Skip to content

Resource · Public

Moving Averages: Smooth the Chart Without Losing Context

Calculate moving averages on an original price series to compare smoothing and lag, including a reversal where price falls but the longer average rises.

Published

A jagged price sequence and smoother moving average accompany a five-observation average diagram.
Conceptual illustration: a jagged price sequence and smoother moving average accompany a five-observation average diagram.
Image credit

Conceptual illustration: a jagged price sequence and smoother moving average accompany a five-observation average diagram.

FavAcademy · Original commissioned AI illustration; public use authorized

At a glance

Smoothing can make a noisy series easier to describe while hiding recent changes. Calculate three- and five-observation averages from the same original dataset, inspect a sudden reversal, and choose settings as part of a test rather than a search for a perfect line.

01

Start with observations rather than a line

A moving average is a transformation of a price series. It does not add new market observations. A simple moving average, or SMA, takes the arithmetic mean of a chosen number of values, often closing prices. The window advances as each new observation arrives, dropping an older value and including a newer one.

Before adding an indicator, identify the instrument, source, interval, input field, and treatment of missing data. Five daily closes and five hourly closes summarize different periods. A chart label showing only five is incomplete. This lesson uses invented index-like numbers to keep the arithmetic visible and avoid suggesting a current market opportunity.

02

Calculate the first windows

Take the first five closing observations as 100, 102, 101, 105, and 107. The five-observation SMA at the fifth close is 103. The three-observation SMA at the same time uses only 101, 105, and 107, producing approximately 104.33. Both describe the same data through different windows.

The shorter average sits closer to the recent higher observations because it excludes the first two lower values. Neither calculation is more truthful: each answers a different summary question. Keep the raw closes visible beside the averages. A smooth line can conceal the size and order of the movements that produced it.

03

Advance one observation

Suppose the sixth close is 109. The five-observation average now uses 102, 101, 105, 107, and 109, giving 104.8. The three-observation average uses 105, 107, and 109, giving 107. The latest close lies above both averages, but the five-observation window retains more of the earlier lower values.

This is the mechanical source of part of the lag: a new value shares the calculation with older ones. A rising average does not require every individual close to rise. Similarly, an average can continue rising after a small price decline if the observation leaving the window was lower than the new observation.

04

Test a sudden reversal

Now let the seventh close fall to 103. The three-observation SMA becomes approximately 106.33 from 107, 109, and 103. The five-observation SMA uses 101, 105, 107, 109, and 103, producing 105. The longer average rises from 104.8 to 105 even while the latest close falls sharply from 109 to 103.

That is not a calculation failure. The new 103 replaced the old 102 in the five-value window. The total increased by one, so its mean increased by 0.2. Understanding the replacement explains the apparent contradiction without pretending the indicator anticipated or confirmed the turn in real time.

05

Choose a window for a defined purpose

Fidelity's SMA guide describes the familiar trade-off: longer windows produce more smoothing and more lag. A shorter window reacts sooner but also follows more small fluctuations. There is no window length that removes uncertainty. An appealing line on one selected chart does not establish useful performance elsewhere.

Define the intended role before choosing a period. It might summarize direction for a descriptive report or form one condition in a testable strategy. Those are different tasks. If the goal is a trading rule, specify when the value becomes available, what action follows, and how costs and unsuccessful signals are recorded.

06

Inspect crosses without hindsight

A price crossing an average is a relation between current price and a historical summary. It is not evidence that the next movement must continue. In sideways data, repeated crosses can create turnover and costs. Waiting for a completed close may reduce intrabar ambiguity while changing the entry timing; that trade-off needs explicit testing.

Do not use a final candle close to simulate an earlier decision within that candle. Also check whether the chart uses adjusted or unadjusted prices for instruments where adjustments matter. Missing observations and session boundaries can alter the window. A strategy description should contain enough detail for another person to reproduce the indicator values.

07

Keep the comparison honest

Make a table with the observation number, raw close, three-value SMA, and five-value SMA. Leave early values blank until the full chosen window exists, unless a different initialization rule is explicitly declared. Do not replace unavailable history with zeros. Preserve the seventh observation as a counterexample to the idea that a rising average always means rising current prices.

Then evaluate the same predefined calculation on a fresh series rather than repeatedly tuning the period on the original one. Moving averages can organize historical information and make changes easier to discuss. Their usefulness depends on the question and evidence; smoothness alone does not establish predictive skill or profitable decisions.

Risk disclosure

Moving averages summarize past prices and lag changes. Period choices and historical crosses do not guarantee useful future signals.