Audit

This synthesis ties together four TMG sports studies and seven business research sources. None of those sources is mine. The TMG sports studies are mine but they are not new in this piece. The work this page audits is the connection itself: the mapping between business research and sports findings, and the claims that follow.

The Mallory Group audits its own work as a standard, and applies that standard more carefully when the work depends on citing other people's research. This page is the result.

Found an error or concern with anything on this site? Issues are tracked publicly on GitHub. Open an issue.

What this synthesis can claim

These claims are directly supported by the cited sources, stated conservatively.

The buyer side of the talent market consistently overpays, and experience does not correct it.

• Bidwell shows external hires receive lower performance evaluations for two years despite an approximately 18 percent pay premium over internal promotions.
• Roll documents that acquirer announcement returns are zero or negative on average across thousands of corporate takeovers.
• Boivie, Gee, Gentry, and Graffin show that boards with more CEO-hiring experience produce slightly worse subsequent hires, not better. Experience does not correct the buyer-side error.
• The TMG NHL Free Agency Research finds offseason UFA spending does not significantly predict next-season change after controlling for prior performance.
• The TMG NFL Analysis finds the team that spent the most in free agency rarely produced the biggest win gain.
• The TMG NHL Play for Contract study finds same-team contracts deliver more time on ice per cap share than new-team contracts, with a tier-controlled coefficient of +0.70 minutes per game-equivalent at p < 0.0001.

Star performance is contextual more often than it is portable.

• Groysberg documents an immediate and lasting decline among star Wall Street analysts who switched firms, attributable to lost firm-specific resources, networks, and colleagues.
• The TMG NBA Analysis finds players who changed teams declined more than players who stayed, after controlling for age and prior performance.
• The TMG NHL Play for Contract loyalty discount finding is consistent with the same mechanism.

Talent does transfer when the value-creating asset is portable.

• Berry-Stolzle and Eckles show insurance salespeople build books of business that travel with them, supported by an 89 percent average policy renewal rate.
• Gurun, Stoffman, and Yonker show approximately 40 percent of client assets follow financial advisors when they move between firms. Verified against the paper, which uses the 2004 creation of the Protocol for Broker Recruiting as its identification strategy.

The sports findings and the business research point the same direction across multiple populations and units of analysis.

What this synthesis cannot claim

These are the inferential leaps the piece is careful not to make. Each one is a place where someone reading carefully could push back, and the honest answer is to acknowledge the limit rather than overclaim.

This synthesis cannot claim causation. Most of the TMG sports findings are correlational. The peer-reviewed business research uses stronger identification strategies, but even those vary. Gurun, Stoffman, and Yonker use the 2004 Broker Protocol as exogenous variation, which is solid. Bidwell uses an internal comparison group within the same firms, which is strong but not experimental. Roll's hubris hypothesis is an inferential argument, not a causal estimate. The synthesis treats these collectively as evidence pointing in one direction, not as proof.

This synthesis cannot rule out effort as a mechanism. The piece argues that performance is contextual: stars decline because they lose their firm's resources, system, and colleagues. An alternative explanation is that stars coast after a big payday. The available research does not isolate effort as a separate mechanism cleanly. A football study found a small but statistically significant effort reduction after players signed follow-up contracts with new clubs, but no equivalent study exists for business stars. Groysberg explicitly attributes the analyst decline to lost firm-specific resources, but his data does not separate effort from context. The honest claim is that performance is contextual. The honest non-claim is that effort is unchanged.

This synthesis cannot generalize across all industries. Insurance and financial advisors are different industries from athletic performance and equity research. The portability spectrum frames across populations, but the comparison crosses industries and structures. Time on ice per cap share is not the same metric as client assets retained. The reader should treat the spectrum as a conceptual organizing tool, not as a single quantitative scale.

This synthesis cannot claim that buying outside talent never works. The portability counterpoint exists for a reason. The piece argues that paying for outside talent rarely delivers the expected return when the asset is firm-specific or contextual. It does not argue that all external acquisitions fail. The implications page treats this carefully.

This synthesis cannot rule out selection effects. NBA players who change teams may already be in decline when they move (the TMG study flagged this). Star analysts who switch firms may already be struggling at their old firm. NHL same-team signings may correlate with players who fit the team well to begin with, and the loyalty discount may partly reflect that selection rather than the retention itself. The studies that use stronger identification (Gurun et al., Bidwell) address this better. The TMG sports findings address it less.

