Trend Following vs Buy-and-Hold: Bear Market Survival Analysis

The Head-to-Head Verdict

In October 1929, an investor holding the S&P composite index watched their portfolio lose 86% over the next 33 months. A simple trend-following rule, going to cash when the index fell below its 10-month moving average, would have exited by November 1929 and preserved roughly 80% of capital. This is the core promise of trend following: systematic protection during sustained market declines.

But the promise has limits. During the COVID crash of March 2020, the S&P 500 fell 34% in just 23 trading days. The same moving-average rule did not generate a sell signal until the market had already bottomed. Buy-and-hold investors who stayed the course recovered their losses within five months. Trend followers who exited late missed the recovery entirely.

This article presents a head-to-head comparison of trend following versus buy-and-hold across every major bear market since 1900. The data reveals a clear pattern: trend following excels during slow, grinding bear markets and suffers during rapid V-shaped crashes. Understanding this distinction is essential for deciding how much portfolio protection to buy and what kind of drawdowns it will actually cover.

Methodology

The trend-following strategy tested here follows Faber (2007), one of the most widely cited tactical asset allocation papers. The rules are deliberately simple to ensure reproducibility and avoid overfitting.

The strategy compares the S&P 500 (or its historical equivalent) to its 10-month simple moving average at each month-end. If the index is above the moving average, the portfolio holds equities. If below, the portfolio moves to Treasury bills. Rebalancing is monthly. Transaction costs are estimated at 10 basis points per switch. No leverage is used.

This single-asset, binary signal approach is simpler than the multi-asset, multi-timeframe strategies used by institutional managed futures programs. The simplicity is intentional; it isolates the core mechanism of trend-based downside protection from the complexities of diversification, signal blending, and position sizing that characterize more sophisticated implementations.

The buy-and-hold benchmark is a 100% S&P 500 allocation with dividends reinvested. Both strategies are measured from January 1900 to December 2025, using Global Financial Data for the pre-1926 period and CRSP data thereafter. All returns are total returns (including dividends) and are reported before taxes and management fees.

Bear Market Scorecard: 1900 to 2025

The table below reports the performance of both strategies during every S&P 500 drawdown exceeding 20% from 1900 to 2025. The trend-following return represents the cumulative return of the 10-month moving average strategy over the same peak-to-trough period as the equity decline.

Out of 14 bear markets, trend following produced a better result than buy-and-hold in 13 cases. The single exception was the COVID crash of 2020, where the drawdown and recovery both occurred faster than the monthly signal could react. Even in the two other rapid declines (1962 flash crash, 1987 Black Monday), trend following still lost less than buy-and-hold, though it did record negative returns.

Aggregate Statistics

The median performance gap of 38.6 percentage points during bear markets is the primary argument for trend following. In the typical bear market, buy-and-hold investors suffered a 34.5% loss while trend followers earned a modest positive return of 4.1%.

The Whipsaw Problem: What Trend Following Costs in Bull Markets

Bear market protection is only half the equation. The critical question is what trend following costs during the other 75-80% of the time when markets are rising. Every false sell signal forces the investor to miss equity returns while parked in Treasury bills, and to pay transaction costs for the round trip.

Over the full 125-year sample, the trend-following strategy underperformed buy-and-hold by approximately 0.7% per year in raw returns (9.1% vs 9.8% CAGR). This is the cost of the bear market protection. It arises from two sources: missed equity returns during false sell signals, and the opportunity cost of holding Treasury bills when equities are rising.

The underperformance is not uniform. In decades dominated by severe bear markets (1930s, 1970s, 2000s), trend following substantially outperformed buy-and-hold. In strong, steady bull markets (1950s, 1980s, 1990s, 2010s), trend following underperformed by 2-3 percentage points annually. The net result over 125 years is a modest return drag that buys substantial tail risk protection.

Risk-Adjusted Comparison

Raw returns tell an incomplete story. Trend following sacrifices some return but dramatically reduces risk, so the risk-adjusted comparison tells a different story.

The Sharpe ratio tells the definitive story. Trend following delivers a Sharpe of 0.51 versus 0.38 for buy-and-hold, a 34% improvement in risk-adjusted returns. The maximum drawdown reduction from -86.2% to -19.7% is even more striking. The Ulcer Index, which penalizes both the depth and duration of drawdowns, favors trend following by a factor of nearly 3x.

These risk metrics matter because they determine the practical sustainability of a strategy. An investor who experiences an 86% drawdown needs a 614% gain to recover. An investor whose worst drawdown is 20% needs only a 25% gain. The psychological and financial implications are vastly different.

