Constructing overlay profiles to map auction value

Build the overlay from session histograms

An overlay profile is constructed by adding successive daily market-profile histograms so accepted value can be read as a multi-session distribution rather than one print.

Session construction starts from a half-hour histogram. Each labeled time bin is a time-price opportunity, so one auction day is partitioned into multiple mini-auctions instead of a single bar.

A market-profile records which prices traded in each successive time bin. That turns one auction day into a stack of mini-auctions that can later be added together.

Read bracketing, trend, and transition on the finished map

On the finished overlay, a bracketing auction appears as one bell-shaped distribution. A trend appears as more than one distribution in different price regions. A transition appears when those distributions blur as prices keep retracing.

A single-distribution overlay marks one accepted value region. More than one hump implies trend or transition. Bracketing is an equilibrium auction in which prices rotate between accepted upper and lower bounds instead of traveling to a new value area.

Confirm with a rotation profile

A companion rotation profile is a day-scale map of high-to-low revisits used to confirm or refute the overlay. Equilibrium shows repeated travel across a bounded range. A trend shows little rotation.

Both the overlay and the rotation profile can be drawn by hand from the same daily market-profile session data.

Prefer time-price overlays over volume

Volume-based overlays can mark trending versus bracketing periods. Later construction prefers time-price-opportunity overlays because one large trade can skew a volume distribution and because time-price counts exist on every auction market.

A volume-profile is a related overlay built from traded size at price. It is useful for locating accepted value, but a single large print can warp the shape, so it is the earlier and noisier cousin of a time-price overlay.

Score rotation from a price-by-time matrix

The rotation index is built from a price-by-time occupancy matrix. The sample range and the sample window are each split into equal bands. A cell is marked if that price band traded in that time band. The raw score is occupied cells divided by all cells.

On a four-by-four matrix the raw occupancy score is 1.0 when every cell is filled and 0.25 on a perfect one-way path. That interval is then renormalized so full rotation reads 1.0 and a complete trend reads 0.

The finished rotation-index is a zero-to-one occupancy score. It is high when every band is revisited and low when trade occupies a one-way path.

Use the close quarter once rotation is weak

A normalized rotation index at or below 0.5 is read as a tendency toward trend. When the index is below about 0.6 the close quartile is used for direction, with a top-quarter close pointing up and a bottom-quarter close pointing down.

The quadrant-of-close is the quarter of the session range that contains the close. It is used as a slower companion alert and a direction cue once rotation is weak.

Session-analysis reads one day, then a short run of days, as an auction: where time was spent, whether price rotated the range, and in which quarter of the range the session ended.