agilon's risks all
reduce to the same
thing: did the
actuaries get it right?
✦ The bottom line
agilon's risks are medical-cost trend (the assumption that determines whether contracts work), member retention (after shrinking from 605K to 536K, the next leg needs growth), regulatory (Medicare Advantage is a political target), and execution (a new CEO mid-turnaround).
↓ the brief below
From the 10-K · the subsidiary-loss warning
If unfunded, such losses have in the past, and could in the future, result in substantial doubt related to such subsidiary's ability to continue operating as a going concern, and the contractual and regulatory consequences of such failure could have a material adverse effect on our business, financial condition, cash flows, and results of operations.
↳ This is unusual disclosure: agilon is telling you its subsidiaries — the risk-bearing entities that hold individual payor contracts — have triggered going-concern flags before and could again if cost trends spike. The parent company stays solvent, but the contracts can break.
What 'medical cost trend' means and why it's the whole game
Each year, agilon signs contracts that say: 'I'll take care of this member for X dollars per month next year.' That X is set based on what care cost last year, plus some assumption for inflation (the cost trend). If actual cost trend comes in higher than the assumption — because more members got sick, or drug prices rose, or hospital prices rose — agilon eats the gap. In 2024, trend ran about 7-9% against contracts priced for 3-5%. That's how an apparently small percentage delta turns into hundreds of millions of dollars of losses.
Wall Street calls this
Medical cost trend / MLR risk
Q1 2026 is reserved at 7.4% MA cost trend — much higher than pre-2024 assumptions. If actual trend exceeds 7.4%, the medical margin shrinks back.
The concentration · membership trajectory
-11
% YoY
Total members declined from 605,000 to 536,000 (-11% YoY). The turnaround needs the shrink to end and growth to resume — or the per-member improvements run out of room.
We use algorithms, AI, and machine learning solutions in, and we may in the future integrate additional algorithms, AI, and/or machine learning solutions into, our platform, offerings, products and services, and these applications may become more important in our operations over time.
↳ A lot of agilon's underwriting and cost-management depends on algorithmic models. If those models are wrong — or if the data they're trained on doesn't represent next year's reality — the same kind of underwriting miss could recur.