Know Sure Thing with stacked horizons and trendline confirmation

This archive article walks through a historical workflow for judging trend direction and maturity. Four rate-of-change series are smoothed, weighted, and summed into one Know Sure Thing (KST) composite. Action is withheld until short-term, intermediate, and long-term KST readings agree and a matching price-chart signal appears.

Three concurrent trends in the market-cycle model

A market-cycle model treats three concurrent trends as having different typical durations. The short-term swing lasts about three to six weeks. The intermediate swing lasts about six weeks to nine months. The primary trend lasts about nine months to two years, with a usual span of 12 to 18 months and rare cases as short as six months or as long as three years.

These nested swings run together. The historical workflow asks that they be aligned before a trade is taken. The primary trend is the longest listed swing and sets the preferred trade direction.

Positioning with the primary trend

Positioning with the primary trend and avoiding its terminal phase is presented as the core risk-control idea. Early bull-market participation is described as more favorable than late-stage participation. Counter-trend intermediate swings are described as less favorable than with-trend ones.

How Know Sure Thing is constructed

Know Sure Thing is a weighted, smoothed composite of several rate-of-change series used to judge trend direction and maturity. Rate-of-change is the percentage change of the close versus a prior close, recentered so zero is the midpoint of the oscillator.

KST is constructed by computing four rate-of-change series, smoothing each with a simple or exponential moving average, multiplying each smoothed series by a horizon-specific weight, and summing the weighted terms into one momentum composite.

The supplied short-term daily KST parameter set uses rate-of-change lengths 10, 15, 20, and 30 with moving-average lengths 10, 10, 10, and 15 and weights 1, 2, 3, and 4.

A worked short-term weekly example recenters each rate of change as close divided by the close n periods earlier, times 100, minus 100, then applies exponential smoothing with alpha equal to 2 divided by n plus 1.

Several windows versus same-length oscillators

Using several rate-of-change windows is presented as more informative than stacking same-length momentum oscillators. Same-length series are highly correlated, and a lone short window can miss larger cycles that may reverse first.

Agreement plus trendline confirmation

The teaching rule is to require agreement among short-term, intermediate, and long-term KST readings and a matching price-chart signal. In the illustrated case, a long-term bull reading, a short-term rally reading, an intermediate buy reading, and a later price buy signal lined up.

Trendline confirmation is the price-structure rule that a KST buy or sell is valid only after price breaks or holds a drawn trendline in the same direction.

Unconfirmed reversals in a linear trend

KST can issue unconfirmed reversal signals in strong linear trends. A linear-trend false signal is a strong one-way price path in which the long-horizon KST can flip while price never confirms the reversal. In the S&P 500 example, long-term sell readings were treated as false because price never broke its rising trendline.