Читать книгу Merger Arbitrage - Kirchner Thomas - Страница 8

The simplest form of merger is a cash merger. It is a transaction in which a buyer proposes to acquire the shares of a target firm for a cash payment.

We will look at a practical example to illustrate the analysis. An announcement for this type of merger is shown in Exhibit 2.1, which is the press release announcing the purchase of Autonomy Corporation, a U.K. – based infrastructure software firm, by Hewlett-Packard Co. It is typical of announcement of cash mergers.

The terminology used in mergers is quite straightforward: A buyer, HP in this case, proposes to acquire a target, Autonomy here, for a consideration of £25.50 per share. The difference between the consideration and the current stock price is called the spread. When the stock price is less than the merger consideration, the spread will be positive. Sometimes the stock price will rise above the merger consideration, and the spread can become negative. This happens occasionally when there is speculation that another buyer may enter the scene and pay a higher price.

In a cash merger, the buyer of the company will cash out the existing shareholders through a cash payment, in this case £25.50 per share. An arbitrageur will profit by acquiring the shares below the merger consideration and holding it until the closing, or alternatively selling earlier.

Arbitrageurs come across press releases as part of their daily routine search for newly announced mergers. This one was released on August 18, 2011, at 4:10 pm Eastern Standard Time, which was 9:10 pm British Summer Time, when markets both in Europe and the United States were closed. For regulatory reasons, companies announce significant events like mergers after the end of regular market hours or in the morning prior to the opening. This is meant to prevent abuse by investors with slightly better access to news. With the growing importance of after-hours trading and the availability of 24-hour trading of U.S. stocks through foreign exchanges, this restraint has already become somewhat pointless but is still considered best practice.

Exhibit 2.1 Press Release Announcing Acquisition of Autonomy by HP (Extract)

PALO ALTO, Calif., and CAMBRIDGE, England, Aug. 18, 2011 – HP (NYSE: HPQ) and Autonomy Corporation plc (LSE: AU. or AU.L) today announced the terms of a recommended transaction under which HP (through an indirect wholly-owned subsidiary, HP SPV) will acquire all of the outstanding shares of Autonomy for £25.50 ($42.11) per share in cash (the “Offer”). The transaction was unanimously approved by the boards of directors of both HP and Autonomy. The Autonomy board of directors also has unanimously recommended its shareholders accept the Offer.

Based on the closing stock price of Autonomy on August 17, 2011, the consideration represents a one-day premium to Autonomy shareholders of approximately 64 percent and a premium of approximately 58 percent to Autonomy's prior one-month average closing price. The transaction will be implemented by way of a takeover offer extended to all shareholders of Autonomy. A document containing the full details of the Offer will be dispatched as soon as practicable after the date of this release. The acquisition of Autonomy is expected to be completed by the end of calendar 2011.

[…]

The first observation an arbitrageur will make is that the stock of Autonomy jumped immediately upon the announcement of the merger. As can be seen in Figure 2.1, Autonomy closed at £14.29 on August 18, the last day before the announcement of the merger. It opened at £25.27 on August 19, quickly peaked at £25.29, and moved down for the rest of the day to close at £24.52. Some unfortunate investors bought shares at the opening price, and because there must be a seller for every buyer, some lucky sellers parted with their investment at the high price for the day. An investor who wanted to enter into an arbitrage on this merger had a realistic chance of acquiring shares at the day's average price of £24.92. Volume that day was brisk: While it had averaged just under 1 million shares per day (precisely 0.97) over the prior month, it reached 48.6 million on August 19 and averaged 3.7 million per day over the next month. Therefore, the assumption that an arbitrageur could have obtained that day's average price is reasonable.

Figure 2.1 Stock Price of Autonomy before and after the Merger Announcement

A chart like that shown in Figure 2.1 is typical of stocks undergoing mergers. The buyout proposal is generally made at a premium to the stocks' most recent trading price. This leads to a jump in the target's stock price immediately following the proposal. As time passes by and the date of the closing approaches, the spread becomes narrower. This means that the stock price moves closer to the merger price. An idealized chart is shown in Figure 2.2, whereas Autonomy's actual chart is more typical of the behavior of most such stocks. Figure 2.1 also shows the FTSE index, the stock index considered a reference for the London market. Its axis has been scaled (right-hand side) to match the percentage change in Autonomy's stock price. If Autonomy and the FTSE have the same percentage change, then their respective lines will move by the same magnitude in the graphic. It can be seen that prior to the merger announcement, Autonomy's moves on a daily basis match those of the FTSE very closely. After the announcement on August 19, Autonomy and the index no longer move in tandem. This is a good visual illustration at the micro level of the low correlation that merger arbitrage has with the overall stock market. Fluctuations in the index do not impact Autonomy once it becomes the target of an acquisition.

