Volume-scaled member versus public short sales as a two-class classroom test

A two-class classroom test

Market-breadth here is not a single-stock short-sale setup. It is a market-wide participation condition: after one published exchange short-sale total is split, each class is judged as unusually light or unusually heavy.

The classroom test is strict. Name the class being faded. Say how volume-scaled-shorting defines heavy or light. Then say which calendar episodes produced a joint signal, a one-sided signal, or no signal.

Split one total into two classes

Market-participant-analysis starts with the published exchange short-sale total. Member-short-sales are the exchange-member component obtained by adding specialist, floor-trader, and other-member shorts. Public-short-sales are the residual non-member component left after those member shorts are removed from the total.

Once both classes exist, each can be scored on its own pessimism and compared with the other. The split is what makes that comparison possible.

Volume scaling defines heavy or light

Showing each class only as share-of-short-interest conceals whether shorting is light or heavy. Share-of-short-interest is a relative split of member versus public shorts. It compares who is shorting, not how intense the shorting is.

The case prefers volume-scaled-shorting. Each class is expressed as a share of that day's trading volume so intensity stays visible across busy and quiet sessions. Once both classes are volume-scaled, they can be compared to see which participant group is doing the shorting and how pessimistic each group appears.

Two readings that can disagree

The case treats elevated public-short-sales as a contrary cue tied to a public selling mood and to later covering demand from existing shorts. That is the contrarian-strategy reading of the public series.

The case treats heavy member shorting as informed pessimism to be followed, not faded. Editorial note: a follow-the-member reading can disagree with a fade-the-public reading of the same pair of series, which is why the faded class has to be named.

An unusual-band chart of both series

The two volume-scaled series were charted against the Dow Jones Industrial Average with dashed lines at the mean and one standard deviation above and below it. That unusual-band convention treats points outside the band as unusually high or low and places about two-thirds of observations inside the band.

Joint, one-sided, and no-signal episodes

In 1983 the case read unusually low public shorting in the first half of the year together with high member shorting, under a contrary-opinion frame, as a forecast of a Dow Jones Industrial Average decline into 1984. Editorial reading: both classes were used, so this is a joint signal.

In 1984 and 1985 both series mostly stayed inside the middle band while public short sales trended higher, a pattern the case treated as bullish. Editorial reading: both series stayed inside the unusual-band, yet the case still treated the rising public series as a one-sided public signal.

In 1986 high public shorting was treated as a bull-market forecast and member shorts were judged not meaningful. Editorial reading: this is a one-sided public signal.

In 1987 public shorts were erratic and member shorts trended down from the first half, which the case read as falling member pessimism, but neither series was described as giving a clear signal before the October break. Editorial reading: this is a no-signal episode.

In 1988 the case described members as optimistic and the public as pessimistic and treated that pairing as two bullish signs looking ahead. Editorial reading: fade the public and follow the members, a joint signal.