2017issue C0335
Pair-spread risk as a unit-by-unit capital sequence
Pair and spread books can be built in successive layers so the outcome does not depend on a single entry. The next unit of capital is then a preplanned scale-in inside a stated envelope, a last-in-first-out-cut that frees a layer, or a stop-loss that bounds a leveraged tail.
- Layered-spread-capital builds a pairs-trading book in successive units across a chosen range so the outcome does not rest on one entry.
- A mean-reversion-outlier can be a stronger planned scale-in-position and a faster growing loss at the same time.
- A size-relative-stop can treat a very small book as less in need of a stop-loss, but leverage makes some form of stop the general rule.
- A last-in-first-out-cut frees the newest layer of a repeatable-range-spread so the same relationship can be rebuilt at a more favorable reading.
A pair book is a spread, built in layers
Pairs trading, as used here, is a hedged long-versus-short relationship between two symbols, managed as the relative spread rather than as a single-name directional bet.
Pair and spread books can be built by placing capital in successive layers across a chosen range, so the outcome does not depend on a single entry being correct. Layered-spread-capital is that account slice: funds sliced into successive units so the pair or spread can be built across a range instead of as one concentrated entry.
A planned add is not an open-ended chase
Scale-in-position is the add itself: successive units of a pair or spread as the relationship moves through a preplanned range, only up to a stated capital maximum.
When the two symbols in a pair keep drifting apart, that wider relationship can be treated as a stronger mean-reversion setup than the original entry if those paths were mapped before the first unit was placed. The farther reading is a mean-reversion-outlier: it is treated at once as a more extreme reversion setup and a larger loss path.
An extreme spread reading creates a conflict: the same outlier that improves the prospective entry also signals a risk of rapidly growing loss. Editorial reading: do not treat the wider reading as an automatic add. Name whether the next unit is still inside the preplanned envelope, a recycle of the last layer, or a bound on the tail.
Stops follow size, edge, and leverage
Whether a pair needs a planned stop depends on position size versus account capital and on the statistical edge; a very small size can make a stop appear unnecessary. That choice is a size-relative-stop: use or skip a stop based on how large the position is versus account capital and the perceived statistical edge.
When leverage is present, some form of stop is treated as the general rule. A stop-loss is a rule that exits or reduces a unit when loss, leverage, time, new information, or a change in the trade’s validity reaches a bound set before or during the position.
A one-time pair is not a repeatable range
A unique, one-time pair setup is treated differently from a setup that can be repeated many times over a defined range. A repeatable-range-spread is a pair pattern expected to recur often enough that capital can be worked in layers across a defined band rather than treated as a single event.
Repeatable, pattern-based pair setups are described as less stop-dependent when capital is divided into layers and deployed only up to a preplanned maximum.
Cut the newest layer to rebuild it
One recovery sequence removes the most recently added layer of a pair that is moving against the book, then attempts to rebuild that layer at a more favorable spread. A last-in-first-out-cut is that removal: the newest pair layer goes first so that capital can be reserved to rebuild the same relationship at a more favorable spread.
All readings on this track · 18 readings
- 1989Measured range, scale-in ladders, and spread replay as a cash-reserve design
- 1990When result pressure rewrites checklist and scale-in rules
- 1992Split-sample floors, ceilings, and standing aside
- 2006A swing operator's scale-in, structure, and cash-floor procedure
- 2011Scale in currency risk with tenth-size futures
- 2012Nested cycle scale-in with rewritten stops
- 2013Volume-confirmed swing breaks with scale-in and trailing stops
- 2014When volume thrusts still require a session procedure
- 2015Fusion entry and scale-in after broken levels
- 2015Volatility and volume gates for breakout entries
- 2015Scale-in and prior-range breakout management
- 2017A pair as a two-clock book
- 2017Pair-spread risk as a unit-by-unit capital sequence
- 2018Moving-average breakout with delayed scale-in
- 2018Steepening price-channel breakout with delayed confirmation and planned adds
- 2018A scale-in needs a gate or it is leverage
- 2020Two-stage longs: golden-cross starter, then cup-breakout add
- 2020Close-gated scale-in for swing breakouts