What it measures
The stochastic oscillator asks one question: where did price close relative to its range over a chosen window? With the standard 14-period setting, it compares the latest close to the highest high and lowest low of the last fourteen bars and expresses the answer as a percentage. A reading of 80 means price closed in the top fifth of its recent range. The two lines, %K and %D, are that value and its moving average. Nothing in the formula knows about trends, volume, or news — it is a position-in-range gauge.
How traders read it
The textbook signals are the cross and the zone. A %K crossing above %D near the bottom of the range is treated as an entry hint; the reverse cross near the top, as an exit hint. Readings above 80 and below 20 are called overbought and oversold. Used carefully, the oscillator helps with timing inside a range: buy when price is pressing the range low and momentum turns up, sell when it presses the high and momentum turns down. That is the whole honest case for the tool.
The failure mode
In a strong trend the stochastic is wrong for weeks. Price can close in the top fifth of its range for an entire advance, pinning the lines above 80 while every early exit leaves money behind. Overbought does not mean "about to fall"; it means the trend is strong enough to keep closing near its highs. This single misreading costs more than the indicator ever earns. The countermeasure is a filter, not a tweak: only take the crosses in the direction of a higher-timeframe structure, or skip the signal entirely when the range itself is expanding.
One worked example
On an hourly chart, price sits between 1.0850 and 1.0900 for two days. The oscillator falls below 20 as price tags 1.0855, then %K crosses above %D. With a stop under the range low and a target near 1.0900, the trade risks about 15 pips to make 40 — a workable range trade whether or not it wins. Now fast-forward to a breakout week: the same 80-plus reading appears at 1.0930 and price runs to 1.1100. Same signal, opposite context. The indicator did not change; the regime did.
Settings and honesty
The 14-3-3 default is a convention, not a law. Shorter windows produce more crosses and more noise; longer windows produce fewer signals that arrive later. Slowing the %D smooths the line but delays the cross. Every adjustment is a trade-off between getting in earlier and being whipsawed more often. Whatever you choose, write it down and test one market at a time — the tuning that flatters last month's chart is the one that will be blamed next month.
What it cannot do
The oscillator does not measure trend strength, predict reversals, or account for the news that causes them. It is a timing aid inside a plan that already defines the setup, the stop, and the target. Used that way, it earns its screen space. Used as a standalone system, it is a coin flip with extra lines.
Download the indicator files
⬇️ Download pipatlas-stochastic-oscillator.mq4 (MT4 · MIT)
⬇️ Download pipatlas-stochastic-oscillator.mq5 (MT5 · MIT)
⬇️ Download pipatlas-stochastic-oscillator.pine (TradingView · MIT)
Install: MT4 / MT5 → File → Open Data Folder → MQL4/MQL5 → Indicators → copy the file in → open MetaEditor and press F7 to compile → drag it onto your chart. TradingView → Pine Editor → paste the file “Add to chart”.
Frequently asked questions — Stochastic Oscillator Explained
Which settings are standard?
14 periods for %K, 3 for slowing, 3 for %D. They are the default because they are old, not because they are optimal — change them only with a test behind the change.
Should I trade every overbought or oversold reading?
No. In trends those readings persist. Treat them as context and wait for a cross plus a price level that makes the risk worth taking.