Open in interactive viewer → charts, metric popovers & call review

Auna's Predictability Problem: The Risk-Sharing Contract That Broke the Risk-Sharing Story

Mexico finally inflects and cash conversion hits a record — but a 500,000-life oncology contract in Colombia delivered the third 'extraordinary' surprise in three quarters, and management's own words say it knows.
AUNA · Earnings Call · 2026-11-18

Three hits, one quarter, and a word nobody used last year

Auna's Q3 2026 was not a bad quarter in the ordinary sense. Revenue grew 10% FX-neutral, free cash flow nearly doubled year-to-date to PEN 626 million, the cash conversion cycle went from negative 9 days to negative 15, and leverage ticked down sequentially to 3.5x against a 3x medium-term target. Underneath the noise, this is a business generating cash better than it ever has. And yet the quarter arrived carrying three separate "extraordinary" items: an oncology risk-sharing contract in Colombia whose medical costs blew through contractual assumptions, regulatory price caps on high-cost oncology medicines enacted mid-August, and PEN 18 million of revenue adjustments from prior-years' billing reconciliation in Peru. Gisele Ferrero put a ruler on the Colombian damage: of the roughly $4 million year-over-year drop in Colombia's EBITDA, she estimated “approximately 70% of that was due specifically to this oncology risk-sharing contract, which has already been terminated. And the remaining 30% is due to the price caps” — Gisele Ferrero, Chief Financial Officer or CFO · 2026-11-18. The new vocabulary in this transcript is telling. oncology risk sharing contract did not exist as a phrase in Auna's prior three calls. Neither did contractual assumptions, nor particular contract. The company spent 2026 up to now selling risk-sharing as the mechanism of safety; this quarter it discovered the mechanism is only as safe as the actuarial guess inside it.

The paradox at the center of the thesis

Go back to May 2026. Asked about Colombia's regulatory risk, Jesús Zamora was emphatic: “I've seen -- we've seen the worst of regulatory changes and political upheaval with President Petro. And I think in both scenarios of the current leading polls and candidates to the presidency, I think the Colombian health care sector will not get worse than we've seen in 2025.” — Jesús Zamora Leon, Chief Executive Officer · 2026-05-20 On the same call, the risk sharing contracts book was pitched on its stability: “Our risk sharing business have high predictability, cash predictability and lower working capital cycle” — Jesús Zamora Leon, Chief Executive Officer · 2026-05-20, with only a small, temporary margin dilution at onset. Both statements now read badly. The worst was not behind Colombia — a single newer PGP contract covering roughly 500,000 lives absorbed an unusually high incidence of advanced-stage cancer cases and cost far more than modeled, and price caps landed on top of it. What was sold as the predictability layer of the model is precisely what produced the least predictable quarter of the year. Zamora also insisted it was idiosyncratic: “we do not think what has occurred with one of our PGP contracts, a newer contract, has any contagion effect or any replicability in our rest of our PGPs and PIPs contract. So we don't think it's a sector issue.” — Jesús Zamora Leon, Chief Executive Officer or CEO · 2026-11-18 Perhaps. But investors are being asked to take that on faith one quarter after being told Colombia had bottomed — and the CEO's own language has changed from confidence to caution, with Colombia now expected to "gradually recover" rather than lead.

it's important that you also take note that we're working on the qualitative nature of predictability. We cannot be about these exogenous events. We need to make sure that we make sure that, first of all, we have full control of what we're doing in the services that we're delivering.

Jesús Zamora Leon, Chief Executive Officer or CEO · 2026-11-18
That is a remarkable thing for a CEO to say out loud. qualitative nature of predictability as an explicit management objective is an admission that the sell-side case has been built on a property the company has not reliably delivered. He called them “dirty quarters with a lot of extraordinary events” — Jesús Zamora Leon, Chief Executive Officer or CEO · 2026-11-18. Three in a row is not an accident; it is a control environment.

Peru: a cleanup that keeps moving to the right

The billing matters saga is now a three-quarter arc with a slipping finish line. In August, Ferrero said prior-period reconciliations would be “finalized during this year” — Gisele Ferrero, Chief Financial Officer · 2026-08-19. In November, the timeline became: “We do expect to be able to reflect the effects related to previous periods within the next coming quarters and then be able to normalize those effects towards the second half of 2027.” — Gisele Ferrero, Chief Financial Officer or CFO · 2026-11-18 That is a year of slippage, and it matters because Peru is the mature, "clockwork" asset the equity story leans on for earnings quality. The offset is real, though underappreciated: PEN 614 million of cash, an 83% increase year-to-date, and CapEx running at 3% of revenue versus 4% guidance. Peru's hospital network is running near 90% utilization, which the CEO frames as comfortable — “we can actually just marginally build asset-light facilities that can deliver services for hundreds of patients and family members a day without huge investments” — Jesús Zamora Leon, Chief Executive Officer or CEO · 2026-11-18. The Centro Ambulatorio Trecca project, fully funded via EsSalud progress certificates, is on schedule with the first construction disbursement made.

Mexico, finally, is the good news

Strip out Colombia and Peru and the quarter reads differently. Mexico posted its first year-over-year Adjusted EBITDA growth of 2026, +5% at segment level, with the Monterrey network up 10%. Segment margin was 26%, with Monterrey at 28% on 41% occupancy — a combination that implies substantial future leverage as utilization climbs. Surgery volumes rose 12% and chemo/radiotherapy sessions 34%. It is the clearest evidence yet that the high complexity care build-out and the renegotiated ISSSTELEON contract are converting into economics, even against a Mexican insurance market that shrank by roughly a million policies under the VAT deductibility change.

Two risks that are not in the guidance

First, El Niño — a genuine global theme this year, and one Auna is treating as an operational event rather than a footnote: “the market has a forecast for a strong to extremely strong El Niño for the end of this year, beginning of next year, and weather can be highly unpredictable” — Gisele Ferrero, Chief Financial Officer or CFO · 2026-11-18, with backup fuel, water, generators and patient transfer flows pre-arranged across Piura, Chiclayo, Trujillo, Arequipa and Lima. Weather-driven continuity costs are not in a "broadly stable to slightly down" EBITDA guide. Second, Sojitz. The Mexican co-investment MOU was announced in late 2025 amid talk of a $500 million five-year pipeline; a year later the answer is “we should have something at the beginning of the year, perhaps to announce” — Jesús Zamora Leon, Chief Executive Officer or CEO · 2026-11-18. Meanwhile the EBITDA margin outlook has softened from reaffirmed guidance to "broadly stable or slightly down," and the buyback question remains an open board discussion rather than a plan.

The trade

Auna is a sub-$400 million market cap healthcare operator whose stock offers no visible tape context here, so the fundamental read is all we have. What changed this quarter is not the growth rate — it is the credibility of the framework. A company that generates 95% free cash flow growth while repeatedly reporting "extraordinary" items has a valuation problem that cash flow alone will not fix. The path to repair runs through Mexico's margin ramp, a 3x leverage target, and — most importantly — a quarter without a new word in the transcript. Watch whether volume growth in Monterrey converts to margin before Colombia's price caps fully annualize in Q4. That is the sequence that decides whether this was a bad quarter or a bad model.