SMA Indicator: Best Settings & the 200 SMA Signal 2026

The SMA indicator is the plainest tool on any chart. You add up the last N closes and divide by N, and that's the whole calculation. Yet the 200-period version gets more attention than almost any other line in trading, and most of what's written about it mixes two very different jobs. One job is a filter: it tells you which side of the market you're allowed to trade. The other is a trigger: it tells you to buy or sell right now. The published evidence treats those two jobs very differently, and this guide keeps them apart from start to finish.

Short version up front: as a regime filter, the 200 SMA has some third-party support behind it. As a standalone entry signal, the numbers people quote actually point the other way. Every figure below comes from a named outside publisher. Quantzee has not run its own 200-period moving-average sweep, so nothing here is presented as our measurement.

What the SMA indicator actually calculates

A simple moving average weights every bar in its window equally, so a 20-period SMA on a daily chart takes the last 20 closing prices, sums them, and divides by 20. Tomorrow the oldest close drops out and the newest one comes in, so the line moves one step at a time. For the formal definition and the variants (weighted, exponential, cumulative), the moving average entry on Wikipedia covers the maths cleanly. The SEC's investor education site also keeps a short moving average glossary entry if you want the plain-English version.

If you want the broader family first, our moving averages glossary page explains SMA and EMA side by side. This post goes deeper on one question: what the SMA can and can't do once you push its length out to 200.

Common SMA settings and what each one is for

There is no single right setting, because each length answers a different question, and the useful way to pick one is by the time span it covers.

The span column assumes about 252 trading days a year, which is the usual count for US equities. The lag column needs its own section.

The lag trade-off, measured in bars

An SMA's centre of mass sits in the middle of its window. For an N-period average, that point is (N − 1) ÷ 2 bars back. So a 50-period SMA describes, on average, where price was about 24 bars ago. A 200-period SMA describes where price was roughly 100 bars ago. On a daily chart, that's close to 5 months.

That number explains almost everything about the 200 SMA: it can't call a turn early, and it's built not to. By the time the line itself flattens and rolls over, price has usually been moving the other way for weeks. If price drops 15% in 3 weeks, the 200-day line might move only a percent or two, because 185 of its 200 inputs are older closes that haven't changed.

This is a feature for one job and a flaw for the other, since a slow line is good at describing a state ("the market has been mostly higher over the past 10 months"). It's bad at timing ("buy today"). Hold on to that split, because the rest of the evidence lines up with it.

The 200 SMA signal: two jobs, kept apart

When people say "the 200 SMA signal," they mean one of two things, and they rarely say which.

Job 1, the state filter. Price closes above the 200-day SMA, so you only take long setups. Price closes below it, so you stand aside or only take shorts. The line never triggers a trade on its own; it just narrows the menu.

Job 2, the entry trigger. Price crosses above the 200-day SMA, so you buy. Or price pulls back to touch the line and bounces, so you buy the bounce, and the line is the signal itself.

These two uses can share a chart and still behave nothing alike. A filter can add value even if every crossing is noise, because it only has to be right about the broad state most of the time. A trigger has to be right at the specific moment it fires, and that's exactly where a 100-bar lag hurts.

The regime evidence: what holds up, and what's missing

The strongest number on page one of search comes from a trading-tool vendor. According to StockAlarm's moving-average guide, "a widely cited study using S&P 500 data from 1929 through 2020" found returns of roughly 14% a year while the index sat above its 200-day average, against roughly -6% a year while it sat below. That's a 91-year span and a gap of about 20 percentage points.

Now the honest part: the page doesn't name the study, and it doesn't disclose a methodology. It also switches between the S&P 500 and the SPY fund for the same figures, even though SPY only launched in 1993. We can't tell whether the figures include dividends, whether the state was judged on the same close as the return, or how trading costs were handled. Treat it as a description of a market state reported by a third party, not as a tested strategy result.

