A case study of checklist-driven opening-auction and pairs work

Opening-auction entries

One opening-auction procedure staged opening-only orders across hundreds of names and several price layers. An opening-only order is a limit meant to participate only at the open. If it does not fill there, it does not work the rest of the session. A fill was allowed only when a gap exceeded a predefined gap criterion, meaning the opening print had to clear a set distance before entry was permitted. Price layers were stacked limits on the same name so a gap could fill at more than one level.

That same opening procedure scaled out of fills with a second person tracking results. Remaining risk was typically flattened within about a half hour after the opening bell. Accepting losses promptly was described as a central control in this procedure, not an afterthought.

A second opening procedure researched hundreds of names for post-gap retracement patterns by time of day. Those patterns were used to hold some names for hours or until the close.

An automated opening variant

An automated variant staged orders on about 50 names at two or three price layers. Premarket bids and offers were set from futures and that day's fair value. After a fill, the procedure automatically sent a retracement order for half the position a small increment from the opening price. The retracement order left the remainder to a discretionary close.

Focus after the open

Focus choices ranged from opening-only work, to repeating the same one to three names daily, to preselecting names expected to move the next session. Repeating one or a few names is the surrogate specialist practice: trading those names again to learn their auction behavior, without exchange specialist obligations.

The largest group combined three parts of the day. Opening-only work occupied the first 15 minutes. A personal focus list, a small set of names studied in advance, was worked through the regular session after the open. The last 20 minutes were the market-on-close imbalance window, used to work published close-auction imbalance information.

Optional correlated pairs

A correlated pairs procedure was described as market-neutral and research-dependent. It mixed technical, fundamental, and timing inputs. Horizons ran from same-day trades to longer-term books. Entries and exits sat at predetermined researched levels.

Controls that sit on every piece of the day

Participants defined success as increasing size after demonstrated confidence and taking what the session offered. They rejected a fixed dollar target because those targets led them to quit early on productive days and stay too long on unproductive ones.

Respondents said losses should be owned in full. Journals were used to correct recurring errors. A contingency plan, a written response for operational glitches or abrupt market events, kept the next action decided before those events occurred. Losses were not to be attributed to external factors.