Spin-Off Performance Tracker

Methodology

The spin-off effect is one of the more durable findings in corporate finance. After a company separates a division into its own public stock, both the new spinco and the remaining parent have tended to outperform the market for a stretch. One common explanation is forced selling by index funds that hold the parent but have no mandate to hold an unfamiliar new small-cap, which can depress the spinco's price for reasons unrelated to its fundamentals.

The event list (which company spun off which division, and when) is hand-compiled from public record, as on Expectations vs. Reality. Every return number comes from Alpha Vantage TIME_SERIES_DAILY_ADJUSTED (adjusted close, so dividends are included). The anchor date for each event is the spinco's first day of regular trading. Both legs are measured against SPY over the same calendar window, so their excess returns are comparable.

19 events are tracked, going back to eBay's 2015 spinoff of PayPal. Two of them (United Technologies' 2020 spinoffs of Otis and Carrier) have no parent leg, because UTC merged into Raytheon the same day and didn't continue as an independent company. A few others carry a footnote. Viatris (Pfizer's Upjohn unit combined with Mylan through a Reverse Morris Trust) isn't a pure spinoff. Kenvue is anchored to the exchange-offer completion date, not the earlier IPO carve-out. For Kellanova/WK Kellogg, the parent ticker K kept trading under a new name. In the 2015 HP split, both HPQ and HPE are technically new companies, and HPQ, which kept the legacy ticker, is treated as the parent leg. A ticker whose price history doesn't reach its anchor date is dropped from that one leg.

The headline chart aligns every leg to trading day 0, its own separation date, then averages cumulative excess return across every leg at each elapsed trading day. This is the standard event-time setup for asking whether an effect shows up around an event. Sample size shrinks as the horizon lengthens, because younger events (GE Vernova, Solventum) haven't reached the 3-year mark. The chart shows N alongside the average.

Significance. At the 1-year mark, run once for spinco legs and once for parent legs, there are two checks. A one-sample t-test asks whether mean excess return across legs differs from zero. An exact two-sided sign test asks whether legs with a nonzero excess return split roughly 50/50 up and down. With roughly 15-19 events, this is far below the sample size of academic spin-off studies. Several events also share a corporate family (three Danaher spinoffs, two General Electric, two DowDuPont), so they aren't fully independent. Treat the result as illustrative, not a rigorous test.

Refreshes weekly. A fixed list of 19 historical events doesn't need daily updates. The only thing that changes week to week is one more week of price history in the "to-date" return column.