Consensus Synthesis: VSCO Proxy Contest — BBRC International vs. Victoria's Secret & Co.

Re-election of Board Chair Donna A. James | 2026 Annual Meeting

---

Consensus Summary

Three of four models recommend supporting the activist (BBRC International), with one model (OpenAI) dissenting in favor of management. The majority view holds that Donna James's 25-year tenure, a multi-year record of structural underperformance, demonstrable capital allocation failures (most notably the Adore Me acquisition), a problematic executive compensation program, and the unilateral adoption of a poison pill collectively constitute sufficient grounds to vote against her reelection as Board Chair. The minority view (OpenAI) places greater weight on the recent operational turnaround under CEO Hillary Super and characterizes management's engagement efforts as sufficiently responsive to shareholder concerns.

All four models agree that the recent financial trajectory under Super is genuinely positive and deserves serious consideration. The central point of divergence is whether that recovery — approximately 18-21 months old — is sufficient to offset the governance failures and structural underperformance that accumulated over the preceding decade-plus under James's oversight. The majority concludes it does not; the minority concludes it does.

The contest's narrow scope — removing the Chair only, with no alternative director proposed and no change of control at stake — further reinforces the majority recommendation, as the downside risk of a vote against James is structurally low.

---

Model Comparison

---

Points of Agreement

All four models converge on the following substantive conclusions:

1. The Recent Turnaround Is Real and Meaningful
Every model acknowledges the exceptional TSR performance since Hillary Super's appointment (193% from appointment through May 2026), five consecutive quarters of consensus beats, and improving comparable sales trends. No model dismisses this evidence. The disagreement is about how much weight it should carry in evaluating James's continued tenure as Chair, not whether it exists.

2. Donna James's Tenure Is Anomalously Long
All models flag 25 years as significantly exceeding governance norms across any major framework — NACD guidelines, ISS/Glass Lewis standards, S&P 500 averages. No model defends the tenure length on its merits; OpenAI's dissent rests on experience and qualifications, not on a defense of the tenure duration itself.

3. The Adore Me Acquisition Represents a Capital Allocation Failure
Every model identifies the $591 million acquisition, subsequent impairments of $155.9 million, missed EBITDA targets, and zero earnout achievement as a significant governance failure on the Board's watch. Management's defense of this transaction is judged inadequate by all four models.

4. Executive Compensation Design Is Problematic
All models note the sandbagging of incentive targets (set below prior-year actuals in 3 of 4 years), the dual-benchmark asymmetry in PSU design, and the failure to substantively modify the program after 25% say-on-pay opposition. Management's rebuttal is found unpersuasive on this point across all analyses.

5. BBRC's Conduct Concerns Are Real but Largely Irrelevant to the Ballot Question
Every model acknowledges the HSR filing violations, the retail store access incident, Blundy's remuneration strikes at Lovisa, and the competitive overlap concern from Léays. All models also agree that since BBRC is not seeking board representation, these concerns are largely deflections from the core question of whether James should continue as Chair.

6. The Narrow Contest Scope Reduces Downside Risk
All models note that BBRC's singular ask — removing the Board Chair without proposing a replacement or change of strategy — is the lowest-risk form of activism. Even the dissenting model (OpenAI) does not argue that removing James would materially disrupt operations.

---

Points of Divergence

1. Weighting of Recent Turnaround vs. Cumulative Governance Record
This is the primary substantive disagreement. The three majority models (Claude, Grok, Gemini) treat the turnaround as attributable primarily to CEO Super rather than to James's chairmanship, and argue that 18-21 months of recovery is insufficient to offset documented failures accumulated over 25 years. OpenAI weighs the turnaround more heavily and treats James's role in selecting and supporting Super as a meaningful positive contribution warranting continued service.

2. Attribution of Turnaround Credit
Claude and Gemini make the most explicit argument that the Company cannot simultaneously credit James for the turnaround while also framing Super's appointment as marking a "new era" — the logical tension undermines the management rebuttal. Grok implies the same. OpenAI does not engage with this attribution tension as directly, crediting management more holistically.

3. Responsiveness to Shareholder Feedback
OpenAI characterizes management's engagement with BBRC as a meaningful positive signal, citing willingness to consider board refreshment and various dialogue attempts. Claude specifically rebuts this framing, noting that the Company's response to 25% say-on-pay opposition — maintaining the same PSU design while adding a discretionary modifier — is the opposite of responsive engagement. Gemini concurs with Claude on this point. Grok notes the governance concerns but does not analyze the engagement history in equal depth.

4. Poison Pill Analysis
Claude treats the unilateral poison pill adoption as "disqualifying" for a governance-focused vote, characterizing it as a direct anti-shareholder action that ISS/Glass Lewis would flag for board chair accountability. Gemini labels it a "significant red flag." Grok mentions it as a signal of entrenchment. OpenAI acknowledges the litigation and rights plan but does not assign it similar weight in the final calculus.

5. Buyback Execution
Claude and Gemini both find management's VWAP rebuttal technically accurate but strategically evasive, arguing that the governance concern is about the mechanical rather than opportunistic framework for buybacks. Grok concurs directionally. OpenAI accepts management's defense of the buyback program more readily.

---

Consensus Recommendation

Support Activist

Strength: Moderate

Three of four models recommend voting AGAINST the reelection of Donna James as Board Chair, using the BLUE proxy card. The dissent from one model reflects legitimate uncertainty about the timing of a Chair transition relative to an early-stage operational recovery, and this minority view should inform institutional shareholders who place high weight on management continuity during turnarounds.

The majority recommendation rests on five convergent pillars that persist across three independent analyses:

1. Tenure at 25 years is indefensible under any major governance framework and is un-rebutted by management on the merits
2. Adore Me represents a documented, material capital allocation failure under Board oversight, with $155.9 million in impairments and zero earnout achievement
3. Compensation program design is persistently problematic, with sandbagged targets, benchmark asymmetry, and failure to respond to 25% say-on-pay opposition
4. Poison pill adoption without shareholder approval in direct response to BBRC's accumulation is a serious anti-shareholder governance action that carries Chair-level accountability
5. Contest scope is narrow, making this a pure governance signal with limited execution risk — a Chair transition does not disrupt Super's operational mandate or the management team's strategic agenda

The minority view merits acknowledgment: if the Board demonstrates continued responsiveness — board refreshment, improved compensation design, shareholder rights plan review — without a contested vote, the case for disruption weakens. Shareholders who believe management has turned a corner may reasonably weigh current momentum more heavily.

---

Confidence Score

Confidence: 7/10

The consensus confidence mirrors the individual model scores precisely, reflecting strong directional alignment among the majority (3 of 4 models) while accounting for meaningful uncertainty introduced by the genuine operational turnaround under Super. Key risk factors reducing confidence from a higher score: (1) the possibility that James-specific removal mid-turnaround introduces unnecessary governance noise; (2) execution risk under the majority voting standard, given the Board's demonstrated reluctance to respond to shareholder signals; (3) BBRC's credibility limitations (HSR violations, store access incident), which may dampen the coalition-building needed to achieve majority opposition; and (4) the OpenAI model's legitimate observation that management engagement and recent performance provide a reasonable basis for the minority view. A score of 7 reflects high directional confidence in the governance case against James with appropriate acknowledgment of the timing and execution dimensions that introduce real uncertainty.