MacroView · Learn · RSI, SMA and EMA — what each period actually shows
RSI, SMA and EMA — what each period actually shows
Last updated: 2026-09-07
Two very different things sit on a MacroView stock row, and they answer different questions. RSI is a momentum reading: it describes how one-sided the recent move was. Moving averages (SMA and EMA) are just the average price over a stretch of time: they tell you where price sits relative to its own recent history. Neither predicts anything on its own — but together they describe a situation quite precisely.
RSI runs from 0 to 100, and the number has a plain-English meaning most explanations skip. It is the percentage of the recent price movement that happened to the upside:
RSI = 100 × average up-move / (average up-move + average down-move) RSI 71 → ~71% of the movement over the window was upward
RSI Watch screens at 35 / 65 rather than the classic 30 / 70 so a name appears on the list before it crosses the level everyone else has set their alerts on.
The period is simply how far back the calculation looks. RSI(14) covers about three trading weeks; RSI(7) about a week and a half. Halving the window doubles the sensitivity, and that shows up everywhere:
- Speed. Starting from balanced, a run of steady up-days lifts RSI(7) past 70 in 6 sessions and RSI(14) in 12. Exactly double, every time.
- Frequency. On AT&T over five years, RSI(7) crossed above 70 about ten times a year, RSI(14) about five. The 7 spent 16% of all sessions in overbought territory; the 14, only 7%.
- Warning time. RSI(7) got there first — a median of 3 sessions before RSI(14) crossed the same line. That head start is what you pay for with twice as many false alarms.
Which is why the two have different jobs: RSI(14) is for finding names (it is also what every other platform screens on, so it is the comparable number), and RSI(7) is for timing one you already care about. Read together they say something neither says alone — RSI(14) at 71 with RSI(7) at 82 means the whole stretch happened in the last week or so.
An SMA (simple moving average) adds up the last N closing prices and divides — every day counts the same. An EMA (exponential) weights recent days more heavily, so it turns sooner and hugs price more closely. A 9-day EMA puts about 20% of its value on today's close alone.
Every average lags, because an average of the last N days sits in the middle of those days — roughly (N−1)/2 sessions behind. Measured on AT&T, the 200-day SMA runs about 100 sessions behind price. At the same length of 20, the EMA lags 6 sessions where the SMA lags 9. That third less lag is the entire practical difference between the two.
The pattern is the whole point: short lines get crossed constantly and describe the current move; long lines are crossed a handful of times a year and describe the regime. A short line can never tell you about the trend, and a long line can never tell you about this week.
T · close 25.77 value price vs line RSI(7) 81.7 RSI(14) 71.2 EMA 9 25.13 +2.5% EMA 20 24.49 +5.3% SMA 50 23.00 +12.0% SMA 150 25.24 +2.1% SMA 200 25.12 +2.6%
Read in order, that row says four things:
- RSI(14) 71.2 — about 71% of the last three weeks' movement was upward. A one-sided run, and among the top 7% of readings this stock produces.
- RSI(7) 81.7 — the one-sidedness is concentrated in the last week and a half. This is fresh, not a long grind.
- +12% over the 50-day — AT&T normally sits about 5% from that line, and reaches 10% only a tenth of the time. On a three-month view, this is genuinely stretched.
- Only +2% over the 150- and 200-day — where it normally sits 9–10% away. On a longer view the stock has barely regained its averages.
Put together: a sharp recent advance off a long, flat base that has just reclaimed its long-term trend. Stretched on the short horizon, early on the long one. That is a completely different picture from a stock showing the same RSI while trading 30% above every line — and you cannot tell the two apart from the RSI alone. It is exactly why RSI Watch carries the moving-average distances next to the RSI.
"More than 10% above the average is extended" is the most common misuse of these numbers. Typical distance grows with both the length of the line and the volatility of the stock. Over five years, average distance from the 200-day was 10% for AT&T, 32% for NVDA and 5% for Coca-Cola. The same 10% reading is extreme for one, normal for another and unremarkable for the third. Always compare a distance against what is normal for that stock, on that line.
These are descriptions of a condition, not instructions. "Overbought" is not a sell and "oversold" is not a buy. Over the last ten years, a month after an overbought RSI(14) reading, NVDA was up a median 4.9% — better than its own baseline, because that is what a strong trend looks like from the inside — while Coca-Cola was down a median 0.8% and positive only a third of the time. Same indicator, same reading, opposite outcome. What decides it is the trend the stock is in, which is the question the moving averages answer.
See RSI Watch for the live tails and Moving Averages for the lines on a chart. One more detail worth knowing: both are computed on settled closes, so during a session the values match a chart's previous close rather than the live tick — that alone explains most small differences against other platforms.
Wilder's original construction, the 70/30 convention and its limits.
The weighting formula, and how it compares with a simple average.