Every financing announcement is an information event, because management chose it and management knows things you do not.
How much does the share price move on each announcement?
Ranked by the average abnormal return around the announcement, from most negative to most positive: a seasoned equity issue at roughly negative 3%, a convertible issue at roughly negative 2%, a straight debt issue at approximately zero, an open-market buyback authorisation at roughly positive 2 to 3%, and a fixed-price tender offer at roughly positive 8%. Dividend initiations are positive and dividend cuts sharply negative, which is the payout half of the same phenomenon and belongs to the dividend-policy work.
Sized on Wexford's $2,000 million of market value, those percentages are negative $60 million, negative $40 million, zero, positive $50 million and positive $160 million. Writing them in money is the discipline, because a percentage sounds like a statistic and a number sounds like a decision.
Why does the ordering hold, and what are the cautions?
The mechanism behind the ordering is consistent. Issuing a claim whose payoff is sensitive to the firm's true value reveals more than issuing one whose payoff is nearly flat, and a repurchase reveals the opposite of an issue. A convertible sits between debt and equity in payoff sensitivity and between them in announcement effect, which is a small piece of evidence that the mechanism is the right one.
Advertisement
Three cautions before you use any of this. These are averages over large samples with wide dispersion, so a single announcement tells you very little on its own. The effect is conditional on what the market already expected, which means a company that pre-announced a funding need has spent the signal already. And a tender offer at a premium, especially the Indian variety, is functionally a dividend wearing a buyback's clothes, so read the mechanics before reading the signal.
{
"skill": "Assembling the announcement-effect evidence behind a financing recommendation",
"use": "Use an assistant to build a candidate list of studies, sample periods and reported effect sizes for a given announcement type and market, and to convert each percentage into money on your subject company's market capitalisation.",
"tools": ["Chat assistants with web access, for locating the papers and their abstracts", "SSRN, NBER and journal publisher pages as the primary sources", "A spreadsheet for the percentage-to-money conversion and the sensitivity band"],
"prompt": "List the published event studies on announcement returns for seasoned equity offerings, straight debt issues, open-market repurchase authorisations and fixed-price tender offers. For each give the authors, year, sample market, sample period, number of events, average abnormal return and event window, and flag every entry you are not confident about. Then compute what each figure is worth in dollars on a market capitalisation of 2,000 million, with the dispersion around each mean where the paper reports it.",
"verify": "Open each paper and read the abstract yourself before any figure reaches a memo, because an assistant will attribute a plausible effect size to a real paper that never reported it and the citation looks correct from the outside. Check the sample period against the market you are advising in, since a United States sample from the 1980s is not evidence about an Indian issuer today. Confirm the event window, because a one-day and a three-day abnormal return are different quantities. Cite the paper, the year and the sample period in the same sentence as the number, and give the dispersion, since a mean without a spread invites a false inference about a single announcement.",
"diy": "You must be able to explain, with no literature at all, why an equity issue is bad news and a repurchase is good news, deriving the ordering from the payoff sensitivity of each claim. That derivation is what the gate rewards, and it is what lets you recognise a reported effect with the wrong sign as an error rather than a finding.",
"market": "US"
}