Back on March 25, 2004, we had mentioned a somewhat obscure indicator — NYSE Specialist / Total Short Ratio. It was one of the reasons we got Bullish on March 22 — a call we didn’t reverse until mid-April.
At the time, we wrote:
“Another sentiment indicator turns Bullish: the Specialist to Public Short Ratio (60 year chart), is at a levels which in the past have been associated with intermediate lows.
Specialists themselves may not be bearish, but often turn out later to be right because the public or the “crowd” was wrong. In declining markets, it’s the public who is doing the selling (and shorting) and the specialists who must buy to maintain an orderly market – even if they are not bullish at all.” (A variation of the NYSE Short interest is the NYSE Member Weekly Buy/Sell ratio.)
Now, the NYT has picked up on the same indicator:
“STOCKS are likely to rebound, at least for a while, if one obscure indicator, reflecting the investment patterns of an important Wall Street constituency, proves as accurate a forecasting tool as it has for the last 60 years.
The buy signal comes from the eight-week moving average of the weekly New York Stock Exchange specialist short-sale ratio. The ratio fell on July 23 to its lowest level, 22 percent, since at least 1943, when reliable records of the indicator were first compiled. That means specialist firms – brokers appointed by the exchange to maintain orderly markets in individual stocks, often by buying and selling shares themselves – accounted for about 22 percent of all N.Y.S.E shares sold short in the eight weeks through July 23. Selling short is a way to bet on declining prices, and the lower the ratio, the less short-selling the specialists are doing compared with other investors.
The weekly ratio can be calculated from the “round lot report,” found by entering those words in the search box of the exchange’s Web site, www.nyse.com (in the search results, look for a document titled “roundlots.html”). Divide specialist short sales (the second figure in the right column) by total short sales (the first figure in the same column) to obtain the ratio. The data issued by the exchange is usually about two weeks out of date.
When the ratio has fallen below 35 percent in the years since 1943, stocks have often rallied.”
While our bullish call 2 weeks ago was wrong — not early, as some have termed it, but flat out wrong — we continue to see signs of a bottoming process.
Note that this is not the one note song of a perma-bull; We made a very fortuitous Bearish call on January 22, 2004 before getting Bullish on March 22. We advised stepping aside in mid-April, only to get Bullish again on May 17, 2004. That long bias lasted until June 30. Finally, the call late July moved us back into the Bullish camp — while we did see a small bounce for a few days, that’s not what I was looking for.
So at this point, while cognizant of the trend and the selloff, I continue to see the signs of the bottoming process.
Click for larger graphic
Graphic courtesy of NYT
Silver lining: A friend commented to me: “The wronger this call is, the worse it is for the incumbent.” So if that’s your politics, you take some cold comfort in losing money. . .
Seeing Signs of a Stock Recovery in Some Obscure Tea Leaves
CONRAD DE AENLLE
NYT, August 8, 2004