Australian divestiture activity: an examination of gains to sell-off announcements.

Abstract:

This study examines the source of gains associated with Australian divestiture activity, defined as a voluntary modification of the firm "s productive assets by a sell-off of a complete operating division or wholly-owned subsidiary of the divesting firm. The sell-oil announcement produces positive average abnormal returns of 1.15% over the two-day announcement period. We conclude that the gains arise predominantly from divestitures that have a strategic focus as demonstrated by, first, the divested unit is unrelated to the firm's core activities (a strategic divestiture), second, the significance of the strategic variable in explaining the positive market reaction in regression analysis, and, third, the finding of more significant results' where the intended use of proceeds of the sell-off is for strategic purposes.

Keywords:

DIVESTITURES; STRATEGIC GAINS; SELL-OFFS.

1. Introduction

There is much evidence from the United States that the announcement of a sell-off is associated with positive abnormal returns. (1) Outside the US, however, this topic has received little attention. To our knowledge, only four non-US-based divestiture studies are reported: Kaiser and Stouratis (1995) with French, German, Swedish and UK data; Afshar, Taffler and Sudarsanam (1992) with UK data; Hamilton and Chow (1993) with New Zealand data; and one Australian study (Capon, Christodolou, Farley & Hulbert 1987) based on survey data and concluding that there is less divestiture activity in Australia than in the US.

The large number of US divestiture studies is accompanied by an equally large number of attempts to explain the observed gains, warranting comment that indeed 'no consensus has been reached as to the source of these gains' (Slovin, Sushka & Ferraro 1995, p. 90). This lack of an accepted theoretical framework motivates our attempt to organise the various theories into distinct paradigms in order to examine the source of the gains. We classify the existing explanations for divestiture activity into two groups, a strategic group and a financial group

We test a sample of divestitures defined as sell-offs of either a complete operating division or a wholly owned subsidiary of the divesting firm. (2) Analogous to the US studies, our tests show that the market reacts positively to Australian sell-off announcements, with average abnormal returns of 1.15% over a two-day announcement period. When the sample is segregated into strategic divestitures (the disposal of units unrelated to the firm's core activities) and non-strategic divestitures (the disposal of related units) the strategic divestitures are associated with a significant average abnormal return of 1.71%. In contrast, returns accruing to the non-strategic group are insignificant.

Using regression analysis we test the comparable strength of a strategic variable (based on whether the divested unit is a related unit or an unrelated unit) and financial variables (using Altman's Z-score to proxy for financial distress) in explaining the market reaction. Only those models including the strategic variable have statistically significant explanatory power, with the strategic variable being the only statistically significant variable. The regression analysis does not support the financial distress theories of divestiture, a finding at odds with some US studies (Afshar, Taffler & Sudarsanam 1992; Denning 1988; Lang, Poulsen & Stulz 1995).

Examination of the stated use of proceeds of the sell-off reveals that higher and more significant abnormal returns (3.0%) are observed for firms that intend to use the proceeds for strategic purposes, compared with insignificant returns for firms that fail to disclose the intended use. The market reaction to the latter may be driven by a fear that, without public disclosure, opportunistic management is more likely to misappropriate the proceeds (Jensen & Ruback 1983). A lower (1.8%), and only marginally significant abnormal return accrues to firms intending to use the proceeds to repay debt. When strategic divestitures are grouped on the basis of the disclosed use of proceeds, the largest abnormal return of 4.3% accrues to the group in which the intended use of proceeds is also strategic, with lower and insignificant returns recorded for the debt group, 1.93%, and the undisclosed use group, 0.94%. Of the non-strategic divestitures, no group on the basis of the intended use of proceeds has significant announcement-period returns. The study provides evidence that, on average, only strategic sell-offs of unrelated units are positively valued by the Australian market.

In the next section, we review the empirical evidence on sell-off activity and the explanations proposed to explain the positive abnormal returns that have been found on announcement.

In section 3, we describe our data and the methodology adopted. Our results are in section 4 and concluding remarks follow in section 5.

2. Divestiture Activity and Competing Rationales

2.1 Divestitures and Abnormal Returns

Divestitures classified as sell-offs occur when the parent (divesting) company sells part of its assets for cash, securities or assets. An act of divestiture is a major structural activity so that, for the purpose of these definitions, the term 'assets' means both whole operating divisions or subsidiary companies. Table 1 summarises the US studies on the market reaction to the announcement of sell-offs. It documents an average positive abnormal return of around 1% accruing to the divesting firm over the two-day announcement period, day -1 and day 0 in event time, where day 0 is the day of the reported divestiture announcement in the Wall Street Journal.

Consistent evidence of positive abnormal returns associated with announcements of sell-offs in the US and Europe suggests that sell-offs in Australia will also result in positive abnormal returns to the divesting firm. That divestitures are widely publicised events (3) and that one of management's primary motives for divesting is to increase the company's share price (Hamilton & Chow 1993; Montgomery, Thomas & Kamath 1984) further support the expectation of positive announcement returns. Our first hypothesis is, therefore, that positive abnormal returns will accrue to Australian firms upon the announcement of a sell-off.

