Carve-Outs and Divestitures

Carve-outs and divestitures are corporate restructuring decisions in which a firm separates a division, subsidiary, or asset from the rest of the company. Because the announcement of such a transaction is a discrete, dated, and largely unanticipated piece of public information, the event study method (ESM) is the natural tool for measuring how capital markets value the decision: it isolates the abnormal stock return attributable to the announcement from normal market-wide movements, and so quantifies whether investors judge the separation to create or destroy value. 

A central point that an authoritative analysis must respect is that "divestiture" is not one event type but several, each with distinct economics, information content, and typical abnormal-return magnitudes. The three canonical modes should be sampled and studied separately:

• Spin-offs: the parent distributes shares of a subsidiary pro rata to its existing shareholders. No cash changes hands; the parent simply splits itself into two independently traded firms.
• Equity carve-outs: the parent sells a minority stake (a partial IPO) of a subsidiary to the public and raises cash, while typically retaining control.
• Sell-offs and asset sales: the parent privately sells a division, plant, or business unit to another firm for cash. This is a two-sided event, with abnormal returns to both the seller and the buyer.

Two further restructuring modes appear in the literature alongside these three and are described in the related modes note below: split-offs (shareholders exchange parent shares for subsidiary shares, a tax-free swap) and tracking stock (a separate class of parent equity that tracks a division's results, with no new legal entity). Across all the cash-or-share separations, the robust empirical finding is a significantly positive average announcement return to the divesting (parent or seller) firm. The magnitudes, drivers, and methodological subtleties differ by mode, and are summarized below.

The practitioner's mental model behind these results is the diversification (conglomerate) discount: diversified North American and Western European firms have historically traded at a meaningful discount (often estimated near 10%) to portfolios of comparable pure-play peers. A divestiture is, in this framing, a way to unlock that discount by letting each business be valued on its own merits. The academic "focus hypothesis" discussed below is the formal counterpart of this intuition.

What the research shows

The table below summarizes the headline pattern by mode before the detailed evidence that follows.

Mode	Typical announcement abnormal return (divesting firm)	Primary window	Key value driver	Anchor citation
Spin-off	about +3% (consensus +3.02%)	[-1,+1]	Focus-increasing, cross-industry separation	Veld and Veld-Merkoulova (2009)
Equity carve-out	about +1.8% (positive, unlike a seasoned equity offering)	[-1,0] or [-1,+1]	Value-revelation and use of proceeds	Schipper and Smith (1986)
Sell-off / asset sale (seller)	about +1% to +1.7%	[-1,0]	Focus-increasing sales; de-levering use of proceeds	John and Ofek (1995)
Tracking stock	about +2% to +3%, but preceded by negative one-year returns	[-1,+1]	Weaker and less durable than spin-offs/carve-outs	see related-modes note

Spin-offs: roughly +3% at announcement

The founding spin-off event studies (Miles and Rosenfeld, 1983; Hite and Owers, 1983; Schipper and Smith, 1983) all document significantly positive parent abnormal returns concentrated in a two- to three-day window around the announcement, on the order of +2% to +4%. Hite and Owers (1983), studying 123 announcements, report parent cumulative abnormal returns (CARs) near +3.3% and find that the gains are not explained by wealth transfer from bondholders. Schipper and Smith (1983) attribute the roughly +2.8% reaction to relaxed regulatory and contractual constraints and improved managerial incentives (a "recontracting" and transparency story), rather than tax avoidance alone. The meta-analysis by Veld and Veld-Merkoulova (2009), covering 26 spin-off event studies, pins the consensus announcement abnormal return at +3.02%, and is the single best citation for the headline number.

Returns scale with the relative size of the divested unit and are higher for tax- and regulation-friendly transactions (Miles and Rosenfeld, 1983; Veld and Veld-Merkoulova, 2009). A second-order finding is a partial wealth transfer from bondholders: Maxwell and Rao (2003) document a significant abnormal bond return of about -88 basis points around spin-off announcements (driven by collateral loss and rating downgrades), but show that combined debt-plus-equity value still rises, so expropriation is not the whole story. Early claims of multi-year post-spin-off outperformance survive less well: Cusatis, Miles and Woolridge (1993) find positive long-run drift concentrated in spin-offs that subsequently become takeover targets, but Veld and Veld-Merkoulova (2009) caution that much of the long-run effect does not hold up under rigorous calendar-time and buy-and-hold tests. A page should not overclaim durable long-run gains.