Even taken at face value, the number says something narrower than it seems to. It says that historical periods above the line were better on average than periods below it, but it doesn't say you could have captured that gap by switching in and out, because every switch happens late, and some happen many times in a few weeks. A state statistic and a strategy return are two different objects: the first is easy to quote, while the second needs a full rule set, a cost model, and a test.

That's why the filter use is the one with support. A filter only borrows the state reading. It doesn't promise that crossing the line is a good moment to act.

The entry evidence runs the other way

The second-strongest number on the results page is, read closely, evidence against using the line as a trigger. LuxAlgo's 200-day guide reports that a study by Quantified Strategies found that adding an above-200-day filter to a short-term RSI system raised the win ratio from 76% to 81%. In the same study, compound annual growth fell from 8.7% to 7.1%.

So the win rate went up by 5 points and the return went down by 1.6 points a year. That pattern isn't a paradox, and it shows up whenever a filter removes trades instead of improving them. If the filter throws out a batch of trades that included some big winners, the remaining trades can win more often while the account grows more slowly. Fewer trades and less time in the market add up to a smaller total. LuxAlgo itself flags that this is one specific RSI strategy, and that a higher win rate can sit alongside a lower compounded return. The study's full rules weren't disclosed on either page, so we can't check the test ourselves.

The lesson for a trader is simple: win rate is the wrong scoreboard for a filter. Compare net compounded return and drawdown over the same dates, with and without the filter. Our equity curve simulator guide shows why a single backtest path hides how bad the drawdown can get, which matters even more once a filter cuts your trade count.

The bounce claim

Tradingsim reports that bounces off the 200-day SMA give a higher win-loss ratio than breakouts, and that "the market only trends impulsively about 20% of the time." Neither claim comes with a sample size, a date range, or a source. They might be right, but as published there's no way to check them. Tradingsim's own reasoning is that a bounce lets you set a tight stop, which can lift a win-loss ratio without telling you anything about expectancy after costs.

Whipsaw: why the 200 SMA fails as an entry signal

The mechanism behind the entry failure has a name: whipsaw. When price chops sideways near a slow line, it crosses back and forth, and each cross fires a signal that usually reverses within days. The result is a run of small losses that are each cheap and together expensive.

Picture a market drifting in a 4% range for 6 weeks while the 200-day line runs flat through the middle. A cross-above rule might buy 5 times and sell 5 times in that stretch. If each round trip loses 1% to the reversal plus costs, that's about 5% gone before any trend shows up. A trend-following rule needs its winners to cover those clusters, and the 100-bar lag means it joins every winning trend late.

Here's the gap on page one: none of the ranking pages counts how often this happens. Nobody publishes how many crossings a year a given market produces, or how long a typical crossing lasts before it reverses. That's the number that decides whether an entry rule can survive, and it differs a lot by instrument. A broad equity index, a single growth stock, and a currency pair will each give a different answer.

In our experience building indicators, this is where most simple-signal ideas quietly die. When we ran the sweep behind our SuperTrend settings guide, choppy stretches drove most of the damage on every setting. That's a different indicator, so don't read it as moving-average evidence, but the pattern it shows, though, is common to trend tools in general: the losses cluster in sideways markets.

SMA vs EMA at 200 periods

An exponential moving average puts more weight on recent closes, so it reacts faster. At 200 periods, a 200 EMA turns sooner than a 200 SMA after a sharp move. If speed were the only goal, the EMA would win.

It isn't the only goal. The simple version matters more here because it's the line most people watch. Charting platforms default to it. Financial news quotes "the 200-day moving average" and nearly always means the simple one. When large numbers of traders and funds watch the same level, orders tend to cluster around it. That's a reflexivity argument, not a performance one, since the line matters partly because people believe it matters.

