically managing drawdowns means avoiding both simultaneously. Fourth, an adaptive framework prioritizing market breadth does not require precise timing. It requires "a disciplined response to observable conditions, not a prediction of future outcomes."

The authors do not claim that a fourth lost decade is inevitable. What history truly shows is that the conditions often present on the eve of lost decades can be identified, and that preparation always provides a more resilient foundation compared to passive acceptance.

Original Link