The moving average crossover strategy is a trend-following technique where you track two moving averages of different time periods on an ETF’s price chart. When the shorter average crosses above the longer one, it signals a potential uptrend (buy). When it crosses below, it signals a potential downtrend (sell or move to cash). It is one of the oldest and most widely used systematic trading approaches, and it works well with liquid ETFs that track broad indexes or sectors.

The Short Version

• A moving average crossover uses two MAs of different lengths — a fast one and a slow one.
• The “golden cross” (fast MA crosses above slow MA) is a buy signal; the “death cross” (fast crosses below slow) is a sell signal.
• The 50-day and 200-day simple moving average (SMA) pair is the most commonly watched crossover on ETFs like SPY and QQQ.
• The strategy works best in strong trending markets and generates false signals (whipsaws) in sideways, choppy conditions.
• You can apply it to any liquid ETF — broad market, sector, international, or even bond ETFs.

What Exactly Is a Moving Average Crossover?

A moving average crossover happens when a shorter-period moving average crosses above or below a longer-period moving average. The idea is simple: when recent prices are rising faster than the longer-term average, momentum is shifting upward. When they’re falling faster, momentum is shifting downward.

Two specific crossovers have earned their own names in market lore:

• Golden cross: The 50-day SMA crosses above the 200-day SMA. Widely interpreted as the start of a bull phase. Institutional investors and financial media watch this signal closely.
• Death cross: The 50-day SMA crosses below the 200-day SMA. Considered a bearish signal that often precedes extended drawdowns. Has historically preceded major downturns including 2000, 2008, and 2022.

You can use simple moving averages (SMA) or exponential moving averages (EMA). SMA gives equal weight to all prices in the period. EMA gives more weight to recent prices, making it more responsive — and more prone to false signals in choppy markets.

Pair	Type	Best For	Lag
10 / 30 EMA	Exponential	Short-term swing trades	Low
20 / 50 SMA	Simple	Medium-term trend	Medium
50 / 200 SMA	Simple	Long-term trend / investing	High
5 / 10 / 20 SMA	Simple (triple)	Confirming strong trends	Medium-High

How Do You Apply a Moving Average Crossover to an ETF?

Applying this strategy to ETFs is straightforward. Here’s the step-by-step process:

1. Pick your ETF. Choose a liquid, low-expense ETF that tracks a broad index (e.g. SPY, VOO, QQQ, IWM) or a sector you want to trend-trade.
2. Choose your moving average pair. Beginners often start with the 50/200 SMA crossover. For shorter holding periods, try 20/50 SMA or 10/30 EMA.
3. Plot both averages on a daily chart. Most brokers and free charting platforms (TradingView, Yahoo Finance) let you overlay multiple MAs with a few clicks.
4. Enter when the fast MA crosses above the slow MA. Buy the ETF at the next open. Set a stop-loss below the recent swing low or below the slow MA.
5. Exit when the fast MA crosses below the slow MA. Sell the position and hold cash (or move to a cash-like ETF like SHV or BIL) until the next buy signal.
6. Repeat. This is a systematic process. The only judgment call is whether to filter signals with a secondary indicator (e.g. RSI or volume) to reduce whipsaws.

A confirmation filter — such as requiring the price to be above the 200-day MA before taking a golden cross buy signal — can reduce false entries. The Pacer Trendpilot® ETF family uses a 5-day confirmation window on the 200-day SMA to avoid getting whipsawed by a single-day breach.

What Are the Most Common Moving Average Crossover Settings for ETFs?

Traders use different crossover pairs depending on their time horizon and risk tolerance. The table below compares four common setups:

Crossover Pair	Signal Speed	Whipsaw Risk	Best Market	Example ETF Application
10/30 EMA	Fast	High	Strong momentum	Short-term swings on QQQ
20/50 SMA	Medium	Medium	Moderate trend	Mid-term rotation in sector ETFs
50/200 SMA	Slow	Low	Persistent bull/bear	Long-term trend on SPY or VOO
5-10-20 SMA (triple)	Medium-Slow	Low-Medium	Clean trends	Trend confirmation on IWM

The 50/200 SMA crossover is the most recognized because financial media reports every golden cross and death cross on the S&P 500. But it comes with a real trade-off: the 200-day SMA lags significantly. By the time a golden cross appears, the market may already be 5-10% above the lows (source: GraniteShares).

For faster responses, the 20/50 SMA crossover strikes a reasonable middle ground — responsive enough to catch mid-trend entries without generating as many false signals as a short EMA pair.

