Seven lines from three numbers

S1, S2 and S3 are supports below a central pivot, and R1, R2 and R3 are resistances above it. All seven lines come from one session's high, low and close.

The central pivot, usually written P, is the average of those three prices. R1 and S1 sit closest to it. R2 and S2 sit one full day's range away from P. R3 and S3 sit further out still.

They're often called floor pivots because exchange floor traders used them to get quick levels before the open. You only needed yesterday's three numbers and a pencil.

For day trading, you compute them from the prior day. Swing traders use weekly or monthly versions, built the same way from the prior week or month. Other variants exist, such as Fibonacci, Camarilla and Woodie pivots. Each one uses a different formula, so the same day gives you different lines depending on which you pick.

Start with yesterday's high H, low L and close C. The pivot is P equals H plus L plus C, divided by 3. Then R1 is 2P minus L, and S1 is 2P minus H. R2 is P plus the range, H minus L. S2 is P minus the range. R3 is H plus twice P minus L. S3 is L minus twice H minus P.

Say a stock had a high of 452.80, a low of 446.20 and a close of 450.10. The three add up to 1,349.10, so P is 449.70.

Twice P is 899.40. R1 is 899.40 minus 446.20, or 453.20. S1 is 899.40 minus 452.80, or 446.60. The range is 6.60, so R2 is 456.30 and S2 is 443.10.

R3 is 452.80 plus 7.00, or 459.80. S3 is 446.20 minus 6.20, or 440.00. A quick check: R3 should equal R1 plus the range, and 453.20 plus 6.60 is 459.80.

A pivot level knows three prices from one day. It doesn't know where buyers stepped in last Tuesday. It can't see a gap from two weeks ago either.

That means a pivot can land in empty air. In the example, R2 at 456.30 might sit in a stretch where price has never paused. Nothing in the chart says anyone cares about that price, other than the formula.

The formulas also disagree with each other. Classic, Fibonacci and Camarilla pivots give different lines for the same day. If the level moves when you switch settings, it isn't telling you much about the market.

They reset every morning too. Yesterday's R1 vanishes and a new one appears, even if the old price was where the market actually turned. Everyone computes the same numbers, too. A pivot that does hold may have a crowd of stops sitting just beyond it.

Read the chart instead of the formula, then see where the two agree. Start at the current candle on the daily, 4-hour and 1-hour charts. Walk left, one candle at a time, and note the highs and lows price would have to clear next.

Give extra weight to the extremes of reversal patterns, like a hammer's low or an engulfing pair's high. Mark gaps as boxes, edge to edge. These are prices where the market actually did something.

Now lay the pivots on top. In the example, yesterday's low was 446.20 and S1 is 446.60. If a 4-hour hammer also bottomed at 446.35, those three prices form one zone, from 446.20 to 446.60. That's a level worth respecting.

If R2 at 456.30 has no candle extreme anywhere near it, treat it as a reference line, not a level. How near is near enough depends on the stock's price and daily range. It's a judgment call, so be consistent with it.

Keep pivots as a second opinion. Only trade a pivot when it lines up with a real candle extreme. You'll take fewer trades, at better prices.

Don't trade inside a zone. When S1 and a prior low sit 40 cents apart, the space between them can't pay. Wait for price to get clearly above or clearly below the whole zone.

Use the next real level as the target, not the next pivot by default. Then check the reward is at least twice the risk before you enter. Say the S1 zone is your stop area and the pivot at 449.70 is the next real level above. Entry at 447.10 with a stop under 446.20 risks 0.90, and a target at 449.70 pays 2.60. That works.

After the close, look at which lines price actually reacted to. Note it in your journal for a few weeks. You'll learn quickly whether the pivots are adding anything for the stocks you trade.