Florida's Medicare Advantage Plans Can Now Close Mid-AEP for 2027

The letter about to land in your clients' mailboxes

If you write Medicare Advantage business in Florida, some of your Humana clients are getting, or have just gotten, a letter telling them their plan won't exist in 2027. Humana disclosed on its July 29, 2026 second quarter earnings call that its planned 2027 plan exits will affect approximately 600,000 members nationally, as the company narrows its Medicare Advantage lineup toward higher performing plans and works back toward a stated pre-tax margin target of at least 3% by 2028.⁷¹⁰ Newsweek independently reported the same figure, tied to the same earnings call, on September 9, 2026.⁸

The timing isn't a coincidence, and it isn't optional for Humana either. Federal rule 42 CFR 422.506(a)(2)(ii) requires a Medicare Advantage organization that won't renew a plan to notify every enrollee by mail at least 90 calendar days before the non-renewal takes effect.⁶ Count 90 calendar days back from the start of a new plan year and you land in the first days of October, which is exactly what trade coverage of Humana's 2027 exits has reported, naming October 2, 2026 as the date the notices had to be in the mail.⁹

Here's the part that makes this AEP different from the last several. In past years, a non-renewal meant your client shops among whatever plans remain, on the same terms as anyone else. For 2027, some of those replacement plans can legally refuse to take them, not because they're ineligible, but because the plan already has enough members.

Why a Medicare Advantage plan can legally close to new members

A Medicare Advantage plan can close to new enrollees in 2027 because CMS's own contract year 2027 bid rules let an insurer request an enrollment capacity limit as part of its annual bid submission. The relevant regulation is 42 CFR 422.254(c)(4) and (e)(1), and CMS's Final Contract Year 2027 Standards for Part C Benefits, Bid Review and Evaluation memorandum spells out how it works: "MA organizations may submit an enrollment capacity limit (if any) as part of their bid submission for a specific plan... Requesting an enrollment capacity limit as part of the bid is optional."⁴

That's new leverage for insurers, and CMS built in a real guardrail. The same memo states that "CMS will only accept enrollment capacity limits at or above current enrollment as of June 1, 2026, for plans' CY 2027 planned service area," and that "MA organizations may not disenroll existing members to reach a capacity limit."⁴ In plain terms, an insurer can't use this to shrink a plan it already has. What it can do is draw a line above today's membership and refuse to let the plan grow past it, which is a very different business decision than exiting the plan altogether, and one your clients have no reason to expect.

An insurer can't use this to shrink a plan it already has. What it can do is draw a line above today's membership and refuse to let it grow past that line.

The path that already existed, and the one that's new

A plan restricting its own enrollment isn't a brand new idea. Before CY2027, the only way an insurer could cap a plan was to request special CMS approval under 42 CFR 422.60(b), and CMS's own bid review memo says it "will consider such requests only if the health and safety of beneficiaries is at risk."⁴ That's a narrow, exceptional path, the kind used when a plan's provider network genuinely can't absorb more patients safely.

The bid based option for CY2027 is a different animal. It doesn't require a health or safety showing at all. An insurer can request a cap simply because it wants one, as a normal part of filing its bid, for any business reason it chooses. That's the actual change worth understanding: CMS didn't just tighten an existing safety valve, it opened a second, much wider door that has nothing to do with network capacity and everything to do with how an insurer wants to manage its own growth.

How an enrollment cap actually works, in CMS's own words

Once CMS approves a plan's cap, the mechanics are strict and impersonal. CMS's Medicare Advantage and Part D Enrollment and Disenrollment Guidance for contract year 2027 states plainly, in section 50.10.1: "An MA plan with an established enrollment capacity limit must accept all valid enrollments in the order they are received until the capacity limit is reached. An MA plan that has reached a CMS approved capacity limit must remain closed to all applicants until space becomes available."⁵

Read that twice. It doesn't say the plan closes to enrollees from agents who didn't sell it. It says closed to all applicants, first come, first served, until someone already on the plan moves, dies, or otherwise leaves, opening one seat. If a plan's enrollment already exceeds its accepted cap before AEP even starts on October 15, the guidance is explicit that it can't accept new AEP applications for a January 1, 2027 start, though it must still honor other election periods tied to a 2026 effective date through December 1, 2026.⁵

The only individual facing document either guidance references is an updated Model Notice of Denial of Enrollment, the letter a beneficiary gets after applying to a capped plan and being turned away.⁵ There's no waiting list display, no "plan is full" badge, nothing an agent can check before submitting an application on a client's behalf. You find out the same way the client would: after the attempt fails.