What Are the Pros and Cons of the Moving Average Crossover Strategy for ETFs?

Strengths

• Simple and systematic. No subjective chart patterns or gut calls. The rules are clear and executable by anyone.
• Works in strong trends. During sustained bull or bear markets, the crossover keeps you in the right direction for most of the move.
• Low maintenance. You only check signals weekly or at month-end for longer settings. No intraday monitoring required.
• Applicable to any liquid ETF. Works on US equities, international (EFA, VXUS), sectors (XLK, XLV), and even fixed income (TLT, AGG) — as long as the ETF trends.

Weaknesses

• Lagging indicator. The crossover confirms a trend that’s already underway. You buy after the bottom and sell after the top. This eats into returns.
• Whipsaws in sideways markets. When prices chop back and forth without a clear trend, the moving averages cross repeatedly, generating a series of losing trades. This is the biggest practical drawback.
• No protection against gap risk. A market gap down overnight can hit before your crossover triggers an exit. This happened in 2020 (COVID crash) and 1987.
• Underperforms buy-and-hold in long bull markets. Because you exit during pullbacks, you miss some recovery days. Over a full bull run, holding through dips often wins.

Historical testing by Quantified Strategies found that a simple 200-day SMA strategy on SPY (a single MA, not a crossover) returned a CAGR of about 2.7% versus 9.8% for buy-and-hold, with a 68% win rate and significantly lower drawdown (approximately 25% versus 55%). A crossover strategy would produce different numbers depending on the settings, but the same trade-offs apply: lower peak returns in exchange for smoother equity curves.

How Do You Avoid False Signals (Whipsaws) in a Moving Average Crossover Strategy?

Whipsaws are the number-one reason traders abandon crossover strategies. Here are proven ways to reduce them:

• Use a confirmation filter. Only take a buy signal if the ETF’s price is already above the longer moving average. This prevents buying during a “dead cat bounce” that fails.
• Add a time delay. Wait N days after the crossover before acting. The Pacer Trendpilot methodology uses a 5-day confirmation period.
• Apply a volatility filter. Skip signals when the market is range-bound. The ADX indicator (values below 20-25) can identify low-trend environments.
• Switch to monthly signals. Instead of checking daily, evaluate crossovers at month-end. Fewer checks mean fewer trades and less whipsaw — though you may enter later.
• Use a wider MA pair. A 100/200 SMA crossover generates far fewer signals (and fewer whipsaws) than a 10/30 EMA cross, but you enter much later.

No filter eliminates whipsaws entirely. The goal is to reduce them enough that the strategy’s winning trades more than cover the losing ones.

Which ETFs Work Best with a Moving Average Crossover Strategy?

The strategy performs best on ETFs that trend reliably. Not all ETFs trend — some mean-revert or stay range-bound for years. Here’s what to look for:

• Broad market index ETFs (SPY, VOO, IVV, QQQ, IWM). These are the most popular for crossover strategies because they exhibit long, persistent trends.
• Sector ETFs (XLK, XLV, XLI, XLE). Sectors can have multi-year trends driven by economic cycles. The crossover catches entries and exits around those cycles.
• International ETFs (EFA, VWO, VXUS). Cross-border trends tend to be slower and longer, which suits the crossover approach well.
• Bond ETFs (TLT, AGG, BND). Bond ETFs trend during rate cycles. The 50/200 SMA crossover on TLT has historically captured most of the long-duration bull and bear phases.

ETFs that lack trend — such as high-dividend funds or low-volatility strategies that are designed to dampen movement — produce more whipsaws and fewer reliable signals. Test a potential ETF on a free backtest tool like ETFreplay before committing real capital.

Quick Self-Check

Test your understanding of the moving average crossover strategy with these quick questions.

What does a golden cross signal?

A golden cross occurs when the 50-day SMA crosses above the 200-day SMA. It signals that short-term momentum has overtaken the long-term trend, and is widely interpreted as the beginning of a new bull phase.

What is the biggest weakness of crossover strategies?

Whipsaws in sideways markets. When prices chop without a clear trend, the moving averages cross repeatedly, generating a series of small losing trades. This is why most traders add a confirmation filter or volatility check.

What’s the difference between SMA and EMA for crossovers?

SMA (simple moving average) gives equal weight to every price in the period — it’s smoother but slower to react. EMA (exponential moving average) weights recent prices more heavily — it’s faster to signal but generates more false crossovers in choppy markets.

