VALUE CREATION THROUGH INSURANCE COMPANY EQUITY CARVE-OUTS By Krupa S. Viswanathan July 2006 Krupa S. Viswanathan Temple University 471 Ritter Annex (004-00) Philadelphia, PA 19122 215.204.6183 215.204.4712 (fax) krupa@temple.edu 9
VALUE CREATION THROUGH INSURANCE COMPANY EQUITY CARVE-OUTS ABSTRACT A major goal of publicly traded firms is to maximize its value in the interest of its shareholders. For a firm that has subsidiaries, the parent company may attempt to achieve this goal by divesting itself of a subsidiary through an equity carve-out. This action would allow for an infusion of capital for the parent. In addition, by creating some degree of independence between the parent and subsidiary, the parent can be on a path to focusing on its core operations, if the subsidiary s line of business differs. In this research, the market reaction to insurance company carve-outs is investigated to determine whether the parent firms are rewarded through a higher stock price for this corporate action. While the initial reaction to the announcement is significantly positive, over the 75-day period surrounding the announcement, the market reaction is negative, in contrast to findings in other carve-out studies that aggregate various industries. INTRODUCTION Value maximization is a primary goal for stock organizations and issuing an initial public offering in subsidiary companies has the potential for creating greater value. Publicly traded insurance companies aim to maximize value for a variety of reasons beyond those considerations held by any stock organization, such as regulatory considerations and increased capital standards. A parent company which is publicly traded could issue additional shares of stock in its own firm, termed a seasoned equity offering, proceeds of which can then be infused into a subsidiary to support positive net present value opportunities. By offering shares directly in the subsidiary, 1
however, the financial market has the opportunity and ability then to directly measure the worth of that company. There are several ways that a parent can divest itself of its ownership in a subsidiary company. In a pure spin-off, shares of stock in the subsidiary are given to shareholders as a taxfree dividend. Thus, the parent s shareholders also become the subsidiary s shareholders through this complete divestiture. In a sell-off, the parent typically sells all of the assets of the subsidiary to another firm while in an equity carve-out, the parent issues an initial public offering of the subsidiary s stock. The parent may choose to retain some portion of ownership however at a later date, the parent could spin off or sell the remaining shares. The focus of this paper is equity carve-outs of insurance companies and the wealth effect it has on the parent at the time of announcement and the effect it has for both the parent and subsidiary firms after the IPO. The market reaction for 18 equity carve-outs of insurance companies is examined in this research. Typically, the parent firm makes a public announcement of its intention to issue an initial public offering in a subsidiary. The stock returns for the parent during the period surrounding the announcement date are examined to measure the extent of any abnormal return. It is investigated whether the market rewards or penalizes the parent for carving out a subsidiary. We then measure the initial returns for the IPO and also calculate the corresponding returns for the parent. We wish to investigate whether the two entities are worth more separately rather than as a whole. The paper is structured as follows. The next section presents a literature review of divestitures, focusing on the impact of equity carve-outs, spin-offs and sell-offs. This leads into a discussion of why an insurer would execute a carve-out, specifically what further considerations would influence insurers. The data and analysis are then presented with a 2
