Swing trading settings
Short answer
No moving average is best for every trend. A 20-bar average is a shorter-horizon candidate, a 50-bar SMA is a common intermediate trend reference, and a 200-bar SMA tracks a broader regime. EMA weights recent prices more heavily than SMA, so it responds faster but can change direction more often. Match the length to the trend you want to follow, define price position and slope separately, then compare the complete strategy after costs on data kept out of tuning.
Treat each value as a starting point. A setting only becomes useful after you test it on the same market, timeframe, costs, and date range you plan to study.
| Parameter | Starting point | What it changes | Caution |
|---|---|---|---|
| SMA or EMA | Compare both at the same length | SMA weights each price in its window equally. EMA gives recent values more weight and generally reacts sooner. | Faster reaction can help after a turn but also creates more short-term changes. It is not evidence of a better trading result. |
| Shorter trend reference | 20 bars, an illustrative baseline | Tracks a shorter segment of price history for swing timing or a fast trend filter. | It can flip repeatedly during a range. Twenty bars means twenty days only on a daily chart, not on a four-hour chart. |
| Intermediate trend reference | 50-bar SMA | Provides a slower direction filter that can be paired with a separate swing entry trigger. | A rebound can put price above the average before the average turns upward. Specify whether your rule needs one condition or both. |
| Longer trend reference | 200-bar SMA | Provides a broad regime reference rather than a fast entry signal. | More smoothing delays recognition of turns. Use the same evaluation dates after all tested averages have enough prior history. |
| Slope window | Compare with five bars ago, for illustration | Turns a rising average into a rule: current average greater than the average five completed bars earlier. Equality fails this positive-slope condition. | A longer comparison window can keep the slope positive after a recent turn. Test the window separately from the average length. |
| Price position | Completed close above the average | Defines a long-side permission condition. The short-side mirror requires the close below the average and, if used, negative slope. | Price above the average does not establish that the average is rising. Requiring both checks may exclude trades and delay entries. |
| Source and confirmation | Close source, zero plot offset, completed bars | Keeps the input series and decision time consistent across comparisons. | The current close and its average can change intrabar. A plot offset only shifts the displayed line; it does not remove calculation lag. |
Compare 20 SMA and 20 EMA as the trend reference
Use when responsiveness is the hypothesis. Track how often each filter changes state during a range and whether earlier exits offset the added turnover.
50 SMA with price-position and five-bar slope checks
Use when the entry rule should ignore some short swings. Compare price-only permission with price plus slope while keeping the entry and exit rules identical.
200 SMA as a filter for a separate swing entry
Use when broad direction matters more than early reversal entries. Review missed recoveries as well as drawdown; a slower filter can stay bearish after price has turned.
These templates define a research process. They are not trade calls or evidence that a setup will make money.
Rule to test
When flat, enter long after a completed close crosses above a 20-bar average. Exit after a completed close crosses below it. Compare SMA and EMA with the same fixed position size, no pyramiding, and next-bar market orders.
This isolates average type. Record fees and slippage explicitly. Keep an untouched later date range, then compare net average trade, drawdown, and turnover there without changing the selected settings.
Rule to test
Use a close above the highest high of the previous ten completed bars as the long entry trigger, and a close below the lowest low of the previous five completed bars as the exit. Compare no filter, close above 50 SMA, and close above 50 SMA with its value above five bars ago. Keep sizing fixed and allow one position.
This isolates the filter contribution. Predeclare the trend and range periods to inspect, include all of them, and compare exposure and missed trades as well as drawdown. The breakout windows are illustrative choices.
Rule to test
Keep the ten-bar breakout entry, five-bar exit, and price-plus-slope filter from the previous test. Compare SMA lengths 20, 50, and 200 with the same five-bar slope window, sizing, and execution settings.
This isolates trend horizon. Start only after the longest average and slope window are available. Select on the training period, freeze the choice, and evaluate later data. Inspect nearby lengths for sensitivity instead of declaring the highest historical profit universally best.
Define slope numerically before testing. An illustrative rising filter requires the average today to exceed its value five completed bars ago. A line that looks steep on a zoomed chart is not a reproducible condition.
Price above a rising average satisfies both direction checks. Price above a falling average may be a rebound within a decline; price below a rising average may be a pullback. A flat average with repeated price crossings is a warning that the filter may generate frequent reversals.
Illustrative example: a 50-bar SMA rises from 100 to 101 over five bars, but price closes at 100.5. Slope is positive and price position is negative. A rule requiring both conditions rejects the long setup. These numbers describe a hypothetical case, not a historical trade.
A trend filter grants permission to look for an entry; it does not define an entry, stop, or position size. For a swing strategy, keep those rules fixed while comparing filters so a better-looking equity curve has an identifiable cause.
Measure whipsaw costs: record trade count, average net trade, exposure, and drawdown in preselected sideways periods. A faster filter can leave a losing position sooner and still perform worse if repeated exits and re-entries consume the gain through fees and slippage.
Use the Pine Script AI Coding Agent to draft an editable strategy with inputs for average type, length, slope window, and a filter on/off switch. Review its rules and load it into TradingView. Pineify Strategy Optimizer can then compare exposed input combinations through the Chrome extension. Keep the test dates, costs, and order settings consistent; reserve a separate period for a later test without retuning. The generated script and optimization results still need review.
Build a moving-average testSources checked 2026-10-16
This page is educational, not investment advice. Parameters and test templates are illustrative research choices with no performance results claimed. Backtests depend on data and execution assumptions and do not predict future returns. Trend filters can delay entries, miss recoveries, and produce repeated losses. Trading can result in loss.
