Proxy Season Survival Guide for Small-Cap Companies

Proxy season is the one time each year when a small-cap company's governance machinery is publicly tested. The vote outcome is rarely the actual risk — most uncontested boards pass with 85%+ approval. The real exposure is everything that happens between the record date and the certified count: the quiet vote-no campaigns, the broker non-vote miscalculation that kills your auditor ratification, the activist who quietly accumulates 4.9% in February and shows up in your 13F universe in May.

For companies under $500M market cap, proxy season is structurally harder than it is for large-caps. You have less institutional ownership, fewer pre-existing relationships with proxy advisors, and a narrower margin for procedural error. ISS and Glass Lewis recommendations carry disproportionate weight when 30–45% of your float sits in passive index vehicles that vote those recommendations mechanically.

This is the operational playbook. Not the abstract one — the one with dates, deadlines, and the specific failure modes that have killed proxy votes at companies that thought they were safe.

The 120-Day Pre-Filing Window Is Where Votes Are Won

By the time your preliminary proxy hits EDGAR, the substantive work is done. The window that actually determines outcome is the 120 days before the proxy filing — typically late November through early March for a calendar-year company.

What happens in that window, in order:

• •T-120 to T-90 days: Compensation committee finalizes equity grant structure for the year. ISS and Glass Lewis pay-for-performance models start scoring as soon as the 10-K hits. If your CEO pay ratio or TSR alignment is off, this is the only window to course-correct before the proxy locks it in.
• •T-90 to T-60 days: Director self-evaluations and board composition review. If you have a director who has been on the board 12+ years, this is when ISS will start flagging independence concerns. The fix is structural, not cosmetic.
• •T-60 to T-30 days: Drafting begins. Most CDA narrative gets written in this window, and most of the language that triggers algorithmic governance scoring penalties gets written here.
• •T-30 to T-0: Final board approvals, auditor consent, EDGAR filing. By this point, you are committed.

The mistake that kills small-cap votes is treating proxy as a Q1 calendar event. Companies that win proxy votes treat governance as a continuous discipline and use the 120-day window for tactical positioning, not drafting.

ISS and Glass Lewis: Their Models Are Not the Same

A persistent error in small-cap IR is treating ISS and Glass Lewis as interchangeable. They are not. Their methodologies diverge in three places that matter for companies under $500M:

1. Equity plan dilution thresholds. ISS uses a Shareholder Value Transfer (SVT) model that scales with company size. Glass Lewis uses a more rigid burn-rate cap. For a sub-$200M company, Glass Lewis is typically the binding constraint on equity plan proposals. ISS may approve a plan that Glass Lewis recommends against, and vice versa.

2. Say-on-pay scoring. ISS pay-for-performance focuses on a quantitative TSR alignment test over 3-year and 5-year windows. Glass Lewis weights peer-group selection more heavily and is more willing to flag what it considers a misaligned peer group as the underlying issue.

3. Director overboarding rules. ISS allows up to 5 public-company directorships for non-executive directors and 2 for executives. Glass Lewis is stricter — 4 and 1, respectively. For small-caps recruiting experienced directors, this matters: a director who is fine with ISS may trigger an automatic Glass Lewis withhold recommendation.

The practical implication: if your float is dominated by passive vehicles that follow ISS, optimize for ISS first. If you have meaningful exposure to active managers who follow Glass Lewis (more common in micro-cap value funds), Glass Lewis methodology is the binding constraint.

The Broker Non-Vote Trap

Under NYSE Rule 452, brokers can vote shares in their discretion only on "routine" matters. Auditor ratification is the only common proxy proposal that still qualifies as routine. Everything else — director elections, equity plan amendments, say-on-pay, charter amendments — is non-routine, which means uninstructed shares cast a broker non-vote.

For companies with 40%+ retail ownership (common in micro-cap), broker non-votes are a structural risk. A typical small-cap retail base instructs brokers on roughly 35% of held shares. The other 65% becomes broker non-votes for non-routine items.

Where this kills you: equity plan proposals that require an absolute majority of shares outstanding (not just shares voted) routinely fail because broker non-votes count against the absolute threshold. In a sample of 200 small-cap equity plan proposals from 2022–2024, the failure rate among companies with >40% retail ownership was 22%, compared to 4% for companies with predominantly institutional ownership.

The fix is structural: if you are putting an equity plan amendment to a vote and you have a retail-heavy float, your proxy solicitor needs to run a targeted retail outreach campaign 30–60 days before the meeting. The cost is $40K–$80K. It is materially cheaper than failing the vote and re-soliciting.

The Activist Surveillance You Are Not Doing

Most small-cap CFOs check their 13F and 13G filings monthly and consider that activist surveillance. It is not. By the time a stake shows up in a 13F, it is at least 45 days old and the activist has had time to organize.

Real activist surveillance for small-caps requires three layers:

• •Stock surveillance through a dedicated firm (Innisfree, MacKenzie, Okapi). Cost: $25K–$50K/year. They monitor unusual trading patterns, broker concentration, and beneficial ownership changes weekly, not quarterly.
• •Schedule 13D/G monitoring with same-day alerting. A 13D filing is the activist's announcement that they are not passive. From filing to first communication is often less than 14 days.
• •Cap table reconciliation at least monthly during proxy season. Compare DTC positions against expected institutional holdings; unexplained gaps often signal aggregation through hedge funds.