Figure 2.2 Idealized Chart of Stock in a Cash Merger

In some instances, the buyout proposal is made at a discount to the most recent trading price. This rarely happens and is limited to small companies where the buyer is in a position to force the sale. It often leads to litigation and a subsequent increase in the consideration. A transaction at a discount to the last trading price is called a takeunder.

Insider Trading

Investors looking at the large jump in Autonomy's stock on August 19 will be tempted to calculate the profits they could have made with a little advance knowledge of the upcoming merger. Insider trading is a crime, not a form of arbitrage.

As readers of the financial press know, every merger cycle is characterized by insiders taking advantage of advance knowledge of mergers. Law enforcement has been successful in prosecuting even the most elaborate insider trading rings. One recent case involved New York bankers who bought options over the Internet through an online brokerage account established in Austria in the name of an elderly woman living in Croatia. Despite the complexity of the scheme, the perpetrators were caught and imprisoned.

Penalties for insider trading are up to 10 years in prison, in addition to monetary penalties, rescission of profits, and potential civil liability in shareholder lawsuits.

Not all that may look like insider trading really is insider trading

It has become a popular sport among academic economists to create models in order to demonstrate malfeasance in one area of finance or another. A particularly fruitful target appears to be insider trading around merger announcements. One study shows that short-term hedge funds increase their holdings of merger targets in the quarter prior to the announcement of a transaction and conclude from this finding that insider trading must be rampant. However, merger announcements do not occur randomly and do not happen in a vacuum. Astute observers can predict potential targets when a firm announces that it is reviewing strategic alternatives or has hired an investment bank. CEOs may talk on quarterly earnings calls about their desire to acquire firms, or be acquired. While these methods are far from perfect, they will be good enough to make variables in a quantitative model statistically significant. Similarly, potential acquirers frequently announce their intent to make acquisitions either explicitly or indirectly – for example, by taking out large lines of credit. Again, experienced observers will read the signals from these announcements and may often enough interpret them correctly. Studies that ignore such signals and consider only merger announcements miss relevant variables and yield meaningless results.7

The problem is certainly not insider trading; it is the misguided attempt to draw overly specific conclusions from naïve linear regression models based on a limited set of data, in particular when much relevant information is not easily quantifiable. It is an old wisdom among statisticians not to fall into the trap of “data availability.” Merger investing clearly has the potential for insider trading; however, considering that insider trading investigations over the last two decades have occurred outside the arbitrage community and concerned mostly individual investors, it is hard to see how there can be a problem.

An arbitrageur who buys the stock on August 19 for £24.92 will receive £25.50 when the transaction closes. The gross profit for the capital gain on this arbitrage is £0.58 on £24.92, or 2.33 percent:

2.1

where

This return will be achieved by the closing of the merger. A key component in investments is not just the return achieved but also the time needed. A more useful measure of return that makes comparisons easier is the annualized return achieved. The relevant time frame starts with the date on which the arbitrageur enters the position and ends with the date of the closing. The press release stated that the “acquisition of Autonomy is expected to be completed by the end of calendar 2011.” Therefore, the last day of the year, December 31, is used as a conservative estimate for the closing of the transaction. Pedantic arbitrageurs would choose December 30 instead because December 31 was a Saturday in 2011. As there is a large degree of uncertainty about when the transaction will actually close and the choice of the closing date is no more than an educated guess the difference between the two dates is not very meaningful. In practice, the companies will work very hard to close the transaction before the Christmas holiday, so that it is equally justifiable to work with a projected closing date of December 23rd. There are 126 days in the period until the anticipated closing to December 23rd. Two methods can be used to annualize the return: simple or compound interest.