2.2 Theories to Explain Divestiture Gains

Numerous explanations have been proposed to explain the source of divestiture gains. These include: the wealth transfer hypothesis (Galai & Masulis 1976); the information asymmetry hypothesis (Krishnaswami & Subramaniam 1999); the focus hypothesis (Daley, Mehrotra & Sivakumar 1997; John & Ofek 1995; Markides & Berg 1992); the fit hypothesis (John & Ofek 1995); the tax-timing option (Mauer & Lewellen 1990); the contracting efficiency hypothesis (Hite & Owers 1983); the removal of an under-performing unit (Cho & Cohen 1997); and the provision of cheap funds (Sicherman & Pettway 1987). Other studies (e.g. Roy & Manley 1997) re-test many of these theories. The majority of these explanations can be categorised into two broad groups: a strategic group and a financial group. (4)

The first category, the 'strategic hypothesis' or the 'corporate focus hypothesis' encompasses such theories as the focus hypothesis, the fit hypothesis, the contracting efficiency hypothesis and the tax-timing option. There are three central propositions in the strategic literature. First, the divestiture will eliminate negative synergies between the divesting firm's operations and the divested unit. This leads to more efficient allocation of management time, enabling the corporation to focus on businesses in which it has a competitive advantage and remove assets which interfere with other operations (John & Ofek 1995; Markides & Berg 1992). A number of studies suggest that a firm focused in one specialised area is more likely to outperform a more diversified firm (Berger & Ofek 1995; Bhagat, Schleifer & Vishny 1990; Comment & Jarrell 1995).

Second, the increase in focus will increase investors' flexibility in managing their own portfolios, consistent with the notion that the diversification function is most efficiently achieved by shareholders (Levy & Sarnat 1970). This additional flexibility is associated with the third proposition which is a tax-timing option. Instead of losses and gains being netted oft, by separating variable operating units of the firm the capital losses can be realised immediately and the capital gains can be deferred (Constantinides 1984; Mauer & Lewellen 1990). The tax-timing option is an increasing function of the variance between the cash flows of the two divisions. The third proposition is that divestiture will eliminate concerns that unrelated units are cross-subsidising the performance of other units (Berger & Ofek 1995).

To test the strategic theory, we follow previous studies and distinguish sell-offs according to the comparability of the divested unit's operations with the divesting firm's operations (Daley, Mehrotra & Sivakumar 1997; Hite & Owers 1983; John & Ofek 1995). Our second hypothesis predicts that the divestiture of an unrelated unit is more value-increasing than the divestiture of a related unit, as the former increases firm focus and hence firm value.

The second category of explanations of divestiture gains can be termed the financial hypothesis, that firms experiencing higher levels of financial distress pre-divestiture compared to other firms will realise larger gains from the divestiture process. Many have observed that often divestitures are undertaken by firms in financial distress. Alexander, Benson and Kampmeyer (1984) and Jain (1985) found that divestiture decisions tend to be preceded by a period of poor market performance. Furthermore, both high pre-distress leverage (Afshar, Taffler & Sudarsanam 1992; Ofek 1993) and poor performance relative to industry counterparts (Montgomery & Thomas 1988) have been shown to increase the probability of operational actions including divestitures. (5)

This literature proposes two explanations of the gains observed by financially distressed firms. As financially distressed firms have more limited access to capital markets, the divestiture provides a 'cheap source of funding' to be used in the pursuit of other projects and, therefore, results in abnormal gains (Lang, Poulsen & Stulz 1995; Sicherman & Pettway 1987). The second reason is that, if the poor performance of the firm is driven by an under-performing unit, the subsequent disposal of this unit should result in a positive share price reaction (Cho & Cohen 1997; Denning 1988).

However, support for the financial theories of divestiture gains is weakened by findings of no relationship between pre-divestiture financial distress and observed returns (Desai & Jain 1999; Miles & Rosenfeld 1983). Furthermore, if the financial motive is to use the proceeds to repay debt, in some situations this could be viewed as a weakening of the corporate governance function of debt and be discounted by the market accordingly (Williamson 1988). While it is expected that both the strategic theories of divestiture and the financial theories of divestiture explain part of the event period abnormal returns, we attempt to test the relative explanatory power of each. The strategic and financial distress proxies are discussed subsequently.

As an additional test of the strength of the two theories, and motivated by qualitative studies which report share price concerns as one of the most influential conditions affecting divestment decisions (Hamilton & Chow 1993; Montgomery, Thomas & Kamath 1984), we examine the effect of disclosure of the use of proceeds associated with the divestiture decision. The sample is categorised according to the stated use of proceeds from the sell-off, and the announcement effect of each category is measured.

Two studies examine announcement contents to assess divestiture activity. Adopting the financing hypothesis (via the 'cheap funding' rationale), Lang, Poulsen and Stulz (1995) examine 93 sell-offs made by 'financially healthy' firms, distinguishing between firms that use the proceeds to pay down debt and firms that retain the proceeds. Extrapolating from Jensen's (1986) free cash flow hypothesis, Lang, Poulsen and Stulz argue that the market will discount sale proceeds that are retained by the firm, as they may be misappropriated by opportunistic management, and reward the repayment of debt. They find that the majority of the abnormal returns accrue to the 'payout sub-sample' (3.92%) rather than the 'retain subsample' (-0.48%) and sensitivity analysis shows that the payout dummy variable has incremental explanatory power over the financial health of the divesting firm. Lang, Poulsen and Stulz conclude that sell-offs are only value-increasing for firms which apply the proceeds to reduce debt.

Brown, James and Mooradian (1994) refine the sell-off sample to only those firms that previously had experienced financial distress. Based on the notion of equity as a call option on the firm, with an exercise price equal to the value of outstanding debt, management of financially distressed firms would prefer to retain the proceeds of a sell-off rather than use it to payoff debt and extinguish the shareholders' 'option'. The results show that the difference between abnormal returns of financially distressed firms that repay debt and those that use the proceeds for other purposes is 3.17%, statistically significant at 0.05.