This synthesis cannot infer where lost assets go. The Cerulli industry research reports a 22 percent asset loss during advisor transitions. It does not state that 78 percent follow the advisor. Lost assets could have stayed with the original firm under a new advisor, moved to a third party, gone independent through the client, or remained unaccounted for. Reporting the complement as portability would be an inference, not a finding.

This synthesis does not represent the full literature on outside hiring. The source set was selected to test one mechanism: whether performance is portable. Every primary anchor (Groysberg, Bidwell, Roll, Boivie) examines cases where outside talent underdelivers, and the two portability counterpoints examine when transfer works, not whether outside hiring beats internal development. There is a separate body of peer-reviewed research on the upside of outside hiring, including the role of external hires in bringing new knowledge, driving innovation, leading turnarounds, and breaking up complacent cultures. That literature is not engaged here. A reader should treat this synthesis as evidence on the portability mechanism specifically, not as a complete account of when outside hiring succeeds or fails.

This synthesis measures performance return only. It treats a premium as overpayment when on-field or on-the-job performance does not justify the price. It does not account for secondary returns that can make a premium rational even when performance does not. In sports, a team may sign a star for jersey sales, fan engagement, or to deny a rival the asset. In business, a firm may hire a known name to signal direction to investors or the market. Where those are the goals, performance was never the only objective, and "overpayment" is the wrong frame. The synthesis speaks to performance return, not to these secondary returns.

Strength of each source-to-sports mapping

A seven-by-one table of how directly each business research source maps to the TMG sports findings.

The pattern across the seven rows: three strong mappings (Groysberg, Bidwell, Boivie et al.), one medium mechanism-level support (Roll), two medium portability counterpoints (Berry-Stolzle, Gurun), and one weak industry corroboration (Cerulli). Bertrand and Mullainathan (2001) is cited as a footnote and does not carry mapping weight.

Further reading and open questions

Two sources came up during the research that did not make the final source list but are worth investigating later.

Kim, H. and Low, A. (2026). Who hires whom? Connected hiring in the CEO labor market. Journal of Empirical Finance, 87, 101696. This paper studies CEOs hired by board members they already know. It tests a related but different question from the one this synthesis asks: whether incoming connections affect hire selection, rather than whether outgoing colleagues affect post-move performance. Worth pulling for a future expansion.

The football effort-reduction paper cited in the no-claim section above (Weimar and Scharfenkamp, 2019, "Effort reduction of employer-to-employer changers: Empirical evidence from football," Managerial and Decision Economics, 40(3), 277-291) is the closest sports analog to the effort question. Worth pulling if a future piece focuses on the effort versus context mechanism.

The Khurana research on star CEOs (referenced briefly in the earlier scoping discussion) was not included in this synthesis but covers adjacent territory and could be pulled for a related future piece.

One open question raised during drafting also belongs here.

The persistent decision-maker question. The cited research identifies that buyers overpay (Roll), that experience does not correct this (Boivie, Gee, Gentry, and Graffin), and that pay decisions are worse when principal oversight is weak (Bertrand and Mullainathan). None of these papers isolates the specific dynamic where executors of bad hiring decisions (talent acquisition, HR, general managers in sports) get replaced while the persistent decision-makers who set the culture (boards, founders, owners) keep their seats and continue making the same decisions through new executors. This is a logical next step from the existing research but not directly studied in the sources used here. If this direction is added to a future piece, the Bertrand and Mullainathan (2001) source should be reassessed. It is currently a footnote-only citation. The "without principals" framing in the original paper's title is directly relevant to the persistent decision-maker question, and the source would likely deserve featured placement rather than footnote placement if the question becomes central.

Summary

This synthesis is a connection piece, not new empirical research. It is as strong as the sources allow and no stronger. The convergence between the TMG sports findings and the business research is the value the piece offers. The limits are documented above so a reader can decide how much weight to put on each claim. The strongest claims are the buyer-side overpayment pattern and the contextual nature of star performance. The weakest claims are anything that requires inferring effort, causation, or where lost assets go. The synthesis avoids those.

Corrections record

Corrections found during audit are recorded here. The corrected content appears throughout the site. The original errors appear only on this page, as a record of what the audit process caught.

