Meb Faber's GTAA: Global Tactical Asset Allocation

In 2007, quantitative researcher Meb Faber published a paper that would become one of the most cited works in tactical asset allocation: "A Quantitative Approach to Tactical Asset Allocation." The paper introduced a deceptively simple idea: apply a 10-month moving average to each asset class in a diversified portfolio, holding the asset when it is above its average and moving to cash when it is below.

The results were striking. Over an 80-year backtest, this simple rule reduced the maximum drawdown of a diversified portfolio from −46% to −13% while maintaining similar long-term returns. The strategy, which Faber later expanded into his Global Tactical Asset Allocation (GTAA) framework, became a blueprint for individual investors seeking institutional-grade risk management.

The Core Concept

Trend Following Meets Asset Allocation

GTAA combines two well-established investment principles:

1. Broad asset class diversification: Holding a mix of asset classes (stocks, bonds, real estate, commodities, international equities) that respond to different economic environments
2. Trend following: Using a moving average to determine whether each asset is in an uptrend (hold) or downtrend (move to safety)

The combination is more powerful than either principle alone. Diversification provides structural resilience across economic regimes. Trend following provides dynamic risk management that static diversification cannot — the ability to step aside when an asset class is deteriorating regardless of the portfolio's overall diversification.

The 10-Month Simple Moving Average

The signal is straightforward: for each asset, compare the current monthly closing price to the average of the previous 10 monthly closing prices.

• If the current price is above the 10-month SMA → hold the asset
• If the current price is below the 10-month SMA → sell the asset and move the proceeds to cash (T-bills)

This is checked once per month, at month-end. No daily monitoring, no intraday decisions, no complex calculations. A spreadsheet or free charting tool is sufficient to run the signals.

Why 10 Months?

The 10-month SMA is approximately equivalent to the 200-day moving average, which is the most widely used trend indicator in technical analysis. Faber tested numerous lookback periods (3, 6, 8, 10, 12, 15 months) and found that performance was relatively insensitive to the exact choice — any period between 8 and 12 months produced similar results.

This robustness is important. It means the strategy's edge does not depend on a precisely calibrated parameter (which would suggest overfitting). It works because trends in asset classes tend to persist for months or years, and a medium-term moving average captures these trends regardless of the exact lookback length.

GTAA Variants

Faber developed several variants of the GTAA framework, differing primarily in the number of asset classes:

GTAA 5

The original five-asset version:

Asset	Weight	ETF
U.S. Stocks	20%	SPY
International Stocks	20%	EFA
U.S. Bonds	20%	AGG
Real Estate	20%	VNQ
Commodities	20%	DBC

Each asset gets a 20% allocation with the 10-month SMA filter applied independently. The portfolio can range from 100% invested (all five above their SMA) to 100% cash (all five below).

GTAA 13

An expanded version with 13 asset classes for finer diversification:

• U.S. large cap, U.S. mid cap, U.S. small cap
• International developed, international emerging
• U.S. 10-year bonds, U.S. 30-year bonds, TIPS, international bonds
• Real estate, commodities
• Broad market

Each asset receives approximately equal weight (roughly 7.7% each), with the same SMA filter applied independently. The finer granularity allows the strategy to differentiate between, say, small-cap stocks (which may be in a downtrend) and large-cap stocks (which may still be trending positively).

Aggressive GTAA

Faber also proposed aggressive variants that invest only in the top-ranked assets by momentum rather than holding all assets above their SMA. For example, GTAA-AGG3 holds only the top 3 of 13 asset classes by 12-month return. This concentrates the portfolio in the strongest trends, producing higher returns but with higher volatility and turnover.

Historical Performance

GTAA's historical performance demonstrates the power of combining diversification with trend following:

GTAA 5 (1973-2023):

• CAGR: Similar to buy-and-hold (approximately 8-10%)
• Maximum drawdown: Approximately −10% to −15% (vs. −46% for buy-and-hold)
• Sharpe ratio: Approximately 0.7-0.9
• Time in cash: Approximately 25-35% of months

The return is similar because the strategy captures most of the upside during bull markets (assets are above their SMA and fully invested) while avoiding the catastrophic downside during bear markets (assets break below their SMA and move to cash). The dramatically reduced drawdown preserves compounding and allows faster recovery.

Key Crisis Performance

2000-2002 Dot-Com Crash: GTAA moved to cash in U.S. equities by early 2001 as they broke below their 10-month SMA. International stocks and REITs provided some continued exposure. Maximum drawdown during this period was approximately −8% vs. −45% for the S&P 500.