discussion of the results. Conclusions about insurer equity carve-outs and future research then follow. THE ISSUANCE OF EQUITY CARVE-OUTS Prior Studies There have been numerous studies in the finance literature that examine the market reaction to divestiture announcements, which include spin-offs, sell-offs and equity carve-outs. There are several major reasons why firms would dilute its ownership of a subsidiary; such reasons include tax advantages and the facilitation of mergers and acquisitions (Hite and Owers, 1983; Johnson, Brown and Johnson, 1994). A divestiture may also be undertaken if the parent and the subsidiary differ in their lines of business but the parent wishes to remove this diversification from its corporate structure. While the focus of this paper is equity carve-outs, it is of much interest to examine reasons why firms would choose any of its divestiture options. Both the divestiture gains hypothesis and asymmetric information hypothesis have been proposed in the papers described below. In Schipper and Smith (1986), the market reaction to equity carve-outs and seasoned equity offerings are compared. Under either transaction, shares of stock are sold with the proceeds generally directed towards the parent. However, these transactions differ in several marked ways, which correspondingly have their own implications. As the authors recognize, in a carve-out, because the subsidiary becomes publicly traded, it must release its financial statements on a periodic basis. Correspondingly, the market can value its net worth more effectively. In addition, having a tradable currency allows for the creation of incentive compensation contracts for managers. Thus, an equity carve-out allows for divestiture gains. 3
Schipper and Smith (1986) estimate the abnormal returns surrounding the announcement dates for 76 equity carve-outs over the period 1963-1983. Using the market model and gathering daily stock prices for the 280-day period prior to the announcement date, abnormal returns are computed for the 85-day period surrounding the announcement dates. Over the announcement period, consisting of day 4 through day 0, an average abnormal return of 1.83% is found, which is statistically significant. The market reaction to equity carve-outs is positive and does not get diluted during the 40-day period after the announcement date. In an asymmetric information model by Nanda (1991), firms have the opportunity to invest in a subsidiary firm s project. To fund this, they could issue a seasoned equity offering, or they could choose to execute an equity carve-out of the subsidiary. Values for the subsidiary and non-subsidiary portions of the corporation are assumed to be uncorrelated and random. As the results of the model show, a firm is more likely to choose a carve-out if as a whole, it is undervalued but the subsidiary is overvalued. This is in contrast to a parent firm that is overvalued; such a company would rather issue a seasoned equity offering to tap that overoptimism. Thus, this corporate action acts as a signal to the financial market about the true value of assets. Nanda (1991) explains that this result supports the positive abnormal returns found in the Schipper and Smith (1986) carve-out study. In Miles and Rosenfeld (1983), the market reaction to 55 spin-off announcements over the period 1963 to 1980 is investigated. The authors anticipate a positive reaction to these announcements, given that a spin-off can offer the shareholders flexibility by allowing them to choose which entity they wish to retain stock in. In addition, the shareholders can benefit from a spin-off in that it could encourage the firm to engage in investment opportunities that would otherwise not be explored. Rosenfeld (1984) further examines the impact of spin-offs on 4
shareholder wealth and compares it with that of sell-offs. Either of these corporate actions could signal information about the firm s financial condition. A sell-off could be interpreted in a positive way: the parent firm will gain financially from the transaction and would be able to streamline its operations. Such positive perception could favorably influence the stock price. A sell-off could be viewed negatively, however, if it is perceived that a firm needs the funds generated by the transaction just to maintain its current operations. In his empirical work, the author examines the stock performance surrounding 62 sell-off announcements and 35 spin-off announcements. Comparing the cumulative average adjusted returns corresponding to the period day 10 through day 0, for both spin-offs and sell-offs, there is a positive and statistically significant effect on stock prices during this announcement period. In addition, the differential between spin-off and sell-off returns indicates that there is a stronger positive effect on price for firms that spin off a subsidiary versus selling off. The author suggests that even though a sell-off infuses the firm with capital, such a consequence could have had a negative signaling effect. Hite and Owers (1983), studying a