Overall, these studies imply that healthy firms are rewarded for paying out the proceeds of an asset sell-off while financially distressed firms are rewarded for retaining the proceeds. One criticism is that in limiting their division of the sample to those firms that repaid debt and those that did not, both Lang, Poulsen and Stulz (1995) and Brown, James and Mooradian (1994) narrow the theoretical paradigm to consideration only of the financial theory of divestiture gains.

To expand this research, this study identifies three categories of use of proceeds: to further other core operations or specific investment areas (based on the strategic theory); to reduce debt (based on the financial distress theory); and a residual category to accommodate those announcements which do not indicate how proceeds are to be used. Based on costs associated with uncommitted free cash (Jensen & Ruback 1983), sell-offs that state the intended use of proceeds are expected to result in larger abnormal returns than sell-offs that are silent on the issue. Further, given the consistently strong support for focus increasing divestitures over and above financially motivated divestitures (John & Ofek 1995; Roy & Manley 1997), the opportunistic concerns of the latter divestitures and the removal of the governance functions of debt by repaying it (Williamson 1988), our third hypothesis is that firms using the proceeds for strategic investment will have significantly larger returns than firms using proceeds to reduce debt, both of which will be larger than the returns accruing to firms that do not state the intended use of divestiture proceeds.

3. Data and Methodology

3.1 Data

The sample comes from all voluntary sell-offs announced by Australian listed companies between 1 July 1994 and 1 July 2000. To be included in the study the sell-off must be voluntary and the divested unit must be a complete operating division or a wholly-owned subsidiary of the company. These requirements eliminate property sales, sale and leaseback arrangements and sales of non-current assets separate from the business for which those assets are normally employed. Furthermore, due to their involuntary nature, asset disposals made under a scheme of arrangement, under a court direction, or during takeover activities are excluded. (6)

The continuous disclosure provisions of the Australian Stock Exchange (ASX) Listing Rule 3.1 require divestitures with a value in excess of 5% of the divesting company's value to be disclosed. A preliminary search of the Securities Industry Research Centre of Asia-Pacific's (SIRCA) Signal G announcement file of 'Asset Disposals', for the period July 1994 to July 2000, produced 410 divestiture events.

Excluding spin-offs (5), divestitures by dual listed foreign companies (22) or by companies with contemporaneous confounding announcements, for example, profit announcements (116), the sample consists of 267 Australian divestitures characterised as sell-offs. This sample of 267 was examined for the market reaction to announcement of the sell-off. Further analyses include only those companies which traded on both event day 0 and event day +1 (223).

A review of the various characteristics of divesting firms highlights two features. First, the financial position of the divesting firms demonstrates no evidence of financial distress, as evidenced by the positive Altman Z-scores (7) (mean = 2.7986, median = 0.9141), current ratios of 1.5485 (1.3266), and standard debt ratios of 0.5677 (0.5587). Second, the relative size of the divested unit is substantial, averaging 17.37% of the divesting firm's market capitalisation, with a median value of 3%.

Prices and dilution factors for the divesting firm as well as the value of the Australian All Ordinaries Accumulation Index on the same day were collected from the SIRCA database for the event period day -6 to day +5, where the day of announcement is day 0 in event time.

To distinguish between firms divesting for strategic reasons and those divesting for financial distress reasons, we collected financial statement data from both Connect 4 and Datastream. (8) Where appropriate, adjustments for substituted accounting periods and the release of interim financial statements were made. Market capitalisation values in the month prior to announcement were extracted from the Centre for Research in Finance (CRIF) database and Shares magazine.

Divestiture-specific data, namely the nature of the divested unit's operations, the value of the divestiture and the company's classification of the divestiture as a core unit or non-core unit, was also sourced from the SIRCA Signal G Announcements database. Where the company is still in existence, its industry classification came from the ASX Website listings. For companies no longer in existence, the industry classification was generated from the descriptive operative information provided by the announcement in accordance with the ASX Industry Codes. (9) These industry classifications also provide the basis for distinguishing between focus-increasing divestitures (sell-offs of unrelated units, 131 units) and non-focus-increasing divestitures (sell-offs of related units, 92 units).

The texts of all the SIRCA divestment announcements were examined to distinguish between firms that intend to use the proceeds for strategic purposes, (10) firms that intend to use the proceeds to reduce debt, (11) and firms that do not identify how the proceeds are to be used. Where multiple uses of proceeds are given, the first stated purpose or the purpose indicated in the news report as the primary purpose of the sale is used (Brown, James & Mooradian 1994). (12)

3.2 Event-Time Methodology

The market reaction to the sell-off announcement is tested with event study methodology (Brown & Warner 1985). Daily time intervals are used and the event date (t = 0) is defined as the day upon which the divestiture is announced to the market. A two-day event period is chosen as more powerful than a longer window (MacKinlay 1997, p. 35) and to reduce the sensitivity of the abnormal returns to the benchmark choice (Desai & Jain 1999). In this study, day 0 is the day on which the divestiture announcement (Signal G) is made to the market; to cater for firms making a divestiture announcement after the close of trading on day 0 our event window includes day 0 and day +1.

Stock price performance is measured by computing the abnormal return for each firm i for event days -5 to +5 using the market-adjusted model which has been shown to be just as powerful in detecting abnormal performance in daily returns as the more complex risk-adjusted models (Brown & Warner 1985). The market is proxied by the Australian All Ordinaries Accumulation Index. (13)

(1) ARit = Rit - Rmt

Average abnormal returns for each relative event day t are calculated by averaging across the N securities traded on that day:

(2) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.]