2008 Financial Crisis: GTAA exited most risk assets by mid-2008. By October 2008, the strategy was nearly 100% in cash. It missed the worst of the decline (September-November) and only re-entered as assets recovered above their moving averages in mid-2009.

2022 Rate Shock: GTAA moved bonds to cash as they broke their SMA. Equities were moved to cash progressively. Commodities remained invested (above their SMA). The portfolio was approximately 60% cash by mid-2022, significantly reducing the drawdown.

Why GTAA Works

Asset Class Trends Persist

The fundamental reason GTAA works is that asset class trends tend to persist. When equities enter a bear market, they typically decline for 12-24 months — far longer than the 10-month lookback period of the SMA. This persistence gives the signal enough time to detect the trend change and move to safety before the worst of the decline.

Conversely, bull markets in asset classes typically last 3-10 years. The SMA stays below the price throughout these extended uptrends, keeping the strategy fully invested and capturing the majority of gains.

Independent Signals per Asset

Applying the SMA filter independently to each asset — rather than using a single signal for the whole portfolio — allows the strategy to adapt to environments where different assets are in different trends simultaneously. In 2022, for example, commodities were trending upward while stocks and bonds were trending downward. GTAA held commodities and moved stocks and bonds to cash — a nuanced response that a single-signal system could not achieve.

The Benefit of Doing Nothing

GTAA generates relatively few trades — typically 2-4 per asset per year. It is fully invested most of the time and only moves to cash when conditions clearly deteriorate. This low turnover reduces transaction costs and tax events while avoiding the whipsaw risk of more frequent trading systems.

Limitations and Challenges

Whipsaw Around the Moving Average

When an asset's price oscillates around its 10-month SMA (neither clearly trending up nor down), the strategy generates false signals — buying when the price briefly crosses above and selling when it crosses back below. These whipsaws generate small losses and transaction costs without capturing meaningful moves.

Mitigation strategies include adding a buffer zone around the SMA (only triggering a signal when the price is more than 1-2% above or below), using a combination of multiple moving averages, or requiring confirmation from additional signals before acting.

Late Entry During V-Shaped Recoveries

GTAA is slow to re-enter after sudden, sharp recoveries. During the March 2020 COVID crash, the S&P 500 fell 34% in three weeks and then recovered to its prior peak within five months. GTAA exited near the bottom (the price crossed below the SMA during the decline) but did not re-enter until the price crossed back above the SMA — missing a significant portion of the recovery.

This is the fundamental trade-off of trend following: it protects against prolonged bear markets (which are far more damaging to wealth) but gives back some gains during V-shaped recoveries (which are less damaging because the decline itself was brief).

Cash Return Matters

GTAA spends a meaningful percentage of time in cash. During the 2009-2021 period of near-zero interest rates, cash earned essentially nothing — creating a drag whenever the strategy was defensive. In the current higher-rate environment, the cash position actually earns 4-5%, making the defensive periods less costly.

GTAA on PortfolioWiser

PortfolioWiser's strategy library includes approaches that build on Faber's GTAA framework — broad multi-asset strategies with trend-following overlays. The platform's Strategy Builder allows you to customize the key parameters: asset universe, moving average period, and defensive asset selection.

You can also blend GTAA-style strategies with other tactical approaches (momentum-ranking, canary-based, macro-driven) to create multi-strategy portfolios that address GTAA's limitations while preserving its strengths. The backtesting tools show exactly how each combination performs across different market environments, including the critical 2008 and 2022 stress periods.

Frequently Asked Questions

Can I run GTAA myself without software?

Yes. GTAA requires only a comparison of each asset's current price to its 10-month average — a calculation that can be done in any free charting tool or spreadsheet. The monthly time commitment is approximately 15 minutes. However, using a platform like PortfolioWiser automates the signal calculation and provides additional context (performance metrics, drawdown analysis, blending tools) that manual tracking cannot easily replicate.

Is GTAA better with 5 or 13 asset classes?

GTAA 13 provides finer diversification and more granular trend signals, but it also requires more ETFs and generates more trades. For most individual investors, GTAA 5 provides the core benefit (trend-filtered diversification) with maximum simplicity. GTAA 13 is better for larger portfolios where the additional complexity is manageable.

What happens when all assets are below their moving average?

The portfolio goes 100% to cash (T-bills). This is rare — it happened during the worst of the 2008 crisis and briefly during other severe selloffs. Being 100% in cash during a market freefall is exactly the right position. The strategy re-enters each asset individually as it recovers above its moving average, gradually rebuilding the portfolio as conditions improve.