Two practical points follow. First, if you use the 200 SMA as a filter, small differences in when it flips don't matter much, so the slower SMA costs you little. Second, if you're testing a rule that depends on reaction around the line, test the SMA, because that's the one the crowd reacts to. Swapping in an EMA may test a level that fewer people are watching.

Golden cross and death cross, handled honestly

A golden cross is when the 50-day SMA crosses above the 200-day SMA, and a death cross is the reverse. Both are derived signals: they come from two averages, so they inherit the lag of both. Because the 50-day line lags about 24 bars and the 200-day line about 100, a cross often confirms a move that's already months old.

StockAlarm's guide lists 8 golden crosses in SPY from April 2003 to July 2023 and 7 death crosses from December 2007 to November 2023. Its own verdicts on the death crosses are the revealing part: 1 "true signal" (December 2007), 1 "mixed signal" (March 2022), and 5 "false signals" (October 2011, August 2015, December 2018, March 2020 and November 2023). The body text adds that only the 2007 cross led to a long, severe decline. So by the publisher's own reckoning, 5 of 7 death crosses in that table were false alarms. No method is given for how a signal was labelled true or false.

That's a small sample, and 7 events over 16 years can't prove anything either way. It's still enough to retire the idea that a death cross is a reliable sell trigger on its own.

What you can measure yourself in half an hour

Our team doesn't hold a 200-SMA sweep, and our data doesn't cover this indicator, so we won't hand you results. What we can give you is a test setup you can run on your own instrument with free daily data and a spreadsheet or a short script. Here's the methodology, step by step, for anyone willing to spend half an hour on it.

1. Get the data. Download at least 10 years of daily closes for the instrument you actually trade. Use adjusted closes if dividends or splits apply.
2. Build the line. Add a column with the 200-day simple average. The first 199 rows will be blank; drop them.
3. Flag the state. Add a column that reads 1 when the close is above the SMA and 0 when below.
4. Measure share of time above. Average the state column. This tells you what fraction of days a filter would have let you trade long.
5. Count crossings per year. Mark every row where the state changes from the row before. Divide the total by the number of years. This is your raw signal count if you used crossings as entries.
6. Measure how long crossings hold. For each crossing, count the days until the next one. Sort the list. Note how many crossings reversed within 5 days, within 20 days, and how many held longer than 60 days.
7. Compare against a baseline. If you go further and test returns, compare the filtered rule to simple buy-and-hold over identical dates, with costs included, and look at drawdown as well as return.

Step 6 is the one that matters most. If more than half the crossings on your instrument reverse inside 5 days, a cross-based entry rule is mostly paying for whipsaw. If most crossings hold for months, the line behaves more like a usable trend marker on that market. The answer depends on the instrument, which is exactly why a single quoted figure from someone else's index can't settle it for you.

If you'd rather test on a chart than in a spreadsheet, our guide on how to backtest a TradingView indicator walks through the setup. Before you trust any rule that comes out of it, run it through the stress-testing checklist: shifted dates, higher costs, and a look at the worst year, not just the average one.

Best SMA settings: a practical answer

So which settings should you use? Pick the length by the job it has to do, not by the indicator's popularity.

• For a long-term state filter, the 200-day SMA is the standard choice, partly because it's so widely watched. Use it to decide which side you trade, not when.
• For an intermediate trend read, the 50-day SMA is a common middle step, about 24 bars of lag.
• For timing entries, don't lean on a single slow average at all. Use a faster tool built for timing, and let the 200 SMA act as the filter above it.

Then check the result on your own market with the steps above, because a setting that works on a broad index can behave very differently on a volatile single stock.

Quantzee builds TradingView indicators that pair a regime read with separate entry logic, instead of asking one line to do both jobs. If that split is how you already think about the 200 SMA, it's worth looking at how our tools handle it. Treat every indicator, ours included, as analytical software to test, not as advice.

For the professional standard on technical analysis as a discipline, the CMT Association sets the curriculum many analysts train on.

Frequently Asked Questions