These are cumulated over the event window ([T.sub.1] to [T.sub.2]) to form a conventional Cumulative Abnormal Return (CAR) for the period:

(3) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.]

Conventional t statistics are used to determine the statistical significance of the ARs and CARs.

3.3 Multiple Regression 

Regression is used to examine the effects of the strategic and financial variables on the abnormal returns observed at announcement of a sell-off, alter controlling for divestiture size. The strategic variable aims to quantify the increase in focus, or decrease in diversity, of a firm's operations. The literature employs a number of strategic measures, including the change in the number of business lines reported by the divesting company (Desai & Jain 1999; John & Ofek 1995); a dummy variable equal to one when the divested unit has a different 2-digit SIC code to the divesting company (Daley, Mehrotra & Sivakumar 1997; Desai & Jain 1999; John & Ofek 1995; Kaplan & Weisbach 1992); and the Herfindahl index, which measures the concentration of sales across business segments (Desai & Jain 1999; John & Ofek 1995). Due to the flexibility afforded to management under the Australian segment reporting standard (AASB1005), the only reliable and comparable strategic variable available is the industry dummy variable which is the ASX industry code dummy variable ([D.sub.IND]). Industry classifications were assigned with strategic divestitures (or sell-offs of units with different ASX industry codes) coded as [D.sub.IND] = 1 and non-focus-increasing (or sell-offs of units with equivalent ASX industry codes) coded as [D.sub.IND] = 0.

The financial distress theory argues that a decrease in financial distress is only advantageous to firms with a previously high financial distress level, as for 'healthy' firms a further decrease in financial pressure may increase the propensity for management to expropriate the wealth of shareholders (Jensen & Ruback 1983) due to the reduction of the debt governance function (Williamson 1988). To test this proposition we use a single measure of financial distress, Altman's Z, which measures the firm's insolvency potential (Altman 1968). The original formula for the Altman Z-score is employed. (14)

The Z-score is negatively correlated with the probability of going bankrupt, therefore, the lower the score the more financially distressed is the firm. The inputs for Altman's Z are obtained from the last available published accounts prior to the divestiture.

As numerous studies have shown a significant positive relation between the relative size of the divestiture and the abnormal returns observed (Hearth & Zaima 1986; Kaiser & Stouratis 1995; Klein 1986; Schipper & Smith 1983), the following size control is included in the regression:

(5) SIZE = ln([V.sub.D]/[V.sub.i])

where: [V.sub.D] = value ($) of the divested unit; and

[V.sub.i] = market capitalisation of the divesting firm in the month prior to divestiture.

The regression to test the comparative strengths of the strategic and financial theories is:

(6) C A[R.sub.0,1] = [alpha] + [[beta].sub.1] [D.sub.IND] + [[beta].sub.2] ALTMANZ + [[beta].sub.3] SIZE + [epsilon]

where: CA[R.sub.0,1] = cumulative abnormal return for each divestiture over event days 0, + 1 ;

[D.sub.IND] = industry code dummy variable for divestiture ([D.sub.IND] = 1 if strategic);

ALTMANZ = Z-SCORE for divesting firm prior to divestiture; and

SIZE = log of the relative size of divestiture.

The [D.sub.IND] and SIZE co-efficients are expected to be positive and the Z-SCORE coefficient to be negative.

4. Results

4.1 Abnormal Returns

The tests of the market reaction to the announcement of a sell-off are summarised in table 2. (15). The only significant abnormal returns over the observed window occur on event days 0 and +1, at the 0.10 and 0.05 levels respectively. The CAR for the announcement period, days 0 and +1, is 1.15% with a t-statistic of 2.816, supporting the first hypothesis that on average sell-off announcements by Australian firms produce positive abnormal returns. This is consistent with the US sell-off literature which reports positive abnormal returns in the vicinity of 1% (John & Ofek 1995; Roy & Manley 1997).

In the first test of the strategic focus theory, the sample is divided according to whether the divestiture is of a related unit ([D.sub.IND] = 0) or an unrelated unit ([D.sub.IND] = 1). The results reported in table 3 support the strategic hypothesis. Over the announcement period, days 0 and +1, abnormal returns of 1.71% (p < 0.01) are observed for non-related (strategic) divestitures while the returns accruing to non-strategic divestitures are not significantly different from zero. While this finding is analogous to that of Daley, Mehrotra and Sivakumar (1997) who also found that only the 'strategic' sample produced significant returns, it is contrary to Desai and Jain (1999) who find that both sub-samples accrue significant positive returns.

4.2 Competing Theories

Multiple regression analysis is employed to test whether both strategic characteristics and financial characteristics explain divestiture gains, with the former exhibiting superior explanatory ability. Based on initial tests for multicollinearity and autocorrelation which show that the regression assumptions have not been violated, a total of five models are examined, all of which include a size variable as a control (Hearth & Zaima 1986; Klein 1986). The results are presented in table 4.

The strongest finding is that the strategic variable, the industry code dummy, is significant in each of the three models in which it is included. Furthermore, these three models are the only regressions with statistically significant F-statistics. This supports the proposition that strategic characteristics explain a significant amount of divestiture gains and that the strategic hypothesis provides the strongest explanation for the observed abnormal returns.

Conversely, the financial hypothesis of divestiture gains is not supported by the regression results. Initially, Altman's Z-score is used as the measure of financial distress. As the Altman Z-score is negatively correlated with the probability of bankruptcy, and divestitures should be valued more by financially distressed firms, the Altman Z-score is expected to be negatively correlated with the observed CAR. That is, as the firm's financial health decreases, the CAR from divestiture should increase. Although in the correct direction, in both the model inclusive of the strategic variable and the model excluding the strategic variable, the Altman Z-score variable is not significant.

