Value to Shareholders: What It Takes For a Deal to Win Stock Market Approval Part II In the short run, investors look beyond the trappings of m&a transactions to reward deals based on good business sense.

The 1990s have been called the era of strategic mergers. Powerful change forces have transformed the economic and financial environments. Global markets and exploding technologies have blurred industry boundaries and intensified competition. Strong economic growth and a rising stock market have rewarded the leaders and penalized the laggards. These pressures have forced companies to pursue mergers and acquisitions, divestitures, joint ventures, and restructurings in the effort to achieve profitable growth. The inevitable question is how successful these deals have been and there are numerous measures for gauging the value they have or have not created. In recently completed research we took a different approach to the question by linking major transactions to such deal elements as terms and accounting treatment. We looked at a sample of 364 transactions that accounted for almost half of the total m&a values between 1992 and mid-1998 to reveal deal patterns. Overall, pooling-of-interest accounting, which is scheduled to be eliminated by the end of 2000, accounted for 52% of the big deals. Pooling accounted for 80% of bank mergers but only 45% of non-bank mergers. So, the demise of pooling should affect less than half of non-bank mergers. In the 1980s, debt and cash from sales of debt accounted for the majority of transactions in the sample. In the 1990s, stock-for-stock deals accounted for 80% of the big deals. Financing is readily available for the other 20%. Eighty-five percent of bank mergers were non-taxable and were mostly poolings. Only 60% of non-bank mergers were non-taxable. When acquisitions were treated as purchases, at least 70% were taxable. The average premium paid over the stock price of the target 30 days before the deal announcement was about 40% whether pooling or purchase accounting was used. But in purchase transactions that were taxable, the target received a 42% premium versus 34% in deals that were non-taxable. Stock market reactions to the deals are measured over a 30-day window beginning 20 days before the announcement date and ending 10 days after. The stock prices of non-bank buyers went up for 52% of the companies when the deals were announced. On the seller or target side, 90% were positive for the non-bank transactions. The predominance of positives for the targets is not surprising since risk arbitrage tends to push the stock price toward the purchase price, although there are some exceptions. However, risk arbitrage activity usually involves shorting the buyer while going long on the target. When the market judges the deal as making good business sense, this overcomes the selling pressure, which occurred for the 52% of the non-bank deals that had positive stock market gains. However, for banks, the market seems to question whether efficiencies will be achieved, and, therefore, negative price responses for buyers occurred 62% of the time. The deals for which the stock market reacted positively, less the negatives, generated a net gain of $48 billion for non-bank buyers and a net gain of $99 billion for non-bank sellers. Banks did not fare as well. Bank buyers had a net negative market response of $15 billion while sellers gained $22 billion. Overall then, the stock market has judged that these large transactions make sound business sense. Academic studies show that the initial stock market response to merger announcements is a good predictor of future performance. Will these predictions hold for the future? The optimism of the times may require some reservations. But the strong change forces suggest that the many responses by business reflect sound entrepreneurial responses. We now will present a more detailed analysis of the patterns of deal structuring for the big deals of the 1990s. We will cover accounting treatment, method of payment, taxability, premiums paid, and measurements of stock market reactions. The merger movement of the 1980s peaked in 1988 at $238.5 billion. Legislative and economic forces resulted in declines in dealmaking in 1989 through 1991 but a resurgence of m&a activity began in 1993. Each successive year recorded a substantial percentage increase over the previous year, with dollar value reaching a record $1.3 trillion in 1998. It is useful to analyze in some detail the blockbuster transactions whose characteristics differ from the patterns for total transactions, which, by number, would be predominantly smaller deals. Our sample consists of 364 transactions, representing 48.5% of the total dollar value of deals for the period under study. Our selection criteria began with all deals in which the price paid for the target exceeded $500 million. By 1997, the annual number of those transactions became so large that we raised our cutoff point to deals valued at $1 billion or more. Our study ended with transactions announced through June 