sample of 123 spin-offs, also observe on average large abnormal returns for the parent firms. Further examining these announcements and the reasons that a firm gives for initiating this divestiture, the authors find varying market reactions. The stock price gains accrue most for parent firms that seek to engage in a formative corporate action, such as a merger, after the spin-off. Interestingly, for those companies that must divest a subsidiary for legal and regulatory reasons, an average negative abnormal return is found. Insurer Equity Carve-outs Through a divestiture of a non-related subsidiary, a parent firm can become more focused on its core lines of business, an argument presented in many of the divestiture papers. This 5
allows for the firm to engage in activities which would better maximize its value since resources do not have to be directed elsewhere. The financial market, recognizing this future value that will be created, could drive up the share price at the time of announcement. In the insurance industry, multi-line insurance groups are a common corporate structure; these groups have also executed numerous equity carve-outs over the past two decades. While on the surface, this could be viewed an intra-industry divestiture, the difference in lines of business, with its corresponding underwriting complexities and uniqueness, could also support the non-relatedness between parent and subsidiary argument. A property-liability parent firm may reduce its ownership of a life-health subsidiary to reasonably focus on its core competencies. In a study examining seasoned equity and debt offerings by insurers, Akhigbe, Borde and Madura (1997) observe that there is no significant market reaction to equity offerings. This is in contrast to the negative and significant abnormal returns found in other seasoned equity offering (SEO) studies, including Schipper and Smith (1986). The authors believe that while a SEO is generally interpreted as a negative signal of the firm s prospects, the heavy monitoring and regulation in the insurance industry dilutes that signal. It will be interesting to observe the market reaction to insurance industry carve-outs. Hulburt, Miles and Woolridge (2002) empirically test the divestiture gains and asymmetric information hypotheses for carve-outs; they argue that when the parent and subsidiary are in the same industry, the implications from these hypotheses on the effect of a carve-out are ambiguous. In this research, we will investigate this effect within the insurance industry. DATA AND ANALYSIS Searching on the Securities Data Company (SDC) New Issues Database and searching IPOs of common stock during the period 1991-2005, we identify 19 initial public offerings of 6
insurance companies that are equity carve-outs of publicly traded parent companies. These firms are classified under insurance Standard Industry Classification (SIC) codes. 1 Examining the company histories listed in Bests Insurance Reports supplemented with searches in Lexis-Nexis, additional details of the transactions are be gathered. Of primary importance is the degree to which the carve-out operates independently from the parent after the IPO. In addition, the reasons for the carve-out transaction will be investigated. For most of these 19 carve-outs, the parent company, according to its SIC, is also an insurance company. Table 1 lists the parent company and its subsidiary carve-out along with the announcement date and IPO date for the transaction. Table 1 Insurance Company Carve-outs, Announcement and IPO Dates, 1991-2005 Subsidiary Parent Announcement Date IPO Date AMBAC Inc Citicorp 03/05/90 07/11/91 Guaranty National Corp Orion Capital Corp 09/13/91 11/14/91 CCP Insurance Inc Conseco Inc 04/24/92 07/14/92 CMAC Investment Corp Reliance Group Holdings 08/24/92 10/30/92 First Colony Corp Ethyl Corp 09/11/92 12/08/92 Bankers Life Holding Corp Conseco Inc 02/11/93 03/26/93 TIG Holdings Inc TransAmerica Corp 07/20/92 04/20/93 Allstate Corp Sears Roebuck & Co 09/29/92 06/03/93 Paul Revere Corp Textron Inc 08/12/93 10/27/93 Vesta Insurance Group Inc Torchmark Corp 07/23/93 11/11/93 Gryphon Holdings Inc Willis Corroon Group 09/24/93 12/21/93 Western National Corp Conseco Inc 09/13/93 02/09/94 PMI Group Inc Allstate Corp 01/17/95 04/11/95 HealthPlan Services Corp Noel Group Inc 03/21/95 05/19/95 American States Financial Corp Lincoln National Corp 03/14/96 05/23/96 Hartford Life ITT Hartford Group Inc 02/10/97 05/23/97 Travelers Ppty Casualty Corp Citigroup Inc 12/19/01 03/22/02 Platinum Underwriters Holdings St Paul Cos Inc 04/15/02 10/29/02 Genworth Financial Inc General Electric Co 11/18/03 05/25/04 1 SIC codes 6311, 6321, 6324, 6331, 6351. 7
