A spin-off and a carve-out are two ways a parent company separates a business unit into a distinct entity, but they differ on the axis that matters most to investors: who ends up holding the new shares and whether cash changes hands. In a spin-off, the parent distributes shares of a subsidiary (SpinCo) directly to its existing shareholders pro rata — no money is raised, no shareholder action is required, and if structured correctly it is tax-free. In a carve-out (equity carve-out), the parent sells a stake in the subsidiary to the public via an IPO, raising cash but creating a taxable event on the shares sold while typically retaining majority control. The core tension is the same: separating a unit can unlock value by removing conglomerate discount and giving managers focused incentives — but it also strips away shared resources and creates a stub of forced, mechanical trading that the academic record says is genuinely worth studying.

How they're formed

Spin-off. The parent declares a distribution: holders of the parent receive a set ratio of SpinCo shares per parent share (e.g. 1 SpinCo share for every 3 parent shares) as of a record date. SpinCo begins trading as an independent public company on the distribution date, often with a brief "when-issued" market beforehand. The parent's share price drops to reflect the value handed off. To qualify as tax-free under IRC Section 355, the transaction must clear several tests: both the parent (Distributing) and SpinCo (Controlled) must have been engaged in an active trade or business for the prior five years; the parent must distribute "control" (generally ≥80%); there must be a genuine corporate business purpose; and the deal cannot be a "device" for distributing earnings tax-free. A "cash-rich" entity made mostly of liquid non-business assets fails the test (Carpenter Wellington; Accounting Today).

Carve-out. The parent forms a subsidiary, registers it, and sells a minority slice (often 10–20%) in an IPO. Existing parent shareholders get nothing automatically; the parent gets cash and a public valuation for the unit. A carve-out is frequently a first step — used for price discovery — later completed by a full spin-off or split-off of the residual stake.

Related variants. A split-off offers parent holders the choice to exchange parent shares for SpinCo shares (like a buyback paid in subsidiary stock; tax-free if structured under §355). A split-up dissolves the parent entirely into two or more new companies.

How they're used in practice

Boards pursue these separations to: (1) dissolve a conglomerate discount where the market values a diversified parent below its sum-of-parts; (2) let each business pursue its own strategy, multiple, and investor base; (3) sharpen management incentives via pure-play equity; (4) raise cash (carve-out) without diluting the parent's core; or (5) defend against activists demanding a breakup. Carve-outs are also used to monetize a hidden gem while retaining upside on the stub.

For investors, the structurally interesting moment is the post-spin technical setup. SpinCos are routinely smaller than the indexes and benchmarks that held the parent, so index funds and institutions sell the distributed shares mechanically — not because they judge SpinCo overvalued but because it falls outside their mandate. This indiscriminate selling can create a temporary price dislocation, the central thesis Joel Greenblatt popularized in You Can Be a Stock Market Genius (1997). Practitioners watch for the spin to "wash out" forced sellers in the first weeks before assessing fundamentals.

Adoption, debate & evidence

Spin-offs are a well-established, regularly used corporate tool, and the "spin-offs beat the market" claim has real academic roots — but the evidence has decayed and is contested.

• The foundational study, Cusatis, Miles & Woolridge (1993, Journal of Financial Economics 33:293–311), found significant positive matched-firm-adjusted abnormal returns for spin-offs, their parents, and the combined entities over a 1965–88 sample (the paper reports roughly 25% over two years and 33.6% over three years, significant at 5%). This is the most-cited source for the outperformance folklore — but the authors themselves note the abnormal performance was largely concentrated in firms that became takeover targets, suggesting the gains may reflect merger activity rather than the spin-off act per se.
• McConnell, Ozbilgin & Wahal (2001, "Spin-offs, Ex Ante," Journal of Business 74(2):245–280) re-tested on a later sample and found only limited evidence of overperformance, with results sensitive to outliers — an early sign the effect was weakening.
• McConnell, Sibley & Xu (2015, Journal of Portfolio Management 42(1):143–152) examined 2001–2013 and found subsidiaries still substantially outperformed size/book-to-market benchmarks, while parents outperformed only modestly (an equal-weighted parent portfolio beat its benchmark by about 0.27%/month over the first 15 months, ~3.7% cumulative). Notably, the Guggenheim Spin-Off ETF (CSD) — the passive way to harvest the effect — underperformed an active strategy of buying all subsidiaries, though it still beat the value-weighted index. The takeaway: any edge appears concentrated in the subsidiary, is sensitive to weighting and implementation, and may be partly an illiquidity/small-cap/value premium rather than a free "spin-off" anomaly.

Honest summary: the abnormal-return claim is empirically supported but smaller and less reliable than the folklore suggests, period-dependent, and hard to capture net of frictions. It is not a guaranteed edge.

Strengths & limitations

When it works: the forced-selling dislocation is real and creates genuine pricing inefficiency in the weeks after distribution, especially for small, under-followed SpinCos that no analyst yet covers. Focused management and a clean balance sheet can re-rate a previously buried business.

When it fails: parents frequently use spin-offs to offload the worse business — saddling SpinCo with debt, pension liabilities, or a declining product (a "garbage barge"). SpinCos lose the parent's shared services, scale, and credit, and many have weak or temporary governance. The carve-out's retained-control structure creates conflicts of interest between parent and minority holders. The #1 misuse is treating "spin-off" as an automatic buy signal — the dispersion across outcomes is enormous, and the average masks both multibaggers and value traps. Each situation requires reading the separation documents (Form 10/prospectus) for who got the debt and why the parent really separated.

Sources

• Cusatis, Miles & Woolridge, "Restructuring through spinoffs: The stock market evidence," Journal of Financial Economics 33 (1993): 293–311 — ScienceDirect / Semantic Scholar.
• McConnell, Sibley & Xu, "The Stock Price Performance of Spin-Off Subsidiaries, Their Parents, and the Spin-Off ETF, 2001–2013," Journal of Portfolio Management 42(1) (2015): 143–152 — Purdue faculty PDF.
• McConnell, Ozbilgin & Wahal, "Spin-offs, Ex Ante," Journal of Business 74(2) (2001): 245–280 — JSTOR (limited post-1988 outperformance, sensitive to outliers).
• Carpenter Wellington PLLC / Lexology, "Spin-Off Transactions: Tax Issues and Section 355."
• Accounting Today, "The basics of a tax-free spin-off transaction in a corporate restructuring."
• ACap Advisors, "Spin-Off vs. Split-Off vs. Carve-Out" (definitions); AnalystPrep CFA Level II, "Equity Carve-Outs, Spin-Offs, Split-Offs, and Liquidation."
• Joel Greenblatt, You Can Be a Stock Market Genius (1997) — origin of the forced-selling thesis (practitioner source, not peer-reviewed).

Disputes flagged: The magnitude and persistence of spin-off outperformance is genuinely contested (Cusatis 1993 strong vs. McConnell 2001 limited vs. 2015 subsidiary-concentrated). Even the strong 1993 result was concentrated in firms that subsequently became takeover targets, raising the question of whether the "anomaly" is really a takeover-likelihood effect. Greenblatt's forced-selling thesis is widely cited but practitioner-sourced, not academically established as the cause of any abnormal return.