A worked example: what this looks like for one client

Say a Humana client in Broward County gets a non-renewal letter in early October. Her agent identifies what looks like the obvious replacement, a similar HMO from a different carrier in the same network, and plans to submit it once AEP opens on October 15. If that replacement plan requested and received a CY2027 enrollment cap set at its June 1, 2026 membership, and enough of AEP's first two weeks of applications already filled that gap, the application submitted on October 15 can be denied for no reason related to the client at all.⁵ Nothing published anywhere would have told the agent that risk existed before submitting it.

The fix isn't a tool that doesn't exist yet. It's sequencing: submit the highest priority replacement application as early in AEP as possible, have a second and third option already identified before the first one is denied, and don't tell the client "you're all set" until the confirmation actually comes back from the carrier.

Florida's Medicare Advantage market, by the numbers

Florida carries more exposure to this than almost any other state, just not for the reason some agents assume. CMS's own September 2026 state and county enrollment file shows California with more total Medicare Advantage enrollees than Florida, 3,751,199 against 3,076,234, which makes California the largest MA market in the country by raw enrollment.³ Florida is second, and it beats California on the measure that matters more for an agency: penetration. 3,076,234 of Florida's 5,362,591 Medicare eligible residents are enrolled in a Medicare Advantage plan, a 57.36% penetration rate, against 51.49% in California.³ More of the Medicare population in Florida is already on an MA plan than in almost any other state, which means more of it sits inside whatever a carrier decides to do with its book.

CMS's state by state landscape fact sheet for plan year 2027 shows Florida's own market tightening at the same time national numbers loosen. Florida will have 560 Medicare Advantage plans available in 2027, down from 611 in 2026, and the average MA premium in the state is projected to rise from $1.83 a month to $2.13.² Every Florida Medicare beneficiary still has access to at least one $0 premium MA plan, per the same fact sheet,² so this isn't a story about beneficiaries being priced out entirely. It's a story about fewer doors to choose from, at the same time one of the largest carriers in the state is pulling plans out of the mix.

Zoom out and the national picture looks calm. CMS projects the weighted average MA premium nationally will fall from $14.37 a month in 2026 to $12.00 in 2027, a 16.5% decline, with the total number of available plans holding roughly steady at about 5,532, down only slightly from 5,553.¹ CMS also projects 2027 MA enrollment at 34 million people, around 47.4% of everyone on Medicare, with more than 99% of beneficiaries having access to at least one MA plan and 97% having access to 10 or more.¹ Nothing in that headline suggests a market under stress. The stress is concentrated, in specific states, specific carriers, and specific plans, which is exactly why a national news cycle about Medicare Advantage in general does your Florida clients no good at all.

Why Humana is the one sending the letters

Humana is doing this for a stated financial reason, not a secret one. In its Q2 2026 prepared management remarks, filed with the SEC on July 29, 2026, Humana said it expects "our approach to 2027 individual MA bids will drive meaningful progress against our commitment of returning to a sustainable pre-tax margin of at least 3% in 2028."¹⁰ On the earnings call the same day, CFO Celeste Mellet told analysts the company expects its 2027 plan exits to affect approximately 600,000 members, and that Humana expects to recapture "just over 40%" of those members onto a different Humana plan, the same share it recaptured after similar exits in 2025.⁷

Run that math yourself, using only Humana's own two numbers, and something like 360,000 people nationally are the ones Humana itself doesn't expect to keep. That's the group actually shopping the open market this AEP, comparing carriers for the first time in years, searching for answers about a letter they didn't ask for. Some meaningful share of them live in Florida, given how much of the country's Medicare Advantage population already does.

What waiting costs an agency this AEP

The cost of sitting this one out isn't abstract. It's a compressed calendar, a client base that's about to start Googling, and a market where the plan they land on could refuse them for a reason no directory will show you in advance.

Start with the calendar. Non-renewal notices are landing in the first days of October, per the 90 day federal notice rule,⁶⁹ and AEP opens October 15 and closes December 7. That's a window of a few weeks where a large share of your displaced Humana book, plus whoever they tell, is actively deciding who to call. Nothing about the enrollment cap rule changes when AEP starts or ends. It only changes whether the plan someone picks is still accepting anyone by the time they pick it.