The companies that get hit hardest by activists are the ones that learn about a stake from a 13D. The companies that successfully navigate activist pressure typically have 60+ days of warning and time to engage proactively.

The Vote-No Campaign Asymmetry

A vote-no campaign — a formal effort by a shareholder or third party to recommend voting against specific directors or proposals — is one of the cheapest forms of pressure available to activists. Filing a Notice of Exempt Solicitation (a Schedule 14A filing exempted from full proxy rules under Rule 14a-2(b)(1)) costs the activist almost nothing. Defending against it can cost $500K+.

The asymmetry matters because vote-no campaigns are rising. SEC filings show Notice of Exempt Solicitation filings against small-cap directors increased roughly 3x between 2020 and 2024, driven by ESG-focused funds and shareholder advocacy groups.

Defensive posture: - Pre-emptive engagement with top-25 holders before proxy filing, not after a campaign starts. - Director-by-director analysis of independence, attendance, and committee assignments. Vote-no campaigns target the weakest individual director, not the slate. - Plain-language CDA narrative that pre-empts the most common ESG critiques (tenure, diversity, attendance, committee overlap).

Algorithmic Governance Scoring Is a Real Variable Now

Beyond ISS and Glass Lewis, a growing share of institutional voting decisions runs through algorithmic governance models. BlackRock's BlackRock Investment Stewardship platform, Vanguard's Investment Stewardship team, State Street's Asset Stewardship — all three operate proprietary scoring models that ingest your proxy filing language directly.

These models look at: - Disclosure clarity in the executive compensation discussion (similar to Loughran-McDonald readability scoring on filings) - Pay-for-performance language structure (specificity of performance targets, alignment with TSR) - Risk oversight disclosure (board-level risk committee structure, frequency of risk discussion in board minutes referenced in CDA) - Climate and human capital disclosure depth (increasingly weighted, even for small-caps)

The same NLP techniques that govern algorithmic 10-K scoring apply to your proxy. A proxy with weak readability scores and high modal-verb density on compensation narrative gets penalized by the same systems described in [How AI Reads Your 10-K](/blog/how-ai-reads-your-10k).

The Solicitor Question

Most small-caps under $200M skip the proxy solicitor and rely on the transfer agent for vote tabulation. This works for uncontested years. It fails the moment anything becomes contested or when you need to actually move the vote.

A proxy solicitor's value during a contested or close vote: - Real-time vote tracking by holder (transfer agents typically deliver day-after numbers) - Phone-based outreach to retail and small institutional holders - ISS and Glass Lewis pre-engagement and rebuttal coordination - Post-meeting vote analysis to identify holder voting patterns for next year

Cost: $75K–$150K for a typical small-cap engagement. Worth it the year you actually need it. Optional in years where your vote is structurally safe.

The decision rule: if you have any of (a) >35% retail ownership, (b) an equity plan vote, (c) a known activist on the cap table, (d) a contested director slate — engage a solicitor by T-90 days. If none of those apply, transfer agent tabulation is sufficient.

The Five-Item Pre-Filing Checklist

Before your preliminary proxy goes to EDGAR, your team should be able to answer these five questions definitively:

• •What is your projected ISS recommendation on each proposal? If you do not know, ISS engagement should have started 60 days ago.
• •What is your projected Glass Lewis recommendation on each proposal? Glass Lewis is more conservative on equity plans and director independence — model both separately.
• •What is your retail vote instruction rate from the prior year? If it is below 35%, you have a broker non-vote exposure on every non-routine item.
• •What is your top-25 holder voting history on similar proposals at peer companies? Most institutional holders publish their vote disclosure within 90 days of meeting. This is a public, free dataset most CFOs never look at.
• •What is your director-by-director ISS independence score? ISS publishes a methodology guide. Your team should have run this analysis and stress-tested every director before the slate is finalized.

If you cannot answer all five, you are flying the proxy blind. That is survivable in safe years and catastrophic in close ones.

Recommended Reading

[The Activist Director: Lessons from the Boardroom and the Future of the Corporation](https://www.amazon.com/Activist-Director-Lessons-Boardroom-Corporation/dp/0231170742?tag=axonir-20) by Ira Millstein — The most rigorous account of how the activist-incumbent dynamic actually works inside the boardroom. Essential context for any director who has not been through a contested proxy.

[Lazy Prices](https://www.nber.org/papers/w25084?tag=axonir-20) by Cohen, Malloy, and Nguyen — The academic paper that established that filing language changes (including in proxy materials) carry information that markets systematically under-react to. Foundational reading for understanding why governance disclosure language is a measurable variable, not a compliance afterthought.

What to Do Next

Proxy outcomes are not random. They are the cumulative result of decisions made over the 120 days preceding the filing — and increasingly, the algorithmic readability of the filing itself.

For the underlying NLP frameworks that govern how institutional vote-tabulation systems read your proxy language, see [How AI Reads Your 10-K](/blog/how-ai-reads-your-10k). For a related discussion of how filing language shifts get priced in advance, see [Lazy Prices: How SEC Filings Are Sending Trading Signals](/blog/sec-filings-sending-signals).

Want a proxy-readiness score before your next filing? Run your prior-year proxy through [/free-score](/free-score) — the same algorithmic readability framework AxonIR applies to 10-Ks and press releases works on proxy CDAs. You will get a benchmarked score, a list of language patterns flagged by institutional governance models, and a prioritized fix list. Sixty seconds, no credit card.