1998. The stock market adjustment that began in July 1998 dampened new deal announcements. For completed transactions, however, the third quarter of 1998 was still high because of deals initiated earlier. The stock market began to recover in mid-October and was associated with a resumption in an active m&a market. Thus, our study captured a distinctive cycle of m&a activity. Our sample accounted for about 40% to 45% of total deal values in most years and reached almost 69% for the first half of 1998. The exploding number of blockbuster transactions is consistent with our data. Deal Structuring Surprisingly the rules governing the accounting treatment of mergers and takeovers had not been fundamentally changed since the 1970 Guidelines issued by the Accounting Principles Board (APB) of the American Institute of Certified Public Accountants (AICPA). The board’s Opinion 16 specified the 12 conditions required to use pooling-of-interest accounting, and Opinion 17 dealt with the treatment of goodwill involved in purchase accounting. Pooling-of-interest accounting permits the financial statements after a merger to be calculated by the simple addition of all balance sheet and income statement accounts of the constituent companies. In purchase accounting, the excess of the market value paid over the book net worth of the target is first assigned to existing tangible assets and the remainder is goodwill to be written off over 40 years. This write-off is generally not tax-deductible and it reduces the bottom-line net income figure. The successor to the APB, the Financial Accounting Standards Board (FASB), announced in its Bulletin No. 197-A (May 18, 1999) that it had voted unanimously to eliminate pooling-of-interest as a method of accounting for business combinations. The FASB stated that the change will become effective after it issues final standards in late 2000. Published data do not provide systematic numbers on the extent to which pooling versus purchase methods of accounting are employed. But this information is available in proxy statements sent to shareholders in connection with approvals for transactions. We obtained copies of these proxy statements for our 364 companies. From these, we compiled data on accounting treatment as well as method of payment, taxability, premiums, and initial stock market reactions. Pooling Versus Purchase Accounting Transactions that use purchase accounting usually involve the acquisition of a smaller firm by a larger business, and the purchase method is used in the vast majority of transactions taking place in the m&a market. By contrast, in our sample of 364 deals, pooling accounted for slightly more than 52% of the deals, as shown in Table 1. However, 75 of the 364 transactions, or 20.6% of the sample, involved banks. Eighty percent of bank transactions were poolings and only 20% were purchases. For non-bank deals, purchases accounted for 55% of the total. The data indicate that the banks had a very strong preference for pooling. One possible explanation is that banks are strongly averse to the negative impact of goodwill write-offs on reported net income. However, the strong avoidance of non-pooling transactions does not manifest itself in non-bank transactions. One possible explanation is that, in general, the economies or synergies in the non-banking transactions are strong enough to overcome the negative effects of goodwill write-offs because of the increase in earnings that the combined firm will be able to achieve. The point is that firms in highly synergistic transactions are less averse to the negative impact of the goodwill write-off because increased earnings will more than offset the negative effects. This also leads to the prediction that highly synergistic mergers will not be deterred when pooling accounting may not be available. Method of Payment The 1980s often are referred to as the decade of mergers propelled by junk-bond financing. The sale of debt raised cash that was used in takeovers often hostile. Data compilations showed these deals as cash transactions, although the underlying source was debt. Our data show that during the 1990s, 60% of the 364 largest deals exclusively used stock, with combinations of stock and cash moving the proportion up to almost 80% (see Table 2). Stock-for-stock transactions are generally non-taxable. In bank mergers, stock was involved, either exclusively or in a combination of stock and cash, in more than 90% of the deals. In non-bank mergers, the proportion dropped to about 75%. In terms of deal numbers, the m&a market is dominated by smaller transactions typically done for cash. As a result, a broader compilation determined that stock is involved in only about one-third of all transactions, regardless of size. A brief generalization is that big deals in the 1990s have been mainly stock-for-stock transactions while the seller was most likely paid in cash in the smaller transactions. Taxability Table 3 shows that 60% of the non-bank transactions were not taxable to selling shareholders because they received the