Frequently Asked Questions

Can I use moving average crossovers on any ETF?

Yes, any liquid ETF that trends. Broad market index ETFs (SPY, QQQ) and sector ETFs work best. Mean-reverting or low-volatility ETFs generate too many false signals. It’s best to test on a free tool like ETFreplay before trading real money.

What is the best moving average crossover setting for beginners?

The 50/200 SMA crossover is the most widely followed and well-documented. It has fewer trades and lower whipsaw risk than shorter settings. The trade-off is that signals come late relative to the market turn.

How do I backtest a moving average crossover strategy for free?

ETFreplay.com offers a free moving average backtest tool limited to 5 ETFs with a 2002 start date. For more flexibility, TradingView’s Pine Script lets you code and backtest any crossover combination, or you can use Python with the yfinance library to run your own analysis.

Does the moving average crossover strategy beat buy-and-hold?

Not in consistent long bull markets. But in choppy or declining markets, it protects capital by moving to cash. Over full market cycles (including bear markets), crossover strategies often produce comparable or slightly lower total returns with significantly lower drawdown — which matters for capital preservation.

Can I use EMA instead of SMA for crossover signals?

Yes. EMA responds faster to recent price changes, so signals come a few days earlier. But that speed comes with more false signals (whipsaws) in sideways markets. Many traders use SMA for longer-term trend trading and EMA for short-term swing trading.

Summary

The moving average crossover strategy is a time-tested trend-following method that works well with ETFs. Its simplicity makes it accessible to beginners, and its systematic nature removes emotional decision-making. The key is matching your MA pair to your holding period, adding filters to reduce whipsaws, and accepting that no trend system catches every move.

For a deeper dive, read our companion articles on the 200-day moving average strategy and how to build a full ETF trend trading system.

Educational content only — not financial advice. Past performance does not guarantee future results. Trading involves risk of loss. Read our disclaimer and affiliate disclosure.

Dual momentum is a systematic ETF strategy that combines relative momentum (comparing assets to pick the strongest) with absolute momentum (checking whether the winner is worth owning at all). Created by Gary Antonacci and formalized in his 2015 book, it uses a simple monthly decision rule across just two or three ETFs. The strategy delivered ~17% annualized returns in its 40-year backtest with roughly one-fifth the drawdown of the S&P 500 — though real-world results since publication have been more mixed, which matters for setting expectations.

The short version

• Dual momentum = relative momentum (which asset is strongest?) + absolute momentum (is it worth owning?)
• The classic implementation is Global Equity Momentum (GEM): compare SPY vs EFA, buy the 12-month winner if it beats T-bills, otherwise buy AGG bonds
• Rebalanced monthly — about 5 minutes of work per month
• Backtest (1974–2013): ~17% annual return, ~10% max drawdown vs S&P 500’s 10% return and 51% drawdown
• Post-publication (2014–2022): underperformed buy-and-hold during the long US bull market — this is expected behavior, not a flaw
• Minimum account size: ~$5,000
• Tax efficiency: moderate — monthly rebalancing can trigger short-term capital gains

What exactly is dual momentum?

Dual momentum is a rules-based tactical asset allocation strategy. It applies two distinct momentum filters before deciding what to buy:

Relative momentum (cross-sectional): Among a defined set of assets, which one has performed best over the lookback period (typically 12 months)? The winner gets the allocation. This is the “which one?” question.

Absolute momentum (time-series): Is the winning asset actually in an uptrend? The strategy compares the winner’s return against a risk-free benchmark (T-bills). If the winner can’t beat T-bills, the strategy moves to defensive bonds. This is the “should I own it at all?” question.

By combining both filters, dual momentum avoids two common pitfalls: buying a weak asset just because it’s the “best of a bad bunch” (relative momentum alone), and buying a strong rising asset without checking whether the broad trend supports it (absolute momentum alone).

Who created dual momentum?

Gary Antonacci, an independent researcher, published the framework in his 2012 SSRN paper “Risk Premia Harvesting Through Dual Momentum” and expanded it into the 2015 book Dual Momentum Investing: An Innovative Strategy for Higher Returns with Lower Risk. Antonacci didn’t discover momentum as a market anomaly — that credit belongs to academics Jegadeesh and Titman (1993) — but he was the first to package relative and absolute momentum into a single executable system for retail investors.

