Moving averages
A moving average smooths price into a trend line. It is always late by design, and that lag is the trade-off you are buying.
What it computes
A 20-day simple moving average is the mean of the last 20 closes, recalculated each day. An exponential moving average weights recent closes more heavily, so it turns sooner and is noisier.
Shorter lookbacks hug price and generate frequent signals; longer ones filter noise but respond slowly. There is no optimal setting — only a choice about how much lag you accept.
Common uses
As a trend filter: price above a rising 200-day average is a common definition of a long-term uptrend. As dynamic support or resistance: in strong trends price often pulls back to a rising average and resumes. As a crossover signal: a shorter average crossing a longer one marks a momentum shift.
Where they fail
In sideways markets crossovers whipsaw repeatedly, generating losing signals in both directions. A moving average contains no information beyond past prices — it cannot anticipate news, and it will always turn after price does.
Worked example
Computing an SMA and seeing the lag
- Last 5 closes: 20, 22, 21, 25, 27. The 5-day SMA = (20+22+21+25+27)/5 = 23.0.
- Tomorrow closes at 29. Drop the oldest (20), add 29: (22+21+25+27+29)/5 = 24.8.
- Price moved from 27 to 29 (+7.4%); the average moved from 23.0 to 24.8 (+7.8%) but still sits 4.2 below price.
- If price now drops sharply to 21, the average stays above 24 for several days before catching down.
The average is a lagging summary. It is useful for defining trend context, not for calling turns.
Common misconceptions
“A golden cross is a reliable buy signal.”
Crossovers are lagging and whipsaw badly in ranges. They describe a momentum shift that already happened.
“Price must bounce off the 200-day average.”
It is a widely watched reference, not a rule. In downtrends price slices through it repeatedly.
Checkpoint