Briefly Speaking: Leaks and Announcement Effects
We're all aware by now that often by the time events are announced, they are widely known to bigger and more sophisticated players. The most vivid example here is the time two Libyan jet fighters were shot down by the US over the Gulf of Sidra in 1981 on a Sunday, New York time. But on the Friday before, gold went up 10 bucks. How could they have known about a cardinal event like the shooting down of a plane? One wants to say it was coincidence, the conjunction of a 1-in-10,000 event with a 1-in-100 event, which isn't too unlikely considering the number of events that occur. But then it turns out that we had been turning up the screws the whole previous week, threatening the Libyans over and over again, and telling them to restrict their flights or else.
My least favorite example came five minutes before Henry Kaufmann turned bullish on bonds for the first time based on his crackpot flow-of-funds analysis: "The amount of fixed-income supply is $200 billion greater than the demand so rates must go up by 5 percentage points to create the weltanschauung." The firm he worked for offered me bonds at a very good price, two ticks below the market (a one and only), and they rose about 4 points in the next five minutes after the announcement much to the cost of the Palindrome and myself.
The most startling example I came across was a little-noticed editorial by Caspar Weinberger in Forbes a few months before the impeachment hearings. He predicted that the day before the hearings, the U.S. would lob some missiles against the Mideast adversaries, to deflect attention and show the need for a strong imperial presidency, even if attenuated by excessive romance et al.
It is well known that the market looked like Hades in the weeks before 9/11. I received a call the weekend before saying that another 10/19/87 was predicted. The market hadn't gone up two days in a row for the previous two months, which may have been a record of some kind. The S&P itself declined from 1330 on May 22 to 1103 at the close on Sept. 10 (using adjusted futures prices). And out of the clear blue sky, volatility on options went up by 14 percentage points from 19% on June 30 to 32% at the close on Sept. 10.
Insider trading, of course, is the heart and soul of much on the Street. Before merger announcements, the acquiree often rises. Put buyers abound before the announcement of inordinately bad earnings. Indeed, when I studied insider trades in detail with James Lorie in the '60s, we found that 75% of all insider buys and sells were at more favorable prices when the insiders traded than the price prevailing when the transactions were announced.
Particularly reprehensible to me are the staged leaks that the great leaders at the Central Bank are so famous for. They leak to selected reporters what their plans are. They meet with the big brokerage houses to discuss their likely reactions to various events.
The Recent Declines.
Of course what brings this to mind at present was the decline in the market in the week ended Oct. 7, 2005. Five straight days covering 34 points, including the inordinately bullish first day of the month. What could possibly have caused this terrible and once in 50 times magnitude first-week decline? It seemed ridiculous that the three staged announcements that the Fed was near the upper band of its limit of tolerance for inflation could have caused the decline because this was a repetition after repetition of what had already been announced. The market isn't supposed to move based on historical events but on changes in anticipation. It's all so much clearer now that the story about the big brokerage house failure is becoming disseminated. It's like the Libyan negotiations. Big fund transfers occurred the previous week involving an Austrian bank. Resignations had to be prepared. Agreements to pay off debts had to be arranged. Doubtless dozens of professionals had to be involved in these and similar meetings, to say nothing of the pre-notifications to the authorities. The big owners and lenders would presumably have been given their opportunity to protect the interests of the public and themselves. There has been much notice about the planned family wine-tasting tour of the chief perp to Vienna that was on tape. Was such taping routine, or part of something else?
A classification of events.
I have long wished to classify events based on their information content and timing. The best I have been able to come up with is a three-category classification into uncertainty classes as follows:
For example, an earnings release from GE would be known as to its date, and would receive a time of announcement of 10; magnitude of announcement would be uncertain, say 4-6; seeming favorability would be 6.
An expected 1% increase in CPI would be 10 for certainty of time of announcement, 4-6 for magnitude and 1 for seeming favorability.
A future change in the fed funds rate at an unscheduled meeting would be 1 for certainty of announcement, 8 for magnitude of effect, and 10 for seeming favorability.
I am open to other ruminations on this deep subject including alternate methods of classifying events relative to their uncertainty.