The academic foundation runs deep. Jegadeesh and Titman (1993, confirmed 2001) showed stocks with strong 3–12 month returns continued to outperform. Asness, Moskowitz, and Pedersen (2013) at AQR Capital extended this across global stocks, bonds, commodities, and currencies. Momentum is widely considered the most persistent market anomaly. Antonacci’s contribution was making it practical: two ETFs, one comparison, one decision per month.

How does the Global Equity Momentum (GEM) strategy work?

GEM is the most popular implementation of dual momentum. Here is the exact decision rule:

1. Compare the 12-month total return of SPY (US equities) vs EFA (developed international equities).
2. If the winner’s 12-month return is higher than the 1-month T-bill rate, buy that winner.
3. If both SPY and EFA have negative absolute momentum (below T-bills), buy AGG (US aggregate bonds).
4. Rebalance once per month. No adjustments between monthly signals.

That’s the entire strategy. Twelve decisions per year, most taking under a minute. The bond position is the “risk-off” parking spot — it preserves capital until one of the equity ETFs re-establishes positive absolute momentum.

What ETFs do you need for dual momentum?

ETF	Ticker	Role	Expense Ratio
SPDR S&P 500 Trust	SPY	US large-cap equities	0.09%
iShares MSCI EAFE ETF	EFA	Developed international equities	0.33%
iShares Core US Aggregate Bond ETF	AGG	Defensive bond position	0.03%

These are the three ETFs used in the standard GEM model. Some practitioners substitute VTI for SPY (total US market instead of S&P 500) or BND for AGG, but the core selection logic remains identical. Antonacci’s own backtests used index data rather than specific ETFs, so these tickers are practical approximations vetted by the community.

Account minimum: because GEM holds only one ETF at a time and AGG trades at roughly $100/share, the strategy works with accounts as small as $5,000. Larger accounts benefit from lower trading costs as a percentage of position size.

How do you calculate the signals each month?

You need two pieces of data for each ETF: the current price and the price from 12 months ago. Most brokers provide 12-month return figures directly. Alternatively, you can calculate it manually:

1. Get the adjusted close price of SPY and EFA from 12 months ago and today.
2. Calculate 12-month return = (current price / price 12 months ago) — 1.
3. Get the 1-month T-bill rate (available at Treasury.gov or via your broker).
4. Apply the GEM decision rule above.
5. If the signal changes, place the trade. If it stays the same, do nothing.

For retail traders who don’t want to calculate this manually each month, several services publish the GEM signal: Antonacci’s own Optimal Momentum site, and tools like Reblnc provide automated signal tracking. The Quant Investing ETF Momentum dashboard also ranks ETFs by 12-month return across global markets.

What performance did the backtest show?

Antonacci’s backtest running from 1974 through 2013 showed these results (source: Antonacci, 2012 SSRN paper):

Metric	GEM (Dual Momentum)	S&P 500 Buy-and-Hold	60/40 Portfolio
Annualized return	~17%	~10%	~10%
Max drawdown	~10%	~51%	~27%
Period	1974–2013	1974–2013	1974–2013

The drawdown comparison is the standout finding. GEM’s worst peak-to-trough loss was roughly 10%, versus 51% for buy-and-hold during the 2000 dot-com crash and 2008 financial crisis. The absolute momentum filter rotated the portfolio to bonds before the worst losses materialized, then rotated back to equities once trends re-established. Independent analyses using ETF data (which are constrained by shorter ETF lifespans) broadly confirm the pattern, though the magnitude of outperformance is smaller over the ETF-only period.

How did dual momentum perform after publication?

This matters because backtests are not reality. From roughly 2014 through 2022, GEM underperformed a simple S&P 500 buy-and-hold. The reasons are straightforward and honest:

• The US bull market ran for years with shallow corrections — GEM’s defensive trigger rarely activated
• When GEM rotated to international equities (EFA), it significantly underperformed SPY due to US market dominance
• Whipsaw trades — brief signals that reversed quickly — generated transaction costs and caused the strategy to miss parts of rallies

This is not a sign the strategy is broken. Long bull markets with shallow drawdowns are the worst-case scenario for any trend-following approach. GEM is designed to protect against crashes, not to maximize returns when the market goes up almost every year. The strategy did what it was supposed to do in March 2020 — rotated to bonds ahead of the COVID crash, then back to equities in April — but the crash and recovery were so fast that some investors missed the rotation entirely.

GEM remains the benchmark that tactical ETF strategies are measured against precisely because its rules are fully transparent and independently verifiable. No black box, no proprietary indicators, no data mining.

