Indicators that matter — course checkpoint

You began this course with a menu of a hundred confident lines. You finish it with about eight formulas, four questions and three permanent limitations — which is a far better inventory, because you can now say what each line does and where it fails.

The formulas, in one place

• SMA(n) = sum of the last n closes ÷ n
• EMA: k = 2 ÷ (n + 1); EMA = Close × k + EMA_previous × (1 − k)
• RSI = 100 − 100 ÷ (1 + RS), where RS = average gain ÷ average loss over 14 Wilder-smoothed sessions — equivalently, the share of recent movement that was upward
• Stochastic %K = 100 × (Close − Low₁₄) ÷ (High₁₄ − Low₁₄); %D = 3-period average of %K
• MACD = EMA(12) − EMA(26); signal = EMA(9) of MACD; histogram = MACD − signal
• Bollinger Bands = SMA(20) ± 2 population standard deviations of those 20 closes (divide by 20, not 19 — the sample form most tools default to draws bands 2.6% wider)
• True Range = max(High − Low, |High − Close_prev|, |Low − Close_prev|); ATR(14) = Wilder-smoothed average of it
• OBV = running sum of ± volume by close direction; VWAP = Σ(price × volume) ÷ Σ(volume)

The four questions

Trend indicators ask which way has price been leaning? Oscillators ask how fast and how one-sided have recent moves been? Volatility indicators ask how far does this thing travel? Volume indicators ask how many shares backed it, and at what prices? One indicator per question is a complete chart. Everything beyond that is decoration.

The three limits that never go away

1. Lag. Every indicator is computed from bars that already printed. A 10-day average needs ten sessions to fully absorb a jump, and its centre of mass sits about (n − 1) ÷ 2 bars in the past. Smoother always means older; the only choice is where on the dial you sit.
2. Redundancy. Williams %R is exactly stochastic %K minus 100. A MACD zero-line crossing is exactly a 12/26 EMA crossing. Five agreeing momentum indicators are one opinion printed five times — they raise confidence without adding information.
3. Parameter fragility. 14, 12/26/9, 50 and 200 are conventions, not constants. Sweeping thousands of combinations guarantees a beautiful winner even in random data: test 200 variants that each have a 5% chance of looking impressive by luck and you should expect about ten impressive results from nothing.

The vocabulary that keeps you honest

"Overbought" and "oversold" are conclusions disguised as labels. RSI at 78 means the last fourteen sessions were strongly one-sided — and in the strongest trends that reading persists for weeks. A band touch means today's close is two standard deviations above its own 20-day average. A golden cross means the 50-day average rose above the 200-day. Every one of those sentences is checkable, and none of them tells you what to do. That is the point.

The two habits that outlast this course

1. State the question before you add the line. If you can't say what a pane answers, delete it. Charts get better by subtraction far more often than by addition.
2. Count the failures, not just the successes. Define a touch, a hold, a signal before you look, then tally every instance including the ones that went straight through. Hindsight marks the bounces and leaves the misses unlabelled — that asymmetry, not the indicator, is what makes folklore feel true.

Before you sit it

Each of these is a minute at your desk. Any one that is not names the lesson to reopen first.

• Say why every indicator lags, in terms of what goes into it — Why does every indicator lag the price?
• Compute a three-day simple moving average from five closes in your head — How does a simple moving average actually work?
• Say what RSI is the ratio of — How is RSI calculated, and what does it actually say?
• Say what ATR measures that a percentage move does not — How do you measure how much a stock moves on a typical day?

Try it now

1. Start from price and volume only — the chart below is exactly that, and nothing else. Write down, one at a time, the questions you would add a line to answer, and stop when you run out of questions rather than of indicators. Most people stop at three.

2. For each question you wrote, write the formula from memory and then check it against a drawn line. The same year is below with a 20-day Bollinger band and a 14-day RSI, both computed on adjusted closes, as the legend under the chart says; pick one bar and see whether your formula lands on it. Anything you cannot reproduce, you do not yet own — and write the window beside every value, because a number without its setting cannot be checked by anyone, including you next month.

3. Sit the checkpoint. Then on to Patterns & Signals, and after that Risk Management for Traders — where the honest limits you learned here become position sizes and stops.

Nothing in this course was a recommendation to buy or sell anything. You learned what a handful of formulas compute, what traders watch, and how often each one misleads — which is a reading skill, not a signal.