BETR (Better Home & Finance Holding Company) — Consensus Proxy Analysis

---

Consensus Summary

All four models, reasoning from the same factual extraction, recommend supporting management against Vishal Garg's written consent solicitation. The contest involves a founder-CEO removed by unanimous independent board vote who then launched a rapid-strike consent campaign to reconstitute the board in a manner that would plausibly facilitate his return to executive control. The models share a consistent assessment across three dimensions: (1) Garg's financial stewardship was objectively poor, with cumulative losses exceeding $1.5 billion and a stock decline exceeding 90% from the de-SPAC transaction; (2) Garg's consent solicitation suffered from material procedural and legal deficiencies, including premature solicitation, false majority claims, and the absence of named replacement directors; and (3) the current board's improving financial trajectory, articulated strategic plan, and governance rationale for removal provide a stronger basis for continuity than Garg's campaign offers for change. The primary source of uncertainty across all models concerns the board's post-firing offer of $750,000 plus 875,000 shares to Garg — a gesture difficult to reconcile with an "unfit to lead" characterization — and the absence of a permanent CEO. Note that all conclusions below derive from a single factual extraction, not independent readings of the underlying filings; convergence reflects a shared source, not verified independent corroboration.

---

Model Comparison

---

Points of Agreement

All four models, drawing from the same extraction, converged on the following readings:

Financial Track Record: Each model characterized the cumulative losses (exceeding $1.5 billion since 2022), the >90% stock decline since the August 2023 de-SPAC, and the improving but still loss-generating trajectory as the central financial backdrop. All treated the improving loss trend — from $871M in 2022 to approximately $165M in 2025 — as a point favoring continuity rather than disruption.

Procedural Deficiencies in the Consent Solicitation: All four models flagged the false majority claims (publicly retracted), premature solicitation before a compliant consent statement was filed, the alleged failure to timely disclose coordinated group formation under Section 13(d), and critically, the absence of named replacement directors as material flaws. All treated the Activant Amendment No. 7 disclaiming coordination as a significant development undermining Garg's claimed majority.

Governance Conduct Concerns: All models noted the same constellation of conduct: the Zoom layoffs, degrading employee communications, the Triaxx fiduciary duty and self-dealing allegations by major institutional counterparties, the jury finding of breach of fiduciary duty and conversion, related-party transactions with Garg-affiliated entities, the refusal to sign Q2 10-Q representation letters, and the allegation of disclosing non-public information in the August 13 press release. These are the filing's characterizations, principally from the board's submissions, and should be understood as the board's stated positions rather than adjudicated facts — though the models generally treated the jury finding as a matter of public record.

Strategic Plan Assessment: All models credited management's articulated plan (Interim CEO Daniel Lewis, $45M+ annualized cost reductions, enterprise partnerships, Tinman AI platform, UK bank sale) as more coherent than Garg's campaign, while noting Garg's disclaimers of CEO ambition were contradicted by documented statements from his counsel and press releases.

Rights Plan as Proportionate: Each model, to varying degrees, characterized the 15% threshold poison pill with 2027 annual meeting expiration as a defensible rather than clearly entrenchment-motivated measure given the context.

---

Points of Divergence

Confidence in the Recommendation — Gemini vs. the Others: The most striking divergence is Gemini's 10/10 confidence rating versus Claude and OpenAI's 7/10 and Grok's 8/10. This is not a difference in available information — all four received identical input. Gemini appears to have weighted the procedural deficiencies and governance conduct record as essentially dispositive, leaving no meaningful residual uncertainty. Claude and OpenAI explicitly identified the board's post-removal financial offer to Garg ($750,000 cash + 875,000 shares) as a genuine credibility problem that limits confidence. Grok acknowledged the same weaknesses but discounted them more heavily in favor of the governance case. Gemini's 10/10 rating is analytically aggressive and likely understates genuine uncertainty, particularly regarding the Delaware Chancery and SDNY litigation outcomes and the possibility that Garg genuinely commands majority stockholder support absent the Activant group.

The Post-Removal Retention Offer: Claude gave the most sustained attention to the board's offer of $750,000 cash plus 875,000 shares to Garg as a Vice Chairman — characterizing it as "difficult to reconcile with a 'unfit to lead' characterization" and treating it as a genuine analytical problem. Grok and OpenAI mentioned it but did not weight it as heavily. Gemini did not address it as a distinct concern. This divergence turns on how much analytical weight to assign a governance inconsistency that Garg exploits rhetorically, but that management can explain as a founder transition management tactic. The disagreement is substantive and unresolved: the board's conduct here is either a cynical contradiction or a pragmatic retention attempt, and the extraction does not definitively resolve which.

Framing of Garg's Strategic Case: OpenAI gave slightly more neutral treatment to Garg's strategic proposals (buyback, $1 salary, UK bank sale completion) than the other three models, which were more explicitly dismissive. Claude noted the buyback would consume significant cash from a still loss-generating company; Grok and Gemini described the plan as lacking credibility beyond personal reinstatement. OpenAI characterized it as a "lack of detailed plans" rather than as affirmatively implausible. This divergence is modest and turns on how skeptically to read strategic commitments made in the context of a contested solicitation.

Emphasis on Litigation Risk: Claude was the only model to give sustained attention to the Delaware Chancery and SDNY litigation as material sources of outcome uncertainty capable of affecting the ultimate result regardless of the underlying merits. The other models noted litigation defensively but did not treat it as a distinct analytical variable affecting confidence. Given that litigation outcomes are genuinely unknowable from the extraction alone, Claude's caution here is analytically sound.

---

Consensus Recommendation

Support Management

Strength: Strong

The case for supporting management rests on four pillars that all models treated as load-bearing: the objectively poor financial stewardship under Garg, the material procedural and legal deficiencies in the consent solicitation (particularly the absence of named replacement directors and the false majority claims), the documented conduct record creating ongoing legal and reputational risk for public stockholders, and the improving operational trajectory under current board oversight. The appropriate vehicle for a governance challenge with genuine majority support is the 2027 annual meeting, where compliant solicitation with disclosed nominees can occur. Public stockholders should revoke any previously granted consents and return the WHITE consent revocation card.

The "Strong" characterization reflects near-unanimity on recommendation and substantial agreement on reasoning, while acknowledging that residual uncertainty — from pending litigation, the board's retention offer inconsistency, and the permanent CEO vacancy — prevents this from being unqualified.

---

Confidence Score

Confidence: 7/10

The consensus confidence is set at 7/10, reflecting the unanimous recommendation direction but appropriately discounting Gemini's 10/10 outlier in favor of the more analytically careful 7-8/10 assessments from Claude, Grok, and OpenAI. Key sources of residual uncertainty: (1) the Delaware Chancery and SDNY litigation outcomes could materially alter the legal landscape; (2) the board's post-removal financial offer to Garg is a genuine credibility inconsistency that remains unexplained; (3) if Garg can demonstrate genuine majority stockholder support without Activant, there is a democratic governance argument that complicates the picture; and (4) the permanent CEO vacancy means the board's own leadership transition is incomplete. Critically, all analysis derives from a single factual extraction — any error or omission in that extraction would propagate uniformly across all four model analyses without detection, which alone warrants holding confidence below 8/10 regardless of the apparent strength of the case.