A modern strand sharpens when and why the value appears, complementing the information channel discussed below. Feldman (2016) shows that spin-offs improve the firm's information environment by inducing equity analysts to revisit their coverage decisions: the announcement gains are larger for spin-offs that remove a business that had been constraining or muddying analyst coverage of the remaining firm. This treats the analyst-coverage response as a mechanism, not just a correlate, and explains why two otherwise similar spin-offs can re-rate very differently.

Equity carve-outs: positive, and opposite in sign to a seasoned equity offering

The canonical carve-out study is Schipper and Smith (1986), which documents a significantly positive parent announcement reaction of roughly +1.8% on a short window. (More recent and longer-window or non-US samples report figures up to about +2.8%; these are larger-window estimates, not a contradiction of the +1.8% short-window result.) The striking contrast is that ordinary seasoned equity offerings by a parent typically draw a negative reaction of around -3% (an adverse-selection signal), whereas a carve-out, which also raises equity, draws a positive one. The market reads a carve-out as revealing the value of an undervalued subsidiary and as a refocusing step, not as a signal that the parent is overvalued. Vijh (2002) reports parent CARs of about +1.2% to +2.3% and, critically, finds that rival firms react negatively, which rejects a pure asymmetric-information explanation and supports the divestiture-gains (refocusing and financing) hypothesis.

The use of proceeds matters sharply for cash divestitures. Allen and McConnell (1998) show that the parent return is about +6.63% when carve-out proceeds are used to repay debt, versus roughly zero (-0.01%) when the cash is retained for investment; firms tend to carve out when capital-constrained and after poor operating performance. The same use-of-proceeds logic appears in sell-offs (Lang, Poulsen and Stulz, 1995): the market rewards distributing or de-levering with proceeds more than reinvesting them, consistent with the agency costs of managerial discretion. Long-run carve-out returns, by contrast, tend to be muted or negative for both parent and unit, paralleling IPO underperformance (Vijh, 2002). European evidence makes the long-run caveat concrete: studies of European carve-outs report a statistically positive announcement-day reaction (on the order of +1.7%) that is stronger where minority-shareholder protection is higher, yet reverses into significant negative returns over the following six to twenty-four months even as the units' operating profitability improves.

Sell-offs and asset sales: positive for the seller, modestly positive for the buyer

Voluntary sell-off announcements raise wealth on both sides of the transaction. Jain (1985), studying a large sample, finds significant positive announcement returns for both sellers and buyers, with seller gains exceeding buyer gains, and significant negative seller returns in the run-up, consistent with divestiture as a corrective response to poor prior performance. John and Ofek (1995) report a two-day seller CAR of about +1% to +1.7% and show that both the announcement gain and the subsequent improvement in the seller's operating performance are concentrated in focus-increasing sales. Hite, Owers and Rogers (1987) document significant positive CARs to sellers in partial sell-offs and total liquidations, and gains to buyers, supporting asset redeployment to higher-valued users; Sicherman and Pettway (1987) confirm modest positive buyer returns.

Why value is created: the focus hypothesis and its rivals

Four hypotheses compete to explain these gains, and the evidence discriminates among them:

• Corporate focus / refocusing (best supported): value is concentrated in focus-increasing, cross-industry transactions. Daley, Mehrotra and Sivakumar (1997) show that significant announcement returns and operating improvements around spin-offs occur only for cross-industry (different two-digit SIC) deals; own-industry spin-offs show little. John and Ofek (1995) find the same pattern for asset sales. This is the cleanest cross-sectional driver and the right explanation for why divestitures create value: separation removes the diversification (conglomerate) discount, sharpens incentives, and creates pure-play securities for specialized investors. In a cross-sectional regression the driver is typically operationalized as a focus-increasing dummy (parent and unit in different two-digit SIC codes) or as the change in a revenue-based Herfindahl index, interacted with the relative size of the divested unit.
• Information / value-revelation: divesting firms tend to have elevated pre-event information asymmetry that falls afterward and predicts the announcement return (Krishnaswami and Subramaniam, 1999). The channel has support for spin-offs, but the negative rival reaction to carve-outs rejects a pure asymmetric-information story (Vijh, 2002). Feldman (2016) gives this channel a concrete mechanism by tracing the value to analysts' coverage responses.
• Financing / agency of managerial discretion: cash divestitures motivated by financing need create more value when proceeds are distributed or used to repay debt than when retained (Allen and McConnell, 1998; Lang, Poulsen and Stulz, 1995).
• Wealth transfer to bondholders: real but partial, and not the source of the equity gain (Maxwell and Rao, 2003).