What are the limitations of dual momentum?

Every strategy has weaknesses. Dual momentum’s are well-documented:

• Underperforms in long bull markets. As noted above, the defensive mechanism is a drag when drawdowns are shallow and brief.
• Tax inefficiency. Monthly rebalancing means holdings rarely qualify for long-term capital gains treatment. Best used in tax-advantaged accounts (IRA, 401k).
• Bond correlation risk. In a rising-rate environment, AGG can fall alongside equities, reducing the defensive benefit.
• Whipsaw in choppy markets. When the 12-month return oscillates around the T-bill threshold, the strategy can trigger multiple round-trip trades in short succession.
• US-centric. GEM only considers US and developed international equities. It misses emerging markets, sector rotation, and other diversifiers that some modified dual momentum strategies include.

These limitations have directly shaped the next generation of tactical strategies — HAA (Heritage Asset Allocation) and DAA (Defensive Asset Allocation) were designed specifically to address GEM’s weaknesses while preserving its core logic.

How is dual momentum different from simple momentum?

“Simple momentum” strategies typically use only relative momentum: rank assets by recent performance and buy the top one (or top few). The problem is that during a broad market crash, the “best” asset might still be down 20%. You’re buying the least bad option in a bad market.

Dual momentum’s second filter — absolute momentum — solves this. It checks whether the winner is in an actual uptrend. If no asset passes the absolute momentum test, the strategy exits to bonds or cash. This is what protected the 1974–2013 backtest from large drawdowns: during the 2008 financial crisis, both SPY and EFA had negative 12-month returns, so GEM sat in AGG and avoided the worst of the crash.

This makes dual momentum more conservative than pure momentum. You sacrifice some upside during strong bull markets in exchange for crash protection. Whether that trade-off suits you depends on your personal risk tolerance.

Quick self-check

Test your understanding of dual momentum with these three questions.

What are the two types of momentum in dual momentum?

Relative momentum (cross-sectional — comparing assets to find the strongest) and absolute momentum (time-series — checking if the winning asset is in an uptrend relative to T-bills). Both must be applied together; using only one misses the strategy’s core protection.

When does GEM rotate to bonds instead of holding equities?

When the winner of the SPY vs EFA comparison fails the absolute momentum test — meaning its 12-month return is lower than the 1-month T-bill rate. In that case, the strategy buys AGG (US aggregate bonds) until one of the equity ETFs re-establishes positive absolute momentum.

Why did GEM underperform buy-and-hold from 2014 to 2022?

The US bull market had very shallow drawdowns, so GEM’s defensive trigger rarely activated. When it did rotate to international equities (EFA), those significantly underperformed US stocks, and brief whipsaw signals caused the strategy to miss parts of rallies. This is expected for any crash-protection strategy during a long, low-volatility bull run.

Frequently asked questions

Can I run dual momentum in a small account?

Yes. Since GEM holds only one ETF at a time and the cheapest option (AGG) trades around $100/share, accounts of $5,000 or more can implement it efficiently. Some brokers offer fractional shares, further lowering the barrier.

What lookback period does dual momentum use?

The standard is 12 months, based on Jegadeesh and Titman’s academic research showing the 3–12 month window produces the strongest momentum effect. Antonacci tested various periods and settled on 12 months as the best balance of responsiveness and signal reliability for the GEM model.

Does dual momentum work in bear markets?

That’s where it performs best. During extended downturns, the absolute momentum filter rotates to bonds before the worst losses hit. In the 2008 financial crisis, GEM avoided most of the ~50% S&P 500 drawdown by holding bonds. The strategy is specifically designed for crash protection, not bull market maximization.

Can I add more ETFs to the dual momentum universe?

Yes. Many practitioners create expanded versions that include emerging markets (EEM), REITs (VNQ), gold (GLD), or sector ETFs. This is sometimes called “sector rotation dual momentum.” The trade-off is more possible signals to track and potentially more whipsaw trades. Start with the classic three-ETF GEM before experimenting.

Is dual momentum suitable for retirement accounts?

Yes — in fact, retirement accounts are the natural home for this strategy. Since monthly rebalancing triggers mostly short-term capital gains, holding GEM in a tax-advantaged account (IRA, 401k, SIPP) avoids the tax drag. The lower drawdown risk also suits investors who cannot afford a 50% portfolio loss near retirement.

Educational content only — not financial advice. Past performance does not guarantee future results. Trading involves risk of loss. Disclaimer · Affiliate disclosure