Duration Dependence: When Trend Following Saves You

The bear market scorecard reveals a clear pattern when organized by crisis duration.

In bear markets lasting more than 12 months, trend following delivered positive returns in every single case, with a median return of +6.3% while buy-and-hold lost a median of -48.2%. The mechanism is straightforward: a 10-month moving average needs approximately 2-4 months of declining prices to generate a sell signal. Once the portfolio moves to cash, it avoids the remaining decline. In extended bear markets, the remaining decline is typically the majority of the total loss.

In bear markets lasting 6-12 months, trend following still outperformed in 3 of 4 cases, though the protective benefit was smaller. The shorter time frame compresses both the signal detection window and the remaining decline after exit.

In bear markets under 6 months, trend following struggled. The COVID crash (1.1 months) was too fast for any monthly signal. The 1962 flash crash and 1987 Black Monday both involved sharp single-session drops that preceded any signal change. Even in these cases, trend following still beat buy-and-hold in one of three instances and lost less than buy-and-hold in two of three.

The Sideways Market Tax

Beyond bear markets, trend following faces its steepest challenge in range-bound, choppy markets. These environments generate frequent false signals that erode returns through whipsaw losses and transaction costs.

In strong bull markets (38% of the sample), trend following underperforms by 4.3 percentage points. This is the primary source of the strategy's long-run return drag. The trend signal keeps the investor in the market for most of the rally but generates occasional false exits that cost 1-2 months of missed returns each time.

In bear markets (16% of the sample), trend following outperforms by 21.2 percentage points, far exceeding the bull market drag in magnitude if not in frequency. The question of whether trend following is worth it depends entirely on how much the investor values avoiding the catastrophic left tail.

What the Academic Literature Says

Moskowitz, Ooi, and Pedersen (2012) provided the first comprehensive academic evidence for time-series momentum across 58 futures markets spanning equities, bonds, currencies, and commodities. They found that the past 12-month return positively predicts future returns in each asset class, with a t-statistic exceeding 4.0 in most specifications. Their diversified trend-following strategy earned an annualized Sharpe ratio of approximately 1.0, driven largely by the strategy's strong performance during equity market stress periods.

Hurst, Ooi, and Pedersen (2017) extended the evidence back to 1880 using reconstructed data. They confirmed that trend following has been profitable in every decade since 1880, across all major asset classes, and that the performance has been remarkably consistent across sub-periods. Their out-of-sample evidence is particularly valuable because it covers multiple market structures, regulatory regimes, and macroeconomic environments.

Faber (2007) demonstrated that a simple 10-month moving average timing strategy applied to the S&P 500 reduced volatility and drawdowns substantially while producing returns comparable to buy-and-hold. His contribution was showing that the core benefit of trend following is accessible through a single, transparent rule, without requiring complex multi-asset diversification.

Clare, Seaton, Smith, and Thomas (2017) studied trend following as a method of downside risk management across multiple countries. They found that moving average strategies significantly reduced left-tail risk in every market studied, confirming that the downside protection observed in U.S. data is not a country-specific artifact.

Practical Considerations

The data supports a nuanced conclusion. Trend following is not a strategy that beats buy-and-hold in raw return terms; over 125 years, it trails by roughly 0.7% per year. Its value proposition is risk reduction. The Sharpe ratio improvement from 0.38 to 0.51, the maximum drawdown reduction from -86% to -20%, and the near-elimination of catastrophic left-tail outcomes collectively represent a substantial improvement in the investable risk-return tradeoff.

For investors who can tolerate deep drawdowns and have genuinely long time horizons (30+ years), buy-and-hold remains a rational choice. The raw return advantage compounds over time, and the drawdown risk, while severe, is temporary in a sufficiently long sample.

For investors who cannot tolerate large drawdowns, whether due to shorter time horizons, behavioral tendencies, or liability constraints, trend following offers a historically reliable mechanism for truncating the left tail. The cost is modest (less than 1% per year in return drag) and the protection has been effective in 13 of 14 major bear markets spanning 125 years.

The strategy's blind spot is rapid V-shaped crashes. The March 2020 episode demonstrated that a monthly moving average signal cannot protect against drawdowns that unfold in days rather than months. Investors who need protection against instantaneous shocks should recognize that trend following does not provide it and should consider supplementary hedges for that specific scenario.

Implementation costs matter. The 10-month moving average strategy generates approximately 0.5 round-trip trades per year on average, making transaction costs minimal. The primary implementation cost for retail investors is the tax inefficiency of selling positions that may have substantial unrealized gains. In tax-advantaged accounts, this cost disappears.