Edward Breen’s four-year tenure heading DuPont was one for the corporate history books. He reshaped the centuries-old industrial giant through a series of acquisitions, including a merger with Dow Chemical. Then he engineered the slicing of $86 billion DowDuPont into three separate public entities.

In June 2019, after the splits were finalized, Breen relinquished the CEO role, becoming chairman of DuPont de Nemours, the newly formed specialty chemical company. But shares in the new DuPont soon tanked, as weak demand for auto parts and U.S.-China trade tensions ate into earnings. The board’s solution? Bring back Breen. He’s CEO again, as of Feb. 18.

It’s a common tactic among companies in dire straits: rehiring an old friend. CEOs who return for a second stint don’t have a learning curve, the thinking goes, and maybe they can spark magic. The iconic example is Steve Jobs, who famously came back to Apple, the company he founded, after 12 years away and transformed the computer maker into the world’s most valuable company.

But Jobs was an exceptional leader—and for second-time-around CEOs, success is more the exception than the rule. That’s the conclusion of researchers at the University of North Carolina who tracked “boomerang CEOs.” In a recent study, the authors reviewed 6,429 CEOs who led S&P 1500 companies between 1992 and 2017, and found 167 cases when a former CEO ricocheted back to the head role. Executives who went from CEO to co-CEO and back at the same company were counted as boomerang leaders.

In general, investors would have been better off if the boomerang hadn’t come back. Companies with such CEOs garnered annualized stock returns (including dividends) that were, on average, 10% lower than those of non-boomerang companies. Their shares also lagged their industries.

What accounts for the disappointing reunions? Some boomerang CEOs returned under unusually tough conditions. A.G. Lafley of Procter & Gamble and Steve Ells of Chipotle each enjoyed great runs in the 2000s, then were overmatched by new crises (e-commerce and E. coli, respectively) in second stints in the 2010s.

But rehiring a familiar face is often a sign of broader management dysfunction. “It’s a failure of CEO succession planning,” says Travis Howell, a researcher on the study. Companies bring in ex-chiefs in part because they have no one else to turn to. (The numbers also show yet another sign of gender imbalance in the C-suite: Only two of the 167 boomerang CEOs were women.)

Boomerang CEOs are also unusually likely to be founders. Founders made up 44% of the UNC study’s boomerang CEOs (compared with only 4% of all CEOs in the sample). And on average, they performed even worse than non-founders. Often a founder’s entrepreneurial vision no longer meshes with the mature company’s needs, Howell notes; the skills involved in growing a new business don’t necessarily translate well to managing a behemoth. (Here again, Jobs is a rare exception.)

Michael Dell’s tenure at his eponymous company illustrates the phenomenon. The stock was a world-beater in its early years. But its shares languished (along with Dell’s personal-computer business) after Dell the man returned as CEO in 2007, falling 43% before Dell took the company private in 2013. (Dell shares have also trailed the market since the company went public again in 2018.) Indeed, boomerang CEOs are particularly unlikely to replicate their past success in tech: The UNC study found that in industries in which the rate of change is rapid, such as technology, returning CEOs perform far worse than in relatively stable sectors, like industrials.

CEOs who aggressively innovate seem more likely to buck the downward trend. Ron Shaich led Panera Bread in the 2000s, then returned as CEO in 2012 after a two-year break. He tore down the organization, investing heavily to reorient the chain around fast-casual service. By 2017, when Panera was sold to private equity firm JAB Holding, its stock had soared. Starbucks’ Howard Schultz was similarly bold when he reoccupied the corner office in 2008, closing stores, retraining staff, and introducing mobile ordering across the company.

The bad aggregate track record notwithstanding, there are a few boomerang CEOs on the current scene who have earned investors’ confidence. Since founder W. Kent Taylor returned in 2011 as CEO of the Texas Roadhouse (TXRH, $52) restaurant chain, the stock has returned 264%, compared with 159% for the broader market. Taylor has earned kudos for keeping price increases modest, allowing the $2.8 billion chain to underprice competitors. Texas Roadhouse’s recent willingness to more efficiently deploy labor at its 611 locations is a positive development for the stock, says KeyBanc analyst Eric Gonzalez.

Then there’s ¬DuPont (DD, $40). Breen’s knack for restructuring remains acute: In early 2021, DuPont will spin off its struggling biosciences and nutrition unit, enabling it to focus on stronger industrial businesses. Its shares trade at valuations far below those of its industrial peers and have been driven even lower by the coronavirus crisis. But many investors believe DuPont’s woes are cyclical and will recede in the long term. “Valuation should be higher,” says Jonas Oxgaard, an analyst at Bern¬stein. Put another way: Breen could deliver the boomerang effect investors like, where fallen share prices return to normal.