Then add the search behavior. A client who gets a letter saying their plan won't exist in 2027 doesn't call their agent first anymore, not reliably. A large and growing share asks a search engine or an AI assistant why this happened and what to do, in almost the exact wording of the letter. If your site doesn't have a clear, sourced, dated answer to that question sitting on it before the letters go out, you're not in the running for that search, however good your relationship with the client already is. Someone else's page answers it instead, and some of those readers never call you at all.

What to do about it this week

You can handle the client side of this yourself, today, without any of it costing you a subscription. Here's the order that actually matters.

1. Pull every Florida client on a Humana Medicare Advantage plan

Run the list from your CRM or AMS, not from memory. Note which plan they're on, since Humana hasn't exited every plan, only specific ones, and your client may not be affected at all.

2. Reach out before the letter creates the panic

A call or text that says "I saw Humana is adjusting some 2027 plans, let's check yours before you get a letter you don't need to worry about" beats a reactive callback every time. It also keeps you, not a search result, as the first source of truth for that client.

3. Re-quote early, and don't assume the obvious replacement is still open

Because caps are enforced without public notice, the "closest alternative" plan you'd normally suggest first could already be near its limit. Submit real applications early in AEP rather than waiting, since the guidance processes requests in the order received, and a plan that's still open in week one of AEP may not be in week four.⁵

4. Publish what you know, sourced, before you're asked

Write a plain page explaining the non-renewal, cite the actual CMS rule rather than a vague reassurance, and date it. This is the step most agencies skip, and it's the one an AI answer engine or a search result actually rewards.

All four of those steps are things you can do with a spreadsheet and a phone. Whether you want to build and maintain the content piece yourself, on a deadline, every time a rule like this changes, is a separate question.

How we build a site that's ready before the letters land

We build insurance agency websites on Astro, which ships zero JavaScript by default, so pages arrive as finished HTML, served from the edge on Cloudflare. Our builds score 90+ on Lighthouse performance on mobile and desktop, and some score 100, which matters here specifically because a confused client searching on a phone the week their plan letter arrives won't wait six seconds for your page to load. This page is built the same way. We measured its production build with Lighthouse on September 30, 2026, before publishing: a median of 98 on mobile across five runs, individual runs between 96 and 99, and 97 on desktop.

Every page we ship carries the schema stack that lets an AI answer engine actually read and quote it: Article, FAQ, breadcrumb, and organization markup that matches what's on the page, not boilerplate. The crawler files, llms.txt, sitemap.xml, and robots.txt, regenerate themselves the moment a new page publishes, with the full absolute URL, and robots.txt is written to welcome every major crawler including the named AI crawlers, so a page like this one is actually reachable by the systems summarizing it for your clients.

That's also where publishing speed stops being a nice to have. Per the live pricing page, verified today, Digital Foundation's Pro tier adds a new blog post, location page and answer page every week, and Scale adds one of each every business day.¹¹ When a rule like the CY2027 enrollment cap starts producing specific, named plan closures county by county over the next few weeks, that's exactly the cadence that gets a sourced, updated answer live the same week it's news, instead of next quarter.

We also build real quote and comparison tools wired to live federal data, not a static form with example numbers from last year's rates. That's the difference between a page that reassures a client in the abstract and one that actually shows them a current option.

What changes for your agency

The practical result of being first with a sourced, fast, AI-citable answer is that you stop competing for a client's attention only on the phone, and start competing for it on the page they land on before they ever pick up the phone.

None of this replaces the phone call, the re-quote, or the relationship you already have with a client on a non-renewed plan. It's what happens for everyone who searches before they call, and for the prospects who were never your client to begin with but are shopping the market this AEP anyway.

When this doesn't change much for you

If your book is mostly life, final expense, or annuities, or you write Medicare Advantage outside Florida with only a handful of Humana clients, this AEP probably looks close to ordinary. A single sourced page on the topic is still worth having, since a prospect researching the news will find it regardless of your book's makeup, but you don't need a publishing cadence built around it. Save the urgency for the agencies actually sitting on a concentrated Humana book in a high penetration state, because that's where the calendar and the cap rule actually collide.

Questions agents ask about the 2027 enrollment caps