Six Flags: The Portfolio Is No Longer the Portfolio

Six Flags: The Portfolio Is No Longer the Portfolio

There are quarters where the numbers are the story, and there are quarters where the numbers are a hostage. Six Flags' third quarter belongs to the second kind. On paper, the report is a decent one: same-park net revenue of $1.17 billion, up 1.4%; same-park adjusted EBITDA of $493.2 million, up 1.1%; a 42% peak-season margin held flat against a worse July 4 calendar. Yet the company also booked a $688 million non-cash impairment on goodwill, trade names and long-lived assets, producing a $366 million net loss, and guided fourth-quarter EBITDA down to roughly $140–150 million from $166.9 million a year ago.

The stock was already down 67% in the ninety trading days before the call. Six Flags now carries a market capitalization of about $945 million against a business that generated $806 million of trailing-twelve-month adjusted EBITDA. That gap is not a valuation opinion. It is the market pricing a tail risk.

Two weeks in October

The tail risk has a name and a date. On October 12, a guest named Hassan Karim died at Six Flags Over Georgia; others were injured. Days later the company announced it was retiring the X2 coaster at Magic Mountain. John Reilly, eleven months into the CEO job, opened the call with the accident rather than the quarter — a sequencing choice that tells you where management's attention actually is.

Safety is the most fundamental commitment we make... Confidence is not restored with a statement. It is earned through action.

The operating consequences show up in the fourth quarter. October is 60% to 65% of Q4 revenue, and attendance fell 9% year-over-year. Ashok Walia, the CFO, split the cause roughly in half: weather (including hurricane Isasis) on several of the most important early-October weekends, and the other half from the Georgia incident and the ride closures that followed it. He was careful to frame the shortfall as temporary — “the difference only represents about 3% of our trailing 12 month adjusted EBITDA” — Ashok Walia, Chief Financial Officer (CFO) · 2026-11-05 — and noted that in the last week of October, excluding directly affected parks, the company returned to growth.

One detail deserves more attention than it received. Q3 insurance costs were already elevated because of claims experience and an actuarial estimate of incurred-but-not-reported claims, which Walia stressed was "all related to the period before the incident in October." In other words, the self-insured liability trend was worsening before anyone died. The new vocabulary on this call — claim, coverage, retention levels, $100 million of general liability cover above the retention — is not boilerplate. It is the sound of a leisure company discovering it is also an insurance underwriter.

The turnaround, before the noise

Strip out October, and the operating story is genuinely better than the stock. Through nine months, same-park adjusted EBITDA is up about 11% to $639.5 million and margin has expanded roughly 200 basis points to 28.3%. Free cash flow of $257 million improved by about $300 million year-over-year, helped by capex falling to $241 million from $408 million. Cost savings tracked inflation dollar-for-dollar in Q3 — $13 million of savings against $13 million of pressure — aided by 54 extra operating days.

The membership engine is also working. The active membership base was up about 7% at September 27, pass holders and members were 61% of attendance, and cross-park visitation is running about five times last year's level as the Gold regional pass lets guests pair parks like Knott's Berry Farm and Magic Mountain in one market. Ride uptime rose to 90.3% from 89.4%. These are the proof points of a turnaround, not a rescue.

The reversal

Here is where the call gets interesting. In May, an analyst asked whether more parks might be sold. Reilly's answer was a flat no:

We have no other plans in 2026. If you're buying a pass, if you're thinking about a pass, the portfolio is the portfolio, and we're focused on the summer and execution.

Five months later, with the season closed, that answer is gone. “We do believe there is opportunity for further portfolio optimization” — John T. Reilly, Chief Executive Officer (CEO) · 2026-11-05, Reilly said, adding that discussions are underway and that parks are being categorized by their role in the network and by their impact on the net leverage ratio. He declined to name a number of parks.

To be fair, the intent has been consistent — in November 2025 the company said “getting the portfolio smaller and more nimble is a priority” — Brian Witherow, Chief Financial Officer · 2025-11-07, and the February call sketched a framework of core versus non-core. What changed is timing and candor. The May freeze was a mid-season commitment to pass buyers; the November thaw is the post-season truth. The tension is that the seven parks sold in April raised cash but barely moved leverage, as Evercore-skeptic analysts pointed out. Reilly conceded the point: those sold at "a multiple that was similar to where the portfolio was trading." Selling at the portfolio multiple does not delever. Selling at a higher multiple does.

That is the real argument on this call, and it sits on top of a hard January obligation: a roughly $370 million payment to buy out the remaining two-thirds of Six Flags Over Georgia, which today absorbs $26 million a year in distributions. Buying it swaps those distributions for about $21 million of revolver interest — a net saving of roughly $5 million, funded with debt, at 5.8x leverage. The portfolio optimization language and the asset sale pipeline exist to make that arithmetic work.

What stopped being said

The keyword record is unusually loud this quarter. Out: Travis Kelce, the celebrity-investor storyline that dominated late 2025 and has vanished; demand generation, the phrase that anchored the spring turnaround narrative and appears nowhere in the current prepared remarks; and the granular underperforming park taxonomy that once drove whole calls. In: sufficient liquidity, revolver, mid 30s margin ambition, and a newly prominent growth CapEx debate.

That last one matters. Management cut capex hard and re-sequenced the pipeline by return on invested capital, then insisted this was "not a change in our growth strategy." Analysts did not buy it quietly. The honest read: capex is now the shock absorber for leverage, and the admission of "excess capacity in these parks" — “we are not bumping up on closures or anything like that” — John T. Reilly, Chief Executive Officer (CEO) · 2026-11-05 — is a polite way of saying the growth problem is demand, not capacity. Reilly also conceded October pass sales ran slower, and the 7% active-base gain is expected to flatten to "flat to slightly up" by New Year.

Meanwhile the valuation has done something remarkable. Price-to-revenue has fallen to 0.7x, down 27% year-over-year and 41% off its 2025 peak, for a business with 91% gross margins and a 42% peak-quarter EBITDA margin. Price to revenue: 0.7x The offset is a balance sheet carrying roughly $4.9 billion of effective net debt against a 5.8x leverage ratio. Effective net cash of negative $4.9 billion This is not a company that can simply wait out a bad October; it is a company whose equity is a call option on deleveraging.

The verdict

Six Flags has a credible operating turnaround, a management team that is at least willing to say the uncomfortable thing, and a liquidity position of $1.06 billion with no near-term maturities and 70% fixed-rate debt. What it does not have is time to be unlucky twice. The Georgia accident cost it a chunk of October, a chunk of Q4 guidance, and — more importantly — the benefit of the doubt. Reilly's long-term targets of mid-30s margins and ~4x leverage remain unchanged in wording; in February he told analysts “I do not have a timeline today to get to a specific target” — John Reilly, CEO · 2026-02-19, and that is still true nine months later. If the park sales arrive at good multiples, the equity re-rates violently. If they do not, the 67% drawdown is not a dislocation — it is a forecast.