Simple interest

2.2

where

Compound interest

2.3

where

Personal preference determines which method is used. Simple interest is useful if the returns are compared to money market yields that are also computed with the simple interest method, such as the London Interbank Offered Rate (LIBOR) or Treasury bills (T-bills). Compound interest is preferable if the result is used in further quantitative studies. If the returns are compared to bond yields, they should be adjusted for semiannual compounding used in bonds. It is an error encountered frequently, even in research by otherwise experienced analysts and academics, that yields calculated on different bases are compared with one another.

A projected annualized return of 6.74 percent is sufficient to make this investment highly desirable at a time when overnight LIBOR rates in Sterling were around 0.58 percent and the 10-year benchmark Gilt yielded around 2.6 percent.

It is helpful to look at the actual outcome of this merger arbitrage. The Autonomy acquisition closed earlier than an arbitrageur would have assumed: November 14, 2011. With this shorter 88-day time frame to closing, the realized annualized return was 9.65 percent and 10.01 percent for the simple and compound interest methods, respectively. Over the same period, the FTSE returned 10 percent, or an annualized 48.5 percent. However, this better short-term performance came at a price of a volatility that was also significantly higher: Autonomy's daily volatility was 3.4 percent, whereas that of the FTSE was 25 percent.

Autonomy was a non–dividend-paying company. In case a company does pay dividends, there is another source of income that the arbitrageur must factor into the return calculation. For an example, consider Australian bulk grain exporter GrainCorp Limited, which was acquired by Archer-Daniels-Midland Co. for A$2.8 billion. The per-share acquisition price was only A$12.20, but an additional A$1 was to be paid in dividends. Due to the large dividends to be received by shareholders, the stock traded above the A$12.20 level following the announcement of the merger. On April 30, 2013, four days after the announcement, an arbitrageur could have acquired GrainCorp for a volume weighted average price of $12.8239, with an expectation that the transaction would close by September 30, 2013, or within 157 days. A back-of-the-envelope calculation for the net return with dividends is to add the dividend to the merger consideration received. This gives an annualized return of 6.95 percent if the compound interest method is used:

2.4

where

A more accurate method is the calculation of the internal rate of return (IRR). Spreadsheets have built-in functions to calculate IRRs that require the user to enter each payment with the associated date, as shown in Figure 2.3. It is important to note that the dividends were spread over different payment dates. A first net dividend payment of A$0.25, consisting of a $0.20 interim dividend and a $0.05 special dividend, was to be paid on July 19. The dates of any future dividends were not yet known, Since Australian companies pay semiannual dividends, it is safe to assume that no dividend will be received during the 10-week period between July 19 and the closing on September 30. The prior final dividend was paid on December 17, 2012, so that the final dividend would probably also be paid in the middle of December should the merger not be completed by then. Since the arbitrageur is working for now with a closing date of September 30, it is assumed that the final dividend payment will be made simultaneously with the payment of the merger consideration on September 30.

Figure 2.3 IRR Calculation of Annualized Return in Excel

The resulting projected return is an annualized return of 7.21 percent, slightly higher than in the simplified calculation. The reason for the difference lies in the earlier receipt of the dividend cash flow in the IRR calculation.

The actual date of the dividend payment and its amount are not always known at the time of the announcement of a merger. In this case, the arbitrageur will make an educated guess for the next payment date based on the company's payment frequency and past dividend amounts. Care must be taken when foreign companies are listed in another country. For example, U.S. companies pay dividends with a quarterly frequency. However, U.K. companies listed in the United States and their ADRs pay semiannual dividends to all of their shareholders, even those who purchased the shares in the U.S. market. Similarly, Swiss companies pay only one dividend per year even when listed in the United States. Whenever the dividend information becomes known, such as the announcement of the dividend date or the exact amount, arbs must update their spreadsheets promptly.

Arbitrageurs must be aware of a few limitations of this approach:

• The returns calculated are projections that are highly dependent on the date of the closing of the merger. A delay can quickly lower the return on the investment.

• The projections say nothing about the path that the investment takes on its way to closing. Sometimes, following an initial spike after the merger announcement, a target company's stock weakens. An arbitrageur will then have to book a temporary loss on the investment. Of course, this marked-to-market loss will be reversed eventually when the merger closes. However, the projection cannot make a prediction on the trading dynamics of the stock between purchase and merger closing.