Related modes: split-offs and tracking stock

Two restructuring forms sit just outside the canonical three and are worth distinguishing in any divestiture sample. A split-off offers shareholders the choice to exchange some of their parent shares for shares in the subsidiary; because it is structured as a share exchange rather than a pro-rata distribution, it can be tax-free and it changes the parent's share count, which affects how the announcement effect should be measured. A tracking stock creates a separate class of parent equity whose performance is tied to a specific division, with no new legal entity and no transfer of assets or control. The empirical record treats tracking stock as the weakest of these structures: announcements tend to show positive CARs in the +2% to +3% range, but they are typically preceded by negative one-year excess returns, and the structures have proved less durable than spin-offs and carve-outs (institutional-trading and information-production evidence on the choice among spin-offs, carve-outs, and tracking-stock issues). Pool these with the canonical modes only with care; better to flag them with their own indicator.

Beyond isolated events: divestiture programs

While the vast majority of prior papers analyze divestitures as isolated events, a more recent strand examines divestiture programs and the experience and timing effects that single-event studies miss. Brauer and Schimmer (2010) studied the abnormal returns of 160 divestitures in the international insurance field over 1999 to 2007, and, instead of statistical clustering, used the ad hoc corporate press releases accompanying each announcement to categorize the divestitures. This program perspective is a useful complement to, not a substitute for, the canonical single-event evidence above.

Recent real-world examples

The pattern is not a 1980s artifact: large-cap conglomerates continue to use these structures, and recent deals illustrate both the value logic and the methodological pitfalls.

• GE three-way breakup (spin-off): announced 9 November 2021, with the shares rising more than 2% on the announcement even as the broader market pulled back. The spins completed in stages, with GE HealthCare separating in January 2023 and GE Vernova beginning to trade on the NYSE on 2 April 2024. This is a textbook demonstration that the wealth effect lands at announcement and that the announcement-to-completion gap can exceed two years, so day 0 must be the press release, not the ex-distribution date.
• Kellogg into Kellanova and WK Kellogg Co (spin-off): announced on 21 June 2022 and completed on 2 October 2023, framed as a refocusing move so the higher-growth snacks business could be valued closer to its pure-play peers. The announcement was bundled with broader strategic-portfolio commentary, a reminder to screen the event window for confounding news.
• 3M and Solventum (spin-off): Solventum began trading on the NYSE on 1 April 2024, separating 3M's healthcare exposure into a standalone so each entity is valued on its own merits. A clean, recent conglomerate-discount case a researcher can run directly.
• Novartis and Sandoz (spin-off, October 2023): Sandoz began trading on the SIX Swiss Exchange on 4 October 2023, explicitly framed by management as completing Novartis's transformation into a pure-play innovative-medicine business. A clean European, healthcare, focus-increasing case that pairs well with the European-carve-out caveat that markets can be slow to re-rate even after a focus-increasing separation.

How to run this kind of event study

For the conceptual foundations see our introduction to event study methodology; for a step-by-step workflow see the event study application blueprint. The points below are the specifics that matter for divestitures.

