Does the 200-day Moving Average Actually Beat Buy-and-Hold?

Short answer: no, but it's the wrong question.

The 200-day moving average is one of the most-cited rules in trend-following. The strategy is dead simple: be long when the price is above its 200-day moving average, otherwise sit in cash. Practitioners swear by it. Skeptics roll their eyes. Most articles you read take a side without showing the actual data.

We ran the numbers across 16 years (2010–2026) on the most popular ETFs people actually use this rule on (QQQ, SPY, TQQQ, UPRO, SOXL) at realistic transaction costs. The data tells a more nuanced story than either camp likes.

The setup

The rule we're testing is the simplest possible:

• Each day after market close, check: is the symbol's closing price above its 200-day simple moving average?
• If yes → hold the position
• If no → be in cash
• Trade at the next day's open

No additional filters, no day-of-week tweaks, no stop-losses, no leverage targeting. Just one binary signal applied daily.

We compare this to buy-and-hold of the same symbol over the same period, with realistic transaction costs (3 basis points per side, conservative for retail commission-free brokerages).

The headline numbers

Here's what 16 years of data (2010-01 through 2026-05) shows:

Symbol	Buy-Hold Ann	SMA200 Filter Ann	Buy-Hold Sharpe	Filter Sharpe	Buy-Hold Max DD	Filter Max DD
QQQ	19.1%	12.4%	0.95	0.83	-35%	-23%
SPY	12.8%	9.5%	0.91	0.79	-34%	-24%
TQQQ	43.4%	24.5%	0.90	0.83	-82%	-34%
UPRO	29.3%	17.7%	0.76	0.64	-77%	-31%
SOXL	41.8%	22.1%	0.84	0.72	-90%	-42%

Buy-and-hold beats the SMA filter on absolute returns. Every single time.

Don't believe anyone who tells you "the trend filter beats buy-and-hold." On a 16-year window where the broad market basically went straight up with two notable interruptions (2020 and 2022), you can't beat being fully invested all the time. Math wins.

But look at the drawdown column. That's where the story gets interesting.

Drawdowns are the real game

Imagine you put $100,000 into TQQQ at the start of 2010.

Buy-and-hold path: rides the wave to roughly $30M at peak in late 2021. Then watches $25M of that vanish over the next 12 months as TQQQ falls 82% in 2022. Down to about $5M. Climbs back over the next two years as the market recovers. Long-term: incredible. Mid-term, during the drawdown: agonizing.

The vast majority of humans sell during the -82% drop. Including most of the people who swore they'd never sell. This is well-documented behavioral finance. Dalbar's annual study of investor returns vs fund returns consistently shows a 2-4 percentage point gap between what funds returned and what their investors actually earned, almost entirely due to poorly-timed exits during drawdowns.

SMA200 filter path: rides the same wave but exits in early 2022 when TQQQ closes below its 200-day SMA. Watches from cash as TQQQ falls another 70% from there. Re-enters in 2023 when the trend signal flips back. Misses some of the early recovery move. Maximum drawdown along the way: about -34%. Painful but survivable.

The buy-hold strategy has higher theoretical return on a screen. The SMA200 strategy has the return you actually capture because you stayed in the strategy through a full cycle.

The "I'd never sell" lie

Here's the awkward truth about backtests: they implicitly assume you held through the worst parts. Most real humans don't.

A strategy with lower headline returns but a max drawdown of -34% instead of -82% is a strategy you can actually stick with. That changes the calculation from "what does this strategy return on paper" to "what does this strategy return in your specific account, given that you're a human with feelings."

This is why headline backtest comparisons are misleading. The honest comparison isn't between two paper-perfect strategies. It's between the strategy you'll actually execute through a -80% year and the one you'll bail on at -50%.

