Averaging down

A second or later tranche opened in the same direction as an existing position, at a price worse than the one already held. The name describes the visible effect on the average entry price, which moves toward the newer and less favourable price. The practice is old, it has adherents and detractors in equal measure, and the disagreement between them is not about the arithmetic, which nobody disputes.

Three things change at once, and only one of them is the average. The exposure rises, so every subsequent point of adverse movement costs more than it did. The margin committed rises with it, so the free margin absorbing that movement falls, and the distance to a close out shortens from both ends at the same time. On a leveraged position the loss that follows is not limited to the amount deposited. The average entry reports none of that, which is why a practice assessed on the average alone is being assessed on the one number that was designed to improve.

Practitioners divide along a line worth stating plainly. One tradition treats a further tranche as a new position requiring its own reason, its own exit and its own sizing calculation, in which case the existing position is irrelevant to the decision and the word averaging describes an outcome rather than a method. The other treats the run of prices as a single position being built into weakness, which is coherent for an investor holding an unleveraged asset and materially different where a margin close out can end the position before any recovery arrives. A fixed schedule of doubling additions is the extreme case of the second tradition and is treated separately under martingale.