acquirer’s stock. Yet, Table 1 showed that 45% of non-bank deals were accorded pooling-of-interest accounting treatment, which revolves around the use of stock. What accounted for the 15% differential? The conclusion is that several non-taxable, non-bank transactions used purchase accounting but still qualified for non-taxable treatment. Acquirers in some stock-for-stock transactions might not have met all of the 12 rules required to qualify for pooling-of-interest accounting. For example, if one of the participants in a merger had engaged in stock buybacks during the two years preceding the year of the deal, it would fail to qualify for pooling-of-interest treatment. But as a stock-for-stock transaction, the deal still would exempt its shareholders from paying taxes on their proceeds. Table 4 shows that almost 54% of all non-bank transactions in which purchase accounting was used were taxable transactions. In another 19%, taxability depended on whether the seller chose to take cash or stock when the buyer provided an election option. More than 85% of bank transactions were non-taxable (as shown in Table 3), reflecting the predominance of pooling in bank deals. By adding the 6.7% of bank deals in which the buyer offered the seller the option to take cash or stock, we found that more than 90% of the bank deals probably qualified for non-taxability. Purchase Price Premiums We based purchase price premium measurements on the stock price of the target stock 30 days before the public announcement of the deal. We did this in order to avoid the inevitable run-up in the price of the seller shares in response to the leaks that typically develop in the five to 10 days before the formal public announcement date. The 30-day percent premium was about 40% for the seller in non-bank transactions when an arithmetic mean is used to average over the deals. In an arithmetic average, the larger numbers receive a higher implicit weighting. The use of a median half the deals below a certain level and half above eliminates that potential distortion. This gives less weight to the larger numbers. Thus the median premium is 33% for non-bank pooling transactions and 37% for non-bank purchase transactions, as shown in Table 5. When non-bank transactions using purchase accounting are grouped by taxability, the target received a 42% mean premium compared with a 34% premium in non-taxable transactions. This implies that the buyer pays more when the seller is in a taxable transaction. Generally, the pattern for premiums over a target’s stock price 30 days before the announcement date was in the 33%-to-40% range for non-taxable non-bank deals. For taxable non-bank deals, the premiums to sellers appear to jump by three to four percentage points. Stock Market Reactions Empirical studies have found that the initial market reactions to merger announcements are good predictors of subsequent performance. These initial market reactions are measured by statistical procedures that track stock price changes for acquirers and targets for periods both before and after the announcement dates. These stock price changes are adjusted for contemporaneous general stock market movements. The resulting residual represents stock price changes, net of the general market movements, associated with a distinctive event, e.g., mergers, new stock issues, product defects, executive changes. For a large sample of companies, the average stock price change net of market movements would be expected to be zero, since a large sample would mimic market behavior. For a sample based on a common event, such as a merger, with changes measured around the event date, such as the merger announcement, positive or negative differences from general market movements are referred to as abnormal residuals from market movements, or event returns. They represent a measure of whether the stock market evaluates an announced transaction as one that is likely to be successful or good versus a deal destined to be unsuccessful or bad. We calculated positive and negative gains for our sample. Stock price data were not available for all transactions, so for this analysis, our sample size drops from 364 to 309. We multiplied the positive or negative net percentage gains or losses times the market value of equity for the acquirer and the target, 20 trading days before the announcement date through 10 trading days after. The results can thereby be presented in absolute dollar terms. The dollar returns to targets measured over the 30-day window are almost always positive. Whether the event returns for the acquiring firm are positive or negative depends on the market’s judgment of whether the premium paid to the seller will be recovered in the subsequent performance of the combined firm. Table 6 presents the overall results for our event-return analysis. For the total sample, about two-thirds of the deals had positive returns, suggesting a predictability of merger success. Our results thus suggest that two out of three large mergers are likely to add value to shareholders. But for the