2020issue C036
Variable-strength pivot highs as falsifiable entry filters
A reader asked how published Leavitt Convolution entry and exit rules that cite pivot-high parameters should be written as ordinary pivot-point tests. The archive illustrates variable-strength constructions, a wrapper function, and an open-equity histogram.
- A pivot-point is a confirmed local extreme defined by a chosen price series plus left-strength and right-strength bar counts, used here as an entry or exit condition rather than as a standalone forecast.
- A variable-strength pivot high is constructed like a swing high, except the left confirmation window and the right confirmation window may differ.
- A wrapper function still returns one value while a multiple-output Pivot routine writes the confirmed pivot price and the pivot bar into input-output parameters.
- An open-equity histogram plots open equity minus closed equity so the same pivot rules can be inspected after a position is open.
The question behind the rewrite
A reader asked how published Leavitt Convolution entry and exit rules that cite pivot-high parameters should be written so the same conditions can be expressed as ordinary pivot-point tests.
In this archive note a pivot-point is a confirmed local extreme defined by a chosen price series plus left-strength and right-strength bar counts. It is used as an entry or exit condition rather than as a standalone forecast.
Two illustrated constructions
One illustrated construction is a double pivot high whose inputs are the prior bar's low together with length 3, left strength 1, and right strength 2.
A second illustrated construction is a pivot high whose inputs are the close together with length 3, left strength 1, and right strength 2.
Those inputs specify unequal confirmation windows. Variable-strength pivot highs are constructed like swing highs except that the left confirmation window and the right confirmation window are allowed to differ, so the lookback that defines the peak can differ from the look-ahead that confirms it.
The wrapper and its input-output parameters
The older single-return pivot-high helper is described as having been replaced by a variable-strength wrapper function that still returns one value while calling a multiple-output Pivot routine.
That wrapper function accepts a price series, a length, and independently specified left-strength and right-strength integers. It then returns the confirmed pivot price after the underlying call writes pivot price and pivot bar into output parameters.
Multiple-output pivot functions distinguish read-only inputs from writable input-output parameters. Those arguments are typically initialized to zero, passed through unchanged, and then filled by the function with computed values such as the confirmed pivot price or the bar at which that pivot occurred.
Inspecting the same rules after a position is open
A companion subsurface construction, the open-equity histogram, plots open equity minus closed equity as a histogram and colors the bars by the sign of that difference.
The plot is used so the same pivot rules can be inspected after a position is open. It is not used as a price predictor.
All readings on this track · 38 readings
- 1988Constructing action-reaction lines from two pivots
- 1988Constructing intradaily point-and-figure boxes and pivot ladders
- 1991Constructing layered support and resistance from swings, pivots, and retracements
- 1994Three locks on a day-session order, then a staged exit
- 1994Building a five-level daily pivot grid
- 1996Constructing daily pivot points from session prices
- 1996Higher time frame balance points as a trend and band filter
- 1998Cup-with-handle construction rules
- 2000Pivot levels as a daily trade hypothesis
- 2001Construct a same-session polarity card around the daily pivot
- 2001Trading inside the cup-with-handle before the breakout
- 2005A lower-low rebound as one entry, abstention, and stop routine
- 2006Constructing session pivot maps from the prior high, low, and close
- 2006Constructing a pivot grid for stops and buy-stops
- 2006Monoparametric automatic trendline construction
- 2008Write the exit before the entry
- 2010Dynamic-pivot range grids for trend bias
- 2010Reverse-entry exits for pairs, pivots and support
- 2011Sequencing pairs, futures pivots, and implied volatility
- 2013Constructing Camarilla levels from prior range
- 2013Camarilla levels as a multi-timeframe map of reversion and breakout
- 2013Constructing a camarilla-grid from a completed lookback range
- 2013Constructing daily pivot support and resistance rungs
- 2014Constructing daily pivot levels from prior-session OHLC
- 2014Next-session pivot support and resistance from daily bars
- 2014Constructing session pivot rails from the prior-day range
- 2014Evaluating moving-average, pivot, and support-resistance filters
- 2016Stage a Trailing stop toward a planned target
- 2016Smoothed RSI and full-cut pivots for option-income exits
- 2017Constructing a weekly seasonality pivot scaffold
- 2017Seasonality and pivot points as scenario maps, not forecasts
- 2018Wave pivots, strength filters, and option premium
- 2018Constructing Fibonacci and daily pivot support maps
- 2018Building a daily pivot lattice with Fibonacci rails
- 2019Prior-session pivot channels for same-day entries
- 2019Constructing intraday pivot channels from prior-session levels
- 2020Variable-strength pivot highs as falsifiable entry filters
- 2020A high-volume-pivot long after a multi-week decline