As this is contrary to the majority of the literature which finds a significantly positive relation between financial distress in the pre-divestiture period and divestiture returns (Afshar, Taffler & Sudarsanam 1992; Denning 1988; Lang, Poulsen & Stulz 1995) alternative measures of financial distress are employed. The measures used by Montgomery, Thomas and Kamath (1984), Afshar, Taffler and Sudarsanam (1992) and Lang, Poulsen and Stulz (1995) are adopted in substitution of the Altman Z-score. These alternative variables are the current ratio (CL/CA) and a scaled measure of earnings (EBIT/TA), the objective being that in aggregate, these variables will determine the firm's financial performance. Models 4 and 5 incorporate these variables. Only the strategic variable in Model 4 is statistically significant.

The final result of interest is the failure of the control variable, the log of the size relative, to be significant. This implies that unlike the US market (Klein 1986) and the European markets (Kaiser & Stouratis 1995) abnormal returns in the Australian market are not positively related to the relative size of the sell-off.

4.3 Disclosed Use of Proceeds

The final hypothesis provides an alternative test of the strategic/financial dichotomy based on the intended use of proceeds. The intended use is categorised as investment (strategic) purposes, debt reduction purposes, or undisclosed.

The cumulative abnormal returns (CARs) over announcement days 0 and +1 for each of the samples, presented in table 5, reflect the predictions of the theoretical framework. Abnormal returns for the 2 day announcement period where the intended use of proceeds is for strategic purposes (3.0%, p < 0.05) are larger than those where the intended use is for repayment of debt (1.8%, p<0.10). Further, the abnormal returns accruing to firms which fail to disclose the intended use of proceeds are not significant, consistent with expectations and fears of opportunistic behaviour (Jensen 1983).

4.4 Interaction Between Type of Divestiture and Disclosed Use of Proceeds

To examine the interaction between the type of sell-off and the disclosed use of the proceeds, firms are first classified as strategic or non-strategic on the basis of the relatedness of the divested unit, and each of these groups is then classified on the basis of the disclosed use of the proceeds of the sell-off. The announcement period returns are presented in table 6. This table shows that the largest announcement period returns (4.5%, p = 0.017) accrue to strategic sell-offs in which the disclosed use of proceeds is for strategic purposes.

In summary, all results strongly support the importance of the strategic rationale for sell-off gains. We find some, but much weaker support for the explanatory power of the financial rationale.

5. Conclusion

The results support the first hypothesis that, on average, Australian divestitures defined as sell-offs produce abnormal returns of 1.15% over the two-day announcement period. The magnitude of this result is comparable to that observed in the US sell-off literature. Further examination reveals that the gains are limited to divestitures of unrelated units, thus supporting the strategic theory of divestiture and recognising the value of increased firm focus.

Strong support is also found for the strategic theories of divestiture gains in both the multiple regression analysis and the disclosure test. Of the models tested, only regressions which include the industry-code dummy variable record significant F-statistics; further, the only explanatory variable of any significance is the industrycode variable. From the disclosure test, the strategic-use-of-proceeds sub-sample records the largest abnormal return, 3.0% (p = 0.017). When divestitures of unrelated units are also associated with a strategic use of the proceeds, an abnormal return of 4.5% (p = 0.017) is recorded.

Overall, the results suggest that the Australian market values strategic (unrelated) divestitures more so than financially motivated divestitures, although scope exists to further explore these results in the Australian market. Extension of this research could include investigation of firms divesting themselves of majority or controlling shareholdings, and firms making spin-offs as opposed to sell-offs.

(Date of receipt of final transcript: August, 2003.

Accepted by Garry Twite, Special Issue Editor.)

Table 1
US Evidence on Divestiture Gains (Sell-Offs)

                                                Period     Abnormal
Study                     Methodology    n      Studied     Return

Jain (1985)                   MAR       1064   1976-1978     0.6%
Alexander, Benson and         MAR         53   1964-1973     0.17%
  Kampmeyer (1984)
Roy and Manley (1997)         MM         174   1981-1988     1.2%
John and Ofek (1995)          MM         321   1986-1988     1.1%
Hite, Owers and Rogers        MM          55   1963-1978     1.66%
  (1987)
Klein (1986)                  MM         202   1970-1979     0.72%
Hearth and Zaima (1986)       MAR         75   1979-1982     1.42%

Note: MM is the market model and MAR is the mean adjusted returns
model;

For ease of comparison all abnormal returns are shown here only
for event period day -1 and day 0; and
n is the number of firms in the sample.

Table 2
Abnormal Returns for Event days -5 to +5

Based on a sample of 267 sell-off announcements occurring between
1 July 1994 and 30 June 2000, the following table includes abnormal
returns only for days on which there were consecutive trades. For
each of event days -5 to +5 the abnormal return, Student's t
statistic, p-value, and Shapiro Wilk W-statistic for normality are
listed. The Shapiro Wilk W-statistic measures the normality of
residuals on a scale of 0 to 1. Smaller values lead to the
rejection of the null hypothesis that the data is a random sample
from a normal distribution.

Event day    n    Abnormal   t stat      Shapiro Wilk
                   return                W-statistic

   -5       218    0.0008     0.349         0.8532
   -4       215    0.0002     0.104         0.8230
   -3       218   -0.0021    -1.016         0.8067
   -2       216   -0.0002    -0.102         0.9109
   -1       220    0.0031     1.463         0.8833
    0       223    0.0039     1.538         0.8305
    1       223    0.0076     2.445 **      0.7595
    2       217   -0.0005    -0.188         0.7662
    3       218   -0.0036    -1.506         0.7858
    4       220    0.0029     1.277         0.7436
    5       217   -0.0022    -1.253         0.9146

Note: ** p < 0.05.