bank sub-sample, the percentage of predicted success declines. The success ratio is higher than the total sample for non-bank deals. Then we looked at the absolute dollar amounts involved. In the non-bank sample, the dollar amount of the increases in the market capitalization of buyers enjoying upward price movements came to about $130 billion (see Table 7). The aggregate decline of non-bank buyers suffering stock price declines was about $82 billion. Thus, there was a net gain to non-bank acquirers of $48 billion during the sample period. On the sell side, the positive movements of target stock prices added up to a $109 billion increase in market cap for non-banks. There were, however, price declines for some sellers that totaled nearly $10 billion. The net gain for all non-bank sellers was $99 billion. Bank buyers had net losses of $15 billion while bank sellers had net gains of $22 billion. We also analyzed stock market gains and losses based on method of accounting, taxability, and related matters. However, the differences are not statistically significant. Our judgment is that the business soundness of a deal determines how the stock market will react to it. If the market thinks that the deal will work out well in the future, the initial market response is likely to be favorable. If the market believes that the deal is misconceived and does not have a sound business foundation, it will react negatively to the deal. If the transaction is soundly conceived, the stock prices of the sellers and/or buyers will increase. The method of accounting used and taxability are of secondary importance. The important lesson is that good deals assuredly will increase stock prices for the sellers and even for the wisest buyers despite some initial shorting by risk arbitrage traders. Bad deals will be bad news for shareholders, both for the acquiring firms and ultimately for the selling firms. The green light The green lightThe short-term stock market reaction to a major merger or acquisition is considered a good predictor of whether the deal will work out for the buyer. The question is how investors reach this judgment. Extensive research into stock price reactions after the largest transactions of the 1990s suggests that investors reward deals that enjoy proven synergies, make good economic sense, and can generate a net increase in earnings for the combined firm. Ability to recover the purchase price premium is critical. Conversely, the researchers at UCLA found little linkage between the stock market’s reaction and accounting treatment purchase versus pooling or tax implications for the sellers. Table 1: Accounting Treatment in Large Deals 1992 to Mid-1998 Bank Non-Bank Combined No. of % of No.of % of No. of % ofMethod Deals Deals Deals Deals Deals DealsPooling 60 80% 130 45% 190 52.2%Purchase 15 20% 159 55% 174 47.8%Total 75 100% 289 100% 364 100%<\TBL> Table 2: Method of Payment in Large Deals1992 to Mid-1998 Bank Non-Bank Combined No. of % of No. of % of No. of % of Method Deals Deals Deals Deals Deals DealsCash 7 9.3% 72 24.9% 80 22%Stock 61 81.3% 159 55% 219 60.2%Cash and Stock 7 9.3% 57 19.7% 64 17.6%Debt 0 0% 1 0.3% 1 0.3%Total 75 100% 289 100% 364 100%<\TBL> Table 3: Taxability of Large Deals1992 to Mid-1998 Bank Non-Bank Combined No. of % of No. of % of No. of % ofMethod Deals Deals Deals Deals Deals DealsTaxable 6 8% 85 29.4% 91 25%Non-Taxable 64 85.3% 174 60.2% 238 65.4%Election 5 6.7% 30 10.4% 35 9.6%Total 75 100% 289 100% 364 100%<\TBL> Table 4: Purchase Accounting and Taxability in Large Deals 1992 to Mid-1998 Bank Non-Bank Combined No. of % of No. of % of No. of % of Method Deals Deals Deals Deals Deals DealsTaxable 6 40% 85 53.5% 91 52.3%Non-Taxable 4 26.7% 44 27.7% 48 27.6%Election 5 33.3% 30 18.9% 35 20.1%Total 15 100% 159 100% 174 100%<\TBL> Table 5: Premiums Over Target Stock Prices* 1992 to Mid-1998 Bank Non-Bank CombinedAccounting TaxMethod Treatment Mean Median Mean Median Mean MedianPooling Non-Taxable 44% 35% 40% 33% 41% 34%Purchase Total Purchase 36% 34% 41% 37% 40% 37% Taxable 23% 12% 42% 37% 41% 37% Non-Taxable 48% 41% 34% 36% 35% 36% Election 34% 45% 52% 49% 47% 46%*30 days before deal announcements.<\TBL> Table 6: Percentage of Positive and Negative Total Gains in Large Deals1992 to Mid-1998 Buyer Seller Combined No. of % of No. of % of No. of % ofMethod Deals Deals Deals Deals Deals DealsBanks Positives 27 38% 63 88.7% 41 57.7% Negatives 44 62% 8 11.3% 30 42.3%Non-Banks Positives 124 52.1% 213 89.5% 161 67.6% Negatives 114 47.9% 25 10.5% 77 32.4%Total Positives 151 48.9% 276 89.3% 202 65.4% Negatives 158 51.1% 33 10.7% 107 34.6%<\TBL> Table 7: Summation of Positive and Negative Total Gains in Large Deals*1992 to Mid-1998 Buyer Seller CombinedBanks Positives $12,782 $26,006 $26,812 Negatives (28,191) (3,946) (20,162) Total (15,409) 22,060 6,650Non-Banks Positives 129,675 108,880 213,947 Negatives (81,641) (9,723) (66,756) Total 48,034 99,157 147,191Total Positives 142,457 134,886 240,759 Negatives (109,832) (13,669) (86,918) Total 32,625 121,217 153,841* (in $ millions)<\TBL>