The Palm/3Com Carve-Out Mispricing
In 2000, 3Com sold a slice of its Palm subsidiary in an IPO while retaining most of Palm itself, yet Palm's IPO-day market value implied 3Com's remaining stake alone was worth more than all of 3Com — a textbook violation of the law of one price.
Prerequisites: Carve-Out vs Spin-Off vs Split-Off
In March 2000, 3Com carried out a partial IPO of its Palm subsidiary, selling about 5% of Palm's shares to the public while announcing it would distribute most of its remaining Palm shares to 3Com shareholders later that year — meaning each share of 3Com came bundled with roughly 1.5 shares of Palm. Simple arithmetic says 3Com's stock price should have been at least 1.5 times Palm's price, since 3Com's value included that whole Palm stake plus 3Com's own separate businesses. Instead, on Palm's first trading day, Palm's price implied a value for 3Com's Palm stake alone that exceeded 3Com's entire market capitalization — implying the market was assigning a negative value to all of 3Com's non-Palm businesses, which were profitable and had billions in cash.
The mispricing persisted for weeks, not seconds, despite being arithmetically obvious and despite a clear arbitrage — short Palm, buy 3Com — being available to any investor willing to take it on. The trade was not riskless in practice: short-selling Palm was constrained by limited share supply for borrowing, which kept the arbitrage from correcting the price quickly and made the episode a widely cited example of how short-sale constraints, not irrationality alone, can let a plainly wrong price persist.
The Palm/3Com carve-out showed a subsidiary's implied stub value exceeding its parent's full market cap — an arithmetic impossibility that persisted for weeks because short-sale constraints on the newly IPO'd, thinly floated stock prevented arbitrageurs from correcting it quickly.
Further reading
- Lamont & Thaler, Can the Market Add and Subtract? Mispricing in Tech Stock Carve-outs (2003, JPE)