Table 3
Cumulative Abnormal Returns for Strategic and Non-Strategic Sell-Offs

Strategic sell-offs are those for which the primary industry code of
the divesting firm differs from the primary industry code of the
divested unit and non-strategic sell-offs are those divestitures for
which the divesting firm and the divested unit have the same primary
industry code. CAR is the average cumulative abnormal return for the
specified events over event days 0 and +1. Student t statistics and
probability levels are shown.

                               N      CAR    t stat
                                            (p-value)

Strategic Divestitures       131   0.0171     3.38
                                             (0.001)

Non-strategic Divestitures    92   0.0026     0.41
                                             (0.680)

Table 4
Multiple Regression of Strategic Variables and Financial Distress
Variables Against CAR

Regression results are for sell-offs occurring between 1 July 1994 and
1 July 2000. The variables in the models are: [D.sub.IND] is a dummy
variable equal to 1 when the primary industry code of the divesting
firm differs from the primary industry code of the divested unit and 0
when the primary industry code of the divesting firm and the divested
unit are the same. ALTMANZ is the Altman Z score for the divesting
company based on the financial statement prior to the divestiture.
SIZE is the log of the size relative, divestiture value divided by
market value of divesting firm in the month prior to the divestiture.
CL/CA is the current ratio and equals current liabilities divided by
current assets. EBIT/TA is earnings before interest and tax divided by
total assets. F-statistics (and their probabilities), adjusted
R-squared values and Durbin-Watson values are presented for each
model. The different numbers in the regressions are determined by data
availability. For example, only 144 announcements provided the actual
dollar size of the divestiture.

   Model           1             2            3

[D.sub.IND]       0.224 ***     0.2 **
(t-stat)         (2.639)       (2.446)
(p-value)        (0.009)       (0.016)
ALTMANZ          -0.029                     -0.01
                (-0.347)                   (-0.119)
                 (0.729)                    (0.906)
SIZE              0.115         0.125        0.140
                 (1.366)       (1.525)      (1.625)
                 (0.174)       (0.130)      (0.107)
CL/CA
EBIT/TA
F STAT            3.250 **      4.425 **     1.332
(p-value)        (0.024)       (0.014)      (0.267)
N               136           144          136
Adj [R.sup.2]     0.048         0.046        0.005
DW                2.126         2.131        2.152

   Model           4             5

[D.sub.IND]       0.226 ***
(t-stat)         (2.584)
(p-value)        (0.010)
ALTMANZ
SIZE              0.127         0.150
                 (1.485)       (1.725)
                 (0.140)       (0.087)
CL/CA            -0.040        -0.080
                (-0.463)      (-0.912)
                 (0.644)       (0.363)
EBIT/TA          -0.033        -0.015
                (-0.386)      (-0.169)
                 (0.700)       (0.866)
F STAT            2.682 **      1.293
(p-value)        (0.035)       (0.280)
N               132           132
Adj [R.sup.2]     0.049         0.007
DW                2.108         2.104

Note: *** P<0.01; and ** p < 0.05.

Table 5
Abnormal Returns for Sub-Samples According to the Stated Use of
Proceeds from a Sell-Off

The table distinguishes between sell-off announcements in which the
proceeds are stated to be used for other strategic investments
(STRATEGIC); proceeds to be used to repay debt (DEBT); and
announcements in which the use of the proceeds is not disclosed
(UNDISCLOSED). For each sub-sample the mean CAR for event days 0 and
+1, the sample size, t-test and p-value are presented.

              Sample Size    CAR      t stat
                            (Mean)   (p-value)

Strategic          41       0.030     2.490
                                     (0.017)
Debt               41       0.018     1.881
                                     (0.067)
Undisclosed       141       0.004     0.842
                                     (0.4011)

Table 6
Interaction Between Type of Sell-Off and Stated Use of Proceeds

This table presents the interaction between the type of sell-off
(strategic or non-strategic) and the stated use of proceeds (strategy,
debt reduction, or undisclosed). Strategic sell-offs are those for
which the primary industry code of the divesting firm differs from the
primary industry code of the divested unit and non-strategic sell-offs
are those for which the divesting firm and the divested unit have the
same primary industry code. CAR is the average cumulative abnormal
return over event days 0 to +1. Student t statistics and sample sizes
are shown.

                            Type of Sell-Off

Use of Proceeds   Strategic   Non-strategic   Sample size
                     CAR           CAR
                   t stat        t stat
                  (p-value)     (p-value)

Strategic           0.045         0.012           41
                    2.59         (0.729)
                   (0.017)       (0.476)
                   n = 23        n = 18
Debt               0.0193        0.0163           41
                    1.57          1.099
                   (0.132)       (0.285)
                   n = 21        n = 20
Undisclosed        0.0094        -0.0055          141
                    1.80         -0.715
                   (0.075)       (0.478)
                   n = 87        n = 54
Sample size       131            92               223

The authors wish to thank Peter Clarkson, Scott Holmes and a referee for their comments on previous drafts of this paper.

(1.) As summarised in table 1.

(2.) A divestiture can also involve a spin-off of part of a firm's activities and transfer of assets to a new firm. As spin-offs are very few in comparison to sell-offs in Australia, the study is limited to a study of sell-offs.