As the table above shows, there was heavy carve-out activity in the early 1990s, perhaps corresponding to increased risk-based capital regulations that went into effect. This focus on adequate capital, following the insolvencies of the 1980s, could have driven this movement towards streamlined operations and the creation of a tradable stock, to be able to tap the capital market when needed. The announcement date for the transaction is identified through researching Dow Jones wire reports; the date on which the firm states its intention to issue an IPO in the subsidiary is denoted as day 0. Using daily stock price data from Center for Research in Security Prices (CRSP), the abnormal return for the parent company will be calculated for the period surrounding the announcement date. This will provide information on the impact a carve-out announcement has on the parent company, conveying how the market perceives this company action. Further evidence on the market valuation effects of these spin-offs will be determined by calculating the initial stock returns for the subsidiary once its IPO is executed. These returns will be compared with the corresponding parent stock returns over the same time period. It will be interesting to observe whether issuance of a spin-off IPO was value enhancing. Examining the public statements made by these parent firms at the time of announcement, several major motivations are given. A primary reason for the carve-out is to allow the parent to focus on its core operations. Indeed, for the last carve-out listed in Table 1, according to its news releases, General Electric wished to divest itself completely of its insurance operations, thus renaming that unit Genworth Financial and issuing its IPO in 2004. This motivation was similar for Sears and its subsequent carve-out of Allstate. 8
Table 2 Returns for the Insurance Subsidiary and Parent Firm on the IPO Date Corresponding Subsidiary Parent IPO Date First-Day Return for Subsidiary First-Day Return for Parent AMBAC Inc Citicorp 07/11/91 0.0188-0.0776 Guaranty National Corp Orion Capital Corp 11/14/91 0-0.0075 CCP Insurance Inc Conseco Inc 07/14/92 0.0167-0.0474 CMAC Investment Corp Reliance Group Holdings 10/30/92 0.0556-0.0250 First Colony Corp Ethyl Corp 12/08/92 0.0759-0.0126 Bankers Life Holding Corp Conseco Inc 03/26/93 0.1136 0.0018 TIG Holdings Inc TransAmerica Corp 04/20/93 0.0771-0.0277 Allstate Corp Sears Roebuck & Co 06/03/93 0.0880-0.0225 Paul Revere Corp Textron Inc 10/27/93 0 0.0179 Vesta Insurance Group Inc Torchmark Corp 11/11/93 0 0.0030 Gryphon Holdings Inc Willis Corroon Group 12/21/93 0-0.0078 Western National Corp Conseco Inc 02/09/94 0.1146 0.0064 PMI Group Inc Allstate Corp 04/11/95 0.1029-0.0043 HealthPlan Services Corp Noel Group Inc 05/19/95 0.1964 0.0000 American States Financial Corp Lincoln National Corp 05/23/96 0-0.0052 Hartford Life ITT Hartford Group Inc 05/23/97 0.1372-0.0079 Travelers Ppty Casualty Corp Citigroup Inc 03/22/02 0.0573 0.0048 Platinum Underwriters Holdings St Paul Cos Inc 10/29/02 0.1107-0.0269 Genworth Financial Inc General Electric Co 05/25/04 0 0.0140 MEAN 0.0613-0.0118 MEDIAN 0.0573-0.0075 9
Analysis To estimate the daily prediction error (which has also been termed the abnormal return in other research), two methods typically utilized will be followed here for the 18 parent firms. 2 Again, the announcement date for the carve-out is denoted as day 0. Under the market model approach, the daily returns for the 120-day period corresponding to day 240 through day 121 are used to estimate the parameters of the market model, as shown in equation (1) Rit = ai + bi Rmt + eit for t = 240, 239,,, 121 (1) R it represents the stock return for parent firm i on day t, R mt denotes the return on the corresponding day t for the CRSP value-weighted index including distributions. The parameters, a i and b i, are used to then forecast firm i s daily stock return for the following period, days 120 through day +30. Following Schipper and Smith (1986), the daily prediction error, u it, is computed for each of these forecasted returns, as shown in equation (2). uit = Rit ( ai + bi Rmt ) for t = 120, 119,... + 30 (2) For each day t, the prediction errors are averaged across the 18 firms and can be denoted as PE t and cumulative average prediction errors are also computed, according to equation (3). 2 For one carve-out, Bankers Life, the parent firm Conseco became its owner only several months prior to the carve-out. Given that the market model requires parent firm returns for the 240-day period prior to announcement, the actual impact of the carve-out cannot be reasonably estimated. 10