June 2026, pre-publication audit. An independent reviewer ran the five verification checks before the site was published. Findings:

• Bidwell premium figure corrected. The draft stated external hires were paid "18 to 20 percent" more than internal promotions. The paper states they are "initially paid around 18 per cent more than the promoted workers and have higher levels of experience and education." The figure was corrected to "around 18 percent" throughout the site, and the experience-and-education detail was added to the finding.
• Protocol name refined. The draft used the common name "Broker Protocol." The Gurun, Stoffman, and Yonker paper uses the official name "Protocol for Broker Recruiting." First mentions throughout the site now use the official name with the common name noted.
• Groysberg verification incomplete. The reviewer could not verify the Groysberg findings against the HBR companion articles because they are paywalled. This limitation is disclosed in the citations annex, in the strength table above, and in Check 4 below.
• All other checked claims (Boivie findings including the "superstitious learning" phrasing, the Gurun 40 percent figure and sample size, and citation details) were confirmed against the sources.

Verify the claims yourself

The Mallory Group audits its own work as a standard. The next step in that standard is letting readers do the same. Five load-bearing claims in this synthesis can be checked against open access sources in about forty-five minutes. Tick the box once you have verified a claim, or open a GitHub Issue if you find something that does not match.

Your progress is saved in your browser. Issues submitted by anyone are visible at the bottom of this section.

Check 1. Bidwell external hire premium

Claim: External hires were paid around 18 percent more than internal promotions and received lower performance evaluations during their first two years.

Source to verify against: Bidwell 2011 PDF (Wharton)

What to check: Search the PDF for the approximately 18 percent premium and the two-year evaluation finding.

• I verified this claim against the source.

[Report an issue with this claim] (link populated by JavaScript)

Check 2. Boivie experience-does-not-help finding

Claim: Boards with more CEO-hiring experience produce slightly worse subsequent CEO performance, not better. The authors describe this as suggestive of superstitious learning.

Source to verify against: Boivie et al. 2025 (Wiley, open access)

What to check: Read the abstract and search the PDF for "superstitious learning" to confirm the authors use that specific phrase.

• I verified this claim against the source.

[Report an issue with this claim]

Check 3. Gurun 40 percent of assets follow

Claim: About 40 percent of client assets follow financial advisors when they move between firms, based on a sample of over 760,000 advisors and approximately $28 trillion in assets.

Source to verify against: Gurun, Stoffman, and Yonker 2021 SSRN working paper

What to check: Read the abstract. Confirm the 40 percent figure refers to client assets following advisors specifically.

• I verified this claim against the source.

[Report an issue with this claim]

Check 4. Groysberg firm capabilities framing

Claim: Star Wall Street analysts who switched firms declined in performance. Exceptions included stars who moved with their teams and stars who moved to firms with higher capabilities and resources.

Source to verify against: Two Harvard Business Review companion articles by the same author:

• The risky business of hiring stars (Groysberg, Nanda, and Nohria, 2004)
• Are leaders portable? (Groysberg, McLean, and Nohria, 2006)

What to check: Confirm the articles describe the analyst decline finding, the team-move exception, and the firm-capabilities-and-resources framing.

Limitation: Both HBR articles are paywalled. A June 2026 pre-publication audit could not verify the Groysberg findings against these primary sources for this reason. The findings are currently supported by secondary summaries and reviews only. The named firms (Credit Suisse First Boston, Salomon Brothers, Goldman Sachs, Merrill Lynch) come from the book Chasing Stars, which is also paywalled. This check is the most valuable one a reader with HBR or institutional access can complete.

• I verified this claim against the source.

[Report an issue with this claim]

Check 5. Citation details and links

Claim: Every citation in the synthesis (journal name, volume, issue, page numbers, author names, links) is accurate.

Source to verify against: The citations annex.

What to check: Click through any source link in the annex. Confirm the journal page shows the same volume, issue, page numbers, and authors as cited in the annex.

• I verified the citation details across the annex.

[Report an issue with a citation]

Your audit progress

Progress: 0 of 5 checks complete. (This counter updates as you tick boxes above.)

Reader-flagged issues

The list below pulls open audit issues from GitHub in real time. If a claim has been flagged, you can see what the concern is and how it has been resolved.

(Live list populated by JavaScript fetching from the GitHub Issues API.)

Submit other feedback

If you have feedback outside the five checks above, [open a general feedback issue](link to GitHub issue creation). All feedback is public and helps the synthesis stay accurate over time.

Where to go next

• Back to welcome.
• Key Findings: Plain-language summary of the four TMG sports studies and the business research synthesis.
• Business Research: Each source mapped onto the sports findings in detail.
• Implications: What this means for hiring, managing stars, and structuring work.
• Citations: Full bibliography with links to every source.