(3.) Listing Rule 3.1; Chapter 15 Listing Rules; Corporations Law

(4.) A third possible group of explanations is founded on claims that improvements to the agency relationships, namely by the decrease of wealth transfers and reduced information asymmetry, are the source of divestiture gains. We consider agency theories the least satisfactory explanation of divestiture activity and focus our study on the strategic theory and the financial distress theory.

(5.) As an aside, Pashley and Philippatos (1990, 1993) study divestiture activity by segregating the sample according to the life-cycle phase of the firm. Firms in phases which involve high debt levels and / or poor performance (i.e. in financial distress) record higher instances of divestiture activity.

(6.) For example, when Colonial Life Limited announced on 2 May, 2000 that it would dispose of its UK Life and Pensions Business, Colonial was in the process of being taken over by Commonwealth Bank of Australia.

(7.) The arbitrary cut-off for bankruptcy concerns is zero (Altman 1968).

(8.) Data collected included current assets, current liabilities, total assets, book value of debt, retained earnings, EBIT, operating profit before tax, operating profit after tax and interest expense.

(9.) Where there was insufficient detail as to the operations of the divested asset it was assumed that the divested unit has the same industry classification as the divesting firm (Brown, James & Mooradian 1994).

(10.) The strategy statement may be indirect, for example, Goodman Fielder Limited (13/09/1999): 'This sale will allow us to focus our energies on further improvements to our remaining operations.' Alternatively, the strategy statement may be direct, for example, Cultus Petroleum NL: (24/12/1997): 'Proceeds from the proposed asset sales will be used to fund the continued exploration and appraisal of the company's core petroleum interests in Australia and NZ and for new acquisitions as appropriate opportunities arise.'

(11.) The statement of intention to use funds to reduce debt proceeds is generally unambiguous, for example, Mcpherson's Limited (03/01/1997): 'Consideration for the sale of shares is in the order of AUD$2.45 million. It is estimated that an abnormal profit on the sale (net of tax) of approximately AUD$1.20 million will be recorded. The proceeds will be used to reduce bank debt.' The majority of other firms in this sub-sample had similar unequivocal statements.

(12.) An example of where multiple uses were indicated is the announcement by Orica Limited (27/08/1998): 'Mr Weickhardt said that the proceeds from this divestment will be initially used to retire debt but will provide funds for reinvestment in higher growth businesses and contribute to the ongoing growth of the Orica group.' As prescribed, the first stated use, the reduction of debt, was the basis for classification.

(13.) The All Ordinaries Accumulation Index is a value weighted index which uses last traded prices where dividends are reinvested in the index on the ex-dividend date.

(14.) The original formula for Altman's Z-score is:

ALTMANZ - 0.012X1 + 0.014X2 + 0.033X3 + 0.006X4 + 0.0999X5

where: X1 = Working capital/Total Assets;

X2 = Retained Earnings/Total Assets;

X3 = Earnings before interest and taxes/Total assets;

X4 = Market value equity/Book value of total debt;

X5 = Sales/Total assets; and

Working capital = Current assets - current liabilities;

EBIT = Earnings before tax + interest expense.

(15.) The use of t-statistics is validated by the Shapiro-Wilk statistics, all above 0.7, which imply that the assumption of normality cannot be rejected.

References

Afshar, K.A., Taffler, R.J. & Sudarsanam, P.S. 1992, 'The effect of corporate divestments on shareholder wealth: The UK experience', Journal of Banking and Finance, vol. 16, pp. 115-35.

Alexander, G.J., Benson, P.G. & Kampmeyer, J.M. 1984, 'Investigating the valuation effects of announcements of voluntary divestitures', The Journal of Finance, vol. 39, no. 2, pp. 503-17.

Altman, E.I. 1968, 'Financial ratios, discriminant analysis and the prediction of corporate bankruptcy', The Journal of Finance, vol. 23, no. 4, pp. 589-609.

Berger, P.G. & Ofek, E. 1995, 'Diversification's effect on firm value', Journal of Financial Economics', vol. 37, pp. 39-65.

Bhagat, S., Schleifer, A. & Vishny, R. 1990, Hostile Takeovers" in the 1980s: The Return to Corporate Specialisation, Brookings Institute, Washington.

Brown, D., James, C. & Mooradian, R. 1994, 'Asset sales by financially distressed firms', Journal of Corporate Finance, vol. 1, no. 2, pp. 233-57.

Brown, S. & Warner, J. 1985, 'Using daily stock returns: The case of event studies', Journal of Financial Economics, vol. 14, pp. 3-31.

Capon, N., Christodolou, C. Farley, J.U. & Hulbert, J. 1987, 'A comparative analysis of the strategy and structure of United States and Australian corporations: An evolutionary perspective', Journal of International Business Studies, vol. 18, Spring, pp. 51-74.

Cho, M.H. & Cohen, M.A. 1997, The economic causes and consequences of corporate divestitures', Managerial and Decision Economics, vol. 18, pp. 367-74.

Comment, R. & Jarrell, G. 1995, 'Corporate focus and stock returns', Journal of Financial Economics, vol. 37, pp. 67-87.

Constantinides, G.M. 1984, 'Optimal stock trading with personal taxes', Journal of Financial Economics, vol. 13, pp. 65-89.

Daley, L., Mehrotra, V. & Sivakumar, R. 1997, 'Corporate focus and value creation: Evidence from spin-offs', Journal of Financial Economics, vol. 45, pp. 257-81.

Denning, K.C. 1988, 'Spin-offs and sales of assets: An examination of security returns and divestment motivations', Accounting and Business Research, vol. 19, no. 73, pp. 32-42.

Desai, H. & Jain, P. 1999, 'Firm performance and focus: Long-run stock market performance following spin-offs', Journal of Financial Economics, vol. 54, pp. 75-101.