t CPEt = PE j for t = 44, 43,..., + 30 j = 44 (3) The 75-day period surrounding the announcement date is of interest, to determine the wealth effect of the carve-out announcement on the parent firm. As a comparison, the daily prediction errors will also be computed using a mean-adjusted approach, a methodology used in Miles and Rosenfeld (1983) and Rosenfeld (1984). Here, the average daily return for each firm i over the days 240 through 121 is calculated, which represents the expected return ER i. Thus, for day 120 through +30, the daily prediction error u it is calculated as uit = Rit ERi for t = 120, 119,... + 30 (4) The average and cumulative daily prediction errors are computed the same way as in the market model approach. RESULTS The daily prediction errors for the parent firms are estimated according to the procedure outlined in the previous section. Table 3 reports the estimated daily prediction errors and the cumulative prediction errors for the period surrounding the announcement date. As shown, over this 75-day period, the parent firms experience a 5.65% average decrease in the expected stock price under the market model, a marked contrast to the results of Schipper and Smith (1986) and other studies which have found a positive abnormal return for carve-outs over such a time period. Using a mean-adjusted approach, the cumulative prediction error is on the same order, 6.14%, across this sample. 11
Table 3 Prediction Errors and Cumulative Prediction Errors for 19 Equity Carve-out Announcements Market Model Mean-Adjusted Day Daily Prediction Error Cumulative Daily Prediction Error Daily Prediction Error Cumulative Daily Prediction Error -44-0.0032-0.0032-0.0033-0.0033-40 0.0053 0.0043 0.0036 0.0072-35 -0.0010 0.0148-0.0045 0.0057-30 -0.0042-0.0007-0.0016 0.0000-25 -0.0048-0.0160-0.0030-0.0161-20 -0.0036-0.0175-0.0026-0.0182-15 0.0011-0.0268 0.0001-0.0274-10 -0.0032-0.0366-0.0005-0.0346-9 0.0060-0.0306 0.0038-0.0308-8 -0.0043-0.0349-0.0042-0.0350-7 -0.0075-0.0424-0.0063-0.0413-6 -0.0018-0.0442-0.0026-0.0439-5 0.0099-0.0343 0.0100-0.0339-4 0.0127-0.0216 0.0084-0.0255-3 -0.0056-0.0273-0.0062-0.0317-2 -0.0023-0.0296-0.0024-0.0341-1 0.0069-0.0227 0.0094-0.0248 0 0.0111-0.0116 0.0087-0.0161 1 0.0097-0.0018 0.0124-0.0037 2 0.0018 0.0000 0.0021-0.0016 3 0.0010 0.0010 0.0028 0.0012 4-0.0026-0.0016-0.0013-0.0001 5 0.0002-0.0014-0.0008-0.0009 6-0.0045-0.0059-0.0041-0.0050 7 0.0001-0.0057-0.0006-0.0056 8-0.0026-0.0083-0.0004-0.0059 9 0.0030-0.0054 0.0032-0.0027 10-0.0003-0.0057 0.0005-0.0022 11-0.0047-0.0104-0.0043-0.0065 12-0.0015-0.0120-0.0033-0.0099 13 0.0035-0.0085 0.0019-0.0080 14 0.0030-0.0055 0.0015-0.0064 15-0.0044-0.0099-0.0075-0.0140 20 0.0011-0.0301 0.0015-0.0349 25 0.0025-0.0455 0.0016-0.0456 30-0.0048-0.0565-0.0074-0.0614 12
Even though over the 75-day period, there is a negative cumulative impact on the stock price, one can observe that the average daily prediction error fluctuates. The average prediction error is positive on day 0 and during the several days after that, suggesting that the market is valuing these parent firms higher than what is expected. To further examine these fluctuations, the cumulative prediction error over various subperiods will be analyzed, focusing on the market model approach. In Table 4, the cumulative daily prediction errors are listed for certain intervals, which aim to capture the stock price effect from informational flows. The t-statistics are calculated using the procedure outlined in Schipper and Smith (1986). Table 4 Cumulative Prediction Errors for 18 Insurance Carve-outs Days Relative to Announcement Cumulative Daily Prediction Error t-statistic 44 through +30-0.0565-1.3189 44 through 11-0.0334-1.1560 10 through 1 0.0106 0.6805 10 through +10 0.0276 1.2176 1 through 0 0.0180 2.5753** 1 through +1 0.0278 3.2387** 0 through +1 0.0209 2.9838** 0 through +10 0.0170 1.0336 + 2 through +10-0.0339-0.2639 + 11 through +30-0.0508-2.2946* Note: *, ** denotes significance at the 0.05 and 0.01 levels, respectively. Again, over the entire 75-day period, there is a negative wealth effect of carve-out announcements on the parent firms, however not significant. Focusing on the days prior to the announcement date, the deviations from expected returns do fluctuate. The run-up in price is evident during the interval [-1,+1] day by day. During this period, there is a strongly positive 13
