Moving averages: the trend, smoothed (and the golden cross myth)

Smoothing out the noise

Daily prices jump around. A moving average irons out the jumps so the trend underneath is easier to see. The simple moving average (SMA) adds the last N closes and divides by N; tomorrow it drops the oldest close and adds the newest - it "moves". The exponential moving average (EMA) does the same but gives recent days more weight, so it turns a little sooner.

• Short averages (10, 20 days) follow price closely - quick, but jumpy.
• Long averages (50, 200 days) are smooth and slow - they show the bigger trend.

The average as a trend filter

The most useful way to use a moving average is also the simplest: is price above or below it, and which way is the average sloping?

• Price above a rising 200-day average → the long-term trend is up.
• Price below a falling 200-day average → the long-term trend is down.
• Price weaving through a flat average → no trend; averages are mostly noise here.

Many investors use this as a filter: "I only buy stocks above their 200-day average." It sounds sensible. Our NEPSE test, below, has a surprise.

The golden cross and the death cross

When the 50-day average crosses above the 200-day, it's called a golden cross - supposedly a buy signal. The opposite is a death cross. You'll see them in headlines. Think about what has to happen first: the 50-day only rises above the 200-day after weeks of rising prices. By the time the cross prints, much of the move has usually happened.

NEPSE daily data, 238 listed companies; signals from late 2022 (each needs a year of history first) to Oct 2026. Returns from the next session's open to the close 20 sessions later, before costs. Green = better than an ordinary day, red = worse. Past results don't guarantee future results.

• The golden cross did worse than an ordinary day on NEPSE: only 37% of cases were higher 20 sessions later, against 48% for any day. It arrived late, often near a short-term top.
• The death cross wasn't the disaster its name suggests - it also came late, often after most of the fall.
• In this period, stocks below their 200-day did better than stocks above it. NEPSE spent 2022-2023 falling and 2024 rallying hard: buying weakness worked better than buying strength. That won't always be true - but it shows a "rule" like "only buy above the 200-day" depends on the market's regime.
• Whipsaws: the average stock saw about 4.5 golden or death crosses in four years. Each one acted on means costs and a late entry or exit.

Real NEPSE case study · PROFL · 20 May 2024

Real NEPSE case study · SMATA · 28 Jul 2025

How to use moving averages properly

• As context, not a trigger. Use the slope and position to describe the trend; make decisions with other evidence.
• Expect lag. The longer the average, the later it confirms. That's the price of smoothness.
• Ignore crosses in flat, choppy markets - they're coin flips with costs attached.
• Mind adjustments. A bonus-share book closure drags a 200-day average down for months on an unadjusted chart.

See it live on MacroLens NP Stretch → How far each stock and sector sits from its moving averages - and what that meant before.

Check yourself

3 quick questions. Pick an answer for each, then check.

Q1 Why does a 200-day moving average react slowly?

It's broken It averages 200 sessions, so one new day barely moves it Brokers control it It only updates monthly

Q2 A golden cross just happened. What has usually already happened?

Nothing yet A large part of the rise - the cross is late by design A crash A dividend

Q3 What is a moving average best used for?

Exact buy and sell signals Seeing and filtering the trend Predicting earnings Valuing a company

Check my answers

Want your scores saved? Create a free account - it also keeps your place in the course.