Galai, D. & Masulis, R. 1976, 'The option pricing model and the risk factor of stock', The Journal of Financial Economics, vol. 3, pp. 53-81.

Hamilton, R.T. & Chow, Y.K. 1993, 'Why managers divest--Evidence from New Zealand's largest companies', Strategic Management Journal, vol. 14, pp. 479-84.

Hearth, D. and Zaima, J.K. 1986, 'Divestiture uncertainty and shareholder wealth: Evidence from the USA (1975-1982)', Journal of Business Finance and Accounting, vol. 13, no. 1, pp. 71-85.

Hite, G.L. & Owers, J.E. 1983, 'Security price reactions around corporate spin-off announcements', Journal of Financial Economics, vol. 12, December, pp. 409-36.

Hite, G.L., Owers, J.E. & Rogers, R. 1987, 'The market for interfirm sales: partial sell-offs and total liquidation', Journal of Financial Economics, vol. 18, pp. 229-52.

Jain, P.C. 1985, 'The effect of voluntary sell-off announcements on shareholder wealth', The Journal of Finance, vol. 40, no. 1, pp. 209-24.

Jensen, M. & Ruback, R.S. 1983, 'The market for corporate control: The scientific evidence', Journal of Financial Economics, vol. 11, pp. 5-50.

Jensen, M.C. 1986, 'Agency costs of free cash flow, corporate finance, and takeovers', American Economic Review, vol. 76, March, pp. 3-27.

John, K. & Ofek, E. 1995, 'Asset sales and increase in focus', Journal of Financial Economics, vol. 37, pp. 105-26.

Kaiser, K. & Stouratis, A. 1995, 'Value creation through corporate restructuring: European divestitures', European Management Journal, vol. 13, no. 2, pp. 164-73.

Kaplan, S.N. & Weisbach, M.S. 1992, 'The success of acquisitions: Evidence from divestitures', The Journal of Finance, vol. 47, no. 1, pp. 107-38.

Klein, A. 1986, 'The timing and substance of divestiture announcements: Individual, simultaneous and cumulative effects', The Journal of Finance, vol. 41, no. 3, pp. 685-96.

Krishnaswami, S. & Subramaniam, V. 1999, 'Information asymmetry, valuation and the corporate spin-off decision', Journal of Financial Economics, vol. 53, pp. 73-112.

Lang, E.H.P., Poulsen, A. & Stulz, R.M. 1995, 'Asset sales, firm performance, and the agency costs of managerial discretion', Journal of Financial Economics, vol. 37, pp. 3-37.

Levy, H. & Sarnat, M. 1970, 'Diversification, portfolio analysis and the uneasy case for conglomerate mergers', The Journal of Finance, vol. 25, no. 4, pp. 793-802.

MacKinlay, A.C. 1997, 'Event Studies in economics and finance', Journal of Economic Literature, vol. 35, pp. 13-39.

Markides, C.C. & Berg, N.A. 1992, 'Good and bad divestment: The stock market verdict', Long Range Planning, vol. 25, no. 2, pp. 10-5.

Mauer, D.C. & Lewellen, W.G. 1990, 'Security holder taxes and corporate restructurings', Journal of Financial Quantitative Analysis, vol. 25, no. 3, pp. 341-60.

Miles, J.A. & Rosenfeld, J.D. 1983, 'The effect of voluntary spin-off announcements of shareholder wealth', Journal of Finance, vol. 38, no. 5, pp. 1597-606.

Montgomery, C.A. & Thomas, A.R. 1988, 'Divestment: motives and gains', Strategic Management Journal, vol. 9, pp. 93-7.

Montgomery, C.A., Thomas, A.R. & Kamath, R. 1984, 'Divestiture, market valuation and strategy', Academy of Management Journal, vol. 27, no. 4, pp. 830-40.

Ofek, E. 1993, 'Capital structure and firm response to poor performance', Journal of Financial Economics, vol. 34, no. 1, pp. 3-30.

Pashley, M.M. & Philippatos, G.C. 1990,' Voluntary divestitures and corporate life cycle: Some empirical evidence, Applied Economics, vol. 22, pp. 1181-96.

Pashley, M.M. & Philippatos, G.C. 1993, 'Have voluntary divestiture of US corporations increased shareholder wealth? Empirical evidence from the life cycle', Applied Financial Economics, vol. 3, pp. 39-49.

Roy, L.D. & Manley, J.F. 1997, 'The effect of finn specialisation on shareholder wealth', The Mid-Atlantic Journal of Business, vol. 33, no. 3, pp. 219-36.

Schipper, K. & Smith, A. 1983, 'Effects of recontracting on shareholder wealth: The case of voluntary spin-offs', Journal of Financial Economics', vol. 12, December, pp. 437-67.

Sicherman, N.W. & Pettway, R.H. 1987, 'Acquisition of divested assets and shareholders' wealth', The Journal of Finance, vol. 42, no. 5, pp. 1261-73.

Slovin, M., Sushka, M. & Ferraro, S. 1995, 'A comparison of the information conveyed by equity carve-outs, spin-offs and asset sell-offs', Journal of Financial Economics, vol. 37, no. 37, pp. 89-104.

Williamson, O.E. 1988, 'Corporate finance and corporate governance', The Journal of Finance, vol. 43, July, pp. 567-91.

Mary Rose Cooney ([dagger])

Frank Finn ([dagger])

Angela Karl ([dagger])

([dagger]) UQ Business School, The University of Queensland, St Lucia, Queensland, Australia, 4072. Email: [e-mail]