Where the SMA filter pays its rent

Year-by-year on TQQQ, the contrast is sharp:

Year	Buy-Hold	SMA200 Filter	Difference
2017	+118%	+89%	-29 pp
2018	-19.8%	+5.2%	+25 pp
2019	+134%	+71%	-63 pp
2020	+110%	+99%	-11 pp
2021	+83%	+27%	-56 pp
2022	-79.1%	-34.1%	+45 pp
2023	+198%	+120%	-78 pp
2024	+58%	+41%	-17 pp

The pattern is clear:

• In trending bull years (2017, 2019, 2021, 2023), the SMA filter underperforms because it's slower to re-enter after dips, missing some upside.
• In bear or volatile years (2018, 2022), the SMA filter outperforms by being out of the market during the worst parts.
• Net result over a full cycle: similar long-term return, dramatically lower drawdown.

You're not getting "free alpha" with the SMA200. You're trading some upside in good years for avoiding most of the downside in bad years. The math on that trade is roughly even on Sharpe ratio. The behavioral math heavily favors the lower-drawdown path.

What about the leverage decay angle?

For 2x and 3x leveraged ETFs (TQQQ, UPRO, SOXL, USD), there's an additional reason the SMA filter shines: volatility decay.

Leveraged ETFs rebalance daily. In choppy sideways markets, this rebalancing erodes the fund's value even if the underlying ends flat. The math: if QQQ goes up 2% then down 2%, it ends roughly where it started. TQQQ goes up 6% then down 6%, but starting from the higher base the loss is bigger in dollar terms. Compound that effect daily over a year of chop and TQQQ can lose 10–20% even if QQQ ended flat.

The SMA200 filter sidesteps the worst chop. Because it gets you out below the SMA200 (typically when volatility spikes during downturns), you're in cash precisely when the leveraged ETF is bleeding the most from rebalancing decay.

This is why the drawdown reduction is so much more dramatic on TQQQ (-82% → -34%) and SOXL (-90% → -42%) than on plain QQQ (-35% → -23%). The filter saves you from the brutal decay regime and from the directional loss.

What this means practically

The honest read of 16 years of data:

1. If you can stomach -80% drawdowns without selling, buy-and-hold leveraged ETFs probably wins on absolute return over a long enough horizon. Most humans cannot.
2. If you want a strategy you'll actually execute through a full cycle, the SMA200 filter delivers similar Sharpe ratios with vastly smaller drawdowns. That's the real return you'll see in your account.
3. Neither approach is "right." They're different risk profiles. Pick the one that matches your actual ability to stay disciplined during the worst part of a cycle.
4. The SMA200 isn't magical. It's a behavioral guard rail. The "edge" isn't in the math. It's in keeping you invested when you'd otherwise capitulate.

This is the unsexy truth that doesn't make great Twitter content. Trend filters aren't get-rich-quick. They're "don't blow up while staying invested" tools. That's their whole job.

If you want to run a variant of this with LEAPS instead of straight leveraged ETFs (pulling the leverage out of the calendar and into the option, sidestepping the daily-rebalance decay entirely), the broker matters more than the platform you read signals on. tastytrade is the one I'd consider for that workflow. The analytics and per-leg commission cap are built for options-led portfolios. See the full broker shortlist for the honest tradeoffs.

Caveats worth naming

• This data covers 2010–2026, dominated by a long bull market with two notable drawdowns (2020 COVID, 2022 rate shock). Different regimes (extended bear markets, '70s-style stagflation, lost-decade conditions) might tilt the comparison differently.
• Costs are modeled at 3bp per side. Real spreads on TQQQ/SOXL around the open can be 3–5bp; less liquid times can be wider. Higher costs hurt the filter strategy more than buy-and-hold (more trades).
• We backtested simple "price > SMA200" as a daily signal. There are many variations on this rule: different SMA windows (50, 100, 150), day-of-week filters, volatility-targeted sizing. Some perform better, some worse.
• No tax modeling. In a taxable account, the buy-hold strategy benefits from long-term capital gains treatment and deferred realization. The filter strategy generates more short-term gains. After-tax results favor buy-and-hold by another margin.
• Past performance is not predictive. The next 16 years won't look like the last 16. The relative performance of these strategies depends heavily on the regime ahead.

---

Want today's SMA200 status for any US stock? Type a ticker on the homepage →

Want the simple definition of the 200-day moving average without the data deep dive? Read the about page →