• Segment by transaction type first. Spin-offs, equity carve-outs, and sell-offs (and, if present, split-offs and tracking stock) have different information dynamics and different return magnitudes; pooling them blurs the result. Build a separate sample for each.
• Identify the event date carefully. The primary event is the first credible public announcement of the divestiture (press release, news wire, or financial press), not the completion or ex-distribution date. Spin-offs in particular follow a multi-stage timeline (announcement, then board and regulatory approval, then record date, then ex-date), and the wealth effect is concentrated at announcement; Veld and Veld-Merkoulova (2009) note that completed spin-offs earn less than the announcement reaction implies. Klein (1986) shows that the reaction concentrates when the transaction price or terms are disclosed, so consider an intent-to-sell date, a price-disclosure date, and a completion date, and test robustness to the choice.
• Source the date from auditable filings. For US deals the SEC Form 8-K (current report) timestamp and the Form 10 (spin-off registration) provide auditable announcement dates, and press wires fill in the rest. These are exactly the documents the Event Date Identifier can parse at scale.
• Use short, symmetric windows. The canonical literature uses two-day [-1,0] or three-day [-1,+1] windows to limit confounding. Report these as primary and longer windows ([-5,+5], [-10,+10]) as robustness. Inspect, rather than ignore, the pre-event window: the seller's negative run-up before a sell-off is itself informative.
• Screen for confounding events. Divestiture announcements are frequently bundled with earnings releases, dividend or buyback changes, management changes, simultaneous restructuring, or the announced use of proceeds, each of which has its own price effect (Klein, 1986). Screen the event window and either exclude contaminated observations or control for them.
• Treat sell-offs as two-sided. Collect and analyze both the seller and the publicly traded buyer as separate samples, since both can earn positive returns and the total gain is split (Jain, 1985). For carve-outs, keep the parent's announcement reaction separate from the subsidiary IPO's own first-day return; the parent effect is the divestiture-relevant statistic.
• Distinguish voluntary from distressed deals. Forced or creditor-pressured asset sales can show muted or negative seller returns (Lasfer, Sudarsanam and Taffler, 1996); keep them in a separate bucket.
• Build the right cross-sectional design. The variables the literature consistently finds significant are: a focus-increasing dummy (parent and unit in different two-digit SIC codes, or a change in revenue-based Herfindahl), the relative size of the divested unit, tax status (tax-free versus taxable), parent leverage or financial distress, and the stated use of proceeds. For international or non-US samples, add a country-level minority-shareholder-protection or governance measure, since the announcement reaction to carve-outs is stronger where minority protection is better.
• Watch for event clustering. Divestitures occur in industry waves, so samples often contain overlapping event windows and cross-sectionally correlated returns. Use calendar-time portfolios or clustering-robust test statistics when windows overlap, rather than treating observations as independent.
• Choose the benchmark and statistics to match the horizon. Use a market model or Fama-French factor model with a clean estimation window (commonly 200 to 250 trading days ending well before the event) for short-window CARs; for a newly spun-off pure-play unit that lacks pre-event history, a matched control-firm or industry benchmark is preferable. For any long-horizon extension, switch to buy-and-hold abnormal returns (BHAR) or calendar-time portfolio regressions to handle the bad-model and cross-correlation problems. To test wealth transfer or total firm value, supplement the equity study with a bond event study (Maxwell and Rao, 2003). For the menu of test statistics see our significance tests documentation and the expected return models overview; full sources are in the references.

Run it with our tools

Our calculators implement this workflow end to end and are free to use:

• Abnormal Return Calculator (ARC): the core tool. Build a request file of announcement dates (one row per divesting firm), pick the estimation and event windows ([-1,0] or [-1,+1] as primary), choose the market model or a factor model, and select the test statistics. Run separate batches for spin-offs, carve-outs, and sell-offs, then add a focus-increasing dummy and use-of-proceeds variable for the cross-sectional regression.
• Abnormal Volume Calculator (AVC) and Volatility Calculator (AVyC): corroborate the return reaction with abnormal trading volume and volatility around the announcement, useful when confounding events make the return signal ambiguous.
• Event Date Identifier (EDI): given the multi-stage divestiture timeline, EDI parses large volumes of press releases and filings (such as SEC Form 8-K and Form 10) and returns the dates mentioned in the text, helping you pin down the correct day 0.
• News Analytics (CATA): classify and code divestiture press releases at scale (for example, to separate spin-offs from carve-outs from sell-offs, or to flag focus-increasing language) before running the return study.

A short worked example: to study the 3M / Solventum spin-off, set day 0 to the original announcement press release (not the 1 April 2024 first-trading date), choose a [-1,+1] event window with a 250-trading-day estimation window ending well before the announcement, fit the market model in ARC, and record the parent CAR. Extending to a sample, add each deal's focus-increasing dummy, relative size, and stated use of proceeds, then run the cross-sectional regression to test whether the focus-increasing, cross-industry deals carry the gains.

Beyond academia, these divestiture event studies are a staple of litigation and valuation work: securities-fraud, damages, and corporate-valuation disputes routinely use the price impact of a restructuring announcement as quantitative evidence, and corporate-development teams use the same analysis to gauge how the market is likely to receive a planned separation.

Related use cases

Divestitures sit alongside other corporate-event applications of the method. See mergers and acquisitions (the mirror-image transaction, where the same focus and payment logic applies), alliances and joint ventures, and competitive dynamics. For the full catalogue of applications, return to the overview of practical applications.