and significant effect of carve-out announcements on the parent firms stock price. This could be due to the optimism towards breaking up the conglomerate into entities that can now be separately valued for its future opportunities. If the firm is divesting itself of a subsidiary that does not follow its planned focus for the future, a reason given by many of the companies in our sample, the market may be valuing this action. Examining the various time periods post-announcement, the positive run-up in price does get diluted. Indeed, over the time periods [+2,+10] and [+11,+30], a negative cumulative prediction error is observed and in that latter period, with an error of 5.08%, is found to be significant. While this will be investigated further in later research, this could be effect of a reversal of initial overoptimism about the carve-out. In many of the IPO studies, the high initial returns are followed by negative returns; an explanation commonly given is that the market is now properly valuing the firm, after the excitement in the early trading days to own the stock drove up the price. CONCLUSIONS In this paper, the market reaction to insurance company carve-out announcements is investigated. A parent firm may wish to divest itself of a subsidiary for numerous reasons, such as needing an infusion of capital or ridding itself of lines of business outside its core focus. In these carve-outs, an initial public offering of the subsidiary is executed. It is found that the average one-day return for these IPOs is positive, while the average return for the parent firms on that first-day is negative. Future expansion of this paper will include further analysis of these initial returns. 14
Using the market model to estimate the impact of these carve-out announcements on the parent firms, over the 75-day period surrounding the announcement date, a negative and significant abnormal return is observed. Further dissecting this time period, for the three-day period around the announcement date, there is a positive and strongly significant impact on the stock price. It is this run-up in price that gets diluted over the immediate days following. Whether this market reaction exists for insurance company spinoffs and selloffs as well will be investigated in the next phase of this research. REFERENCES Akhigbe, A., S. Borde and J. Madura, 1997, Valuation Effects of Insurers Security Offerings, Journal of Risk and Insurance, 64: 115-137. Alexander, G., P. Benson and J. Kampmeyer, 1984, Investigating the Valuation Effects of Announcements of Voluntary Corporate Selloffs, Journal of Finance, 39: 503-517. Allen, J. and J. McConnell, 1998, Equity Carve-outs and Managerial Discretion, Journal of Finance, 53: 163-186. Aron, D., 1991, Using the Capital Market as a Monitor: Corporate Spinoffs in an Agency Framework, RAND Journal of Economics, 22: 505-518. Hite, G. and J. Owers, 1983, Security Price Reactions around Corporate Spin-off Announcements, Journal of Financial Economics, 12: 409-436. Hogan, K. and G. Olson, 2004, The Pricing of Equity Carve-outs during the 1990s, Journal of Financial Research, 27: 521-537. Hulburt, H., J. Miles and J. Woolridge, 2002, Value Creation from Equity Carve-outs, Financial Management, 83-100. Johnson, G., R. Brown and D. Johnson, 1994, The Market Reaction to Voluntary Corporate Spinoffs: Revisited, Quarterly Journal of Business and Economics, 33: 44-59. Miles, J. and J. Rosenfeld, 1983, The Effect of Voluntary Spin-off Announcements on Shareholder Wealth, Journal of Finance, 38: 1597-1606. Nanda, V., 1991, On the Good News in Equity Carve-outs, Journal of Finance, 46: 1717-1737. 15
Rosenfeld, J., 1984, Additional Evidence on the Relation Between Divestiture Announcements and Shareholder Wealth, Journal of Finance, 39: 1437-1448. Schipper, K. and A. Smith, 1986, A Comparison of Equity Carve-outs and Seasoned Equity Offerings, Journal of Financial Economics, 15: 153-186. Slovin, M., M. Sushka and S. Ferraro, 1995, A Comparison of the Information Conveyed by Equity Carve-outs, Spin-offs, and Asset Sell-offs, Journal of Financial Economics, 37: 89-104. Vijh, A., 2002, The Positive Announcement-Period Returns of Equity Carveouts: Asymmetric Information or Divestiture Gains?, Journal of Business, 75: 153-190. 16