The calculation of returns during seasoned equity offers

Abstract

The article analyses how the returns to a shareholder and the returns for an event study are calculated during the three types of seasoned equity offer (SEO) in use in the UK, namely rights issues, open offers and placings. The calculations differ across the two types of return and the three types of offer. Evidence from a sample of SEOs shows the large impact that the choice of calculation method has on returns. An unresolved question is whether to use discount-adjusted returns in event studies of placings.

Keywords:

• rights issues
• open offers
• placings
• returns
• price adjustments

JEL Classification::

• G24
• G14

Acknowledgement

I am grateful to three referees for detailed comments and to participants at the British Accounting Association Scottish Conference 2009 and BAA Conference 2010.

Notes

The focus of the article is on the calculation of returns, not on why researchers are interested in SEO announcements. See Eckbo, Masulis, and Norli (2007) for a comprehensive review of research on SEOs that includes both event-study evidence and research on the determinants of offer-price discounts.

If the SEO itself is a complete surprise, then arguably the returns should be gross of the impact of both the fees for the issue and the discount, since both are unexpected costs for the company. This is discussed further in Section 2.4.

In a scrip issue, the number of shares in issue is increased by giving shareholders n new shares for each existing share, where n exceeds zero. In a share consolidation, the number of shares in issue is reduced by replacing n existing shares with one new share, where n exceeds 1.

This ignores for simplicity the fact that the present value of the new cash as at day ex-1 is slightly less than .

It is sometimes the case that the new shares are not entitled to the next dividend, Div. Then the TERP formula is .

Rights are call options in which the exercise price is the offer price and the expiry date is the offer close. So in theory a right should be worth slightly more than . We ignore this for simplicity.

If the new shares are not entitled to the next dividend, Div, the value of the entitlement will be Ent.

An exception is Hull and Kerchner (1996), who argue that a large part of the negative AAR on announcement of US firm commitments represents the costs of issue. They measure the information effect of the SEO announcement by adding the costs of issue per old share to the prevailing share price at the end of the event window. Eckbo and Masulis (1992, 322) also present some results with the costs of issue added back.

For example, consider a project with cash in hand of £50 and cash flows of −£50 at date 1 and +£120 at date 2. Assume a discount rate of zero. The project's value at date 0 is £120. At date 1, the £50 cash is paid out, but its value remains at £120.

US firm-commitment offers, not used in the UK, are different again. There is no pro rata offer to existing shareholders. The offer price is set the day before the shares are issued, about 1 month after the AD. The offer price used to be set at or very close to the market price the day before the issue day, but discounts of a few per cent have become normal in the USA since the 1990s. Adler and Shea (2010) discuss returns during firm commitments. They include the value of the discount in their 2-day return for the issue date.

Twenty-one of the 31 shares with a discount-adjusted AR with an absolute value in excess of 30% are very small, with a market capitalisation as at AD-1 of less than £10 m. It could be argued that for such small, infrequently traded shares, the share price is a rather rough guide to the market's valuation. In some cases, the issue is extremely large in relation to the pre-issue size of the company, and so the discount-adjusted AR is very sensitive to the difference (see Equation 14, from Equation (9))).

Other errors in Datastream (and CRSP) prices are documented by Ince and Porter (2006).

The Extel Financial database was established as an electronic database in the mid-1980s, originally to contain the company information recorded on hard-copy Extel cards, which had been published since 1922. In the early 1990s, the database was expanded to include share price data. Extel was eventually acquired by Thomson Reuters, who ceased to make the database available in 2005.

It is easy to show that the derivative of Equation (A1) with respect to , exceeds the derivative of Equation (5) with respect to P _ex, 1/TERP. With no change in equity value, and . Both Equations (5) and (A1) switch sign at this value for P _ex.

What if the shareholder sells their rights at the start of the ex-day, the instant the shares go ex-rights? The return for day ex-1 would then be , where is the ex-rights price at the start of the ex-day, and . The return for the ex-day would be . If , the TERP adjustment is correct; otherwise the analysis is similar to that in the text.

Let N _j be the number of rights owned by holder j and x be the required proportion of the rights to be sold. x is found by solving

The left-hand side is the number of rights that is retained and subscribed for if the proportion x is sold, and the right-hand side is the number of new shares that can be bought from the proceeds of selling proportion x.