Yesterday TF79 asked a great question about Venture Capital (VC) backed insurers in comments:
“The VC backed insurers that attempted to buy marketshare without a concern of profitability…” can you expand on this? I would have assumed that VC’s in general would be hyper-focused on profitability at the exclusion of all else. Are they taking a loss in the short-run in exchange for long-run market power or something?
There have been three major venture capital backed ACA insurers, Oscar, Bright and Friday. Oscar is still alive at a stock price 85% below their IPO but with decent cash reserves. Bright and Friday were shut down by regulators over the past year as the state regulators were terrified at the capital reserve to potential obligation ratio.
So what do I mean “buy marketshare without a concern of profitability…”
Let’s assume that the insurers are competent-ish. This is actually a fairly large assumption as the alternative story is that these three insurers lit several billion dollars on fire to pay for learning by doing (and at that, slow learning by doing….)
We know that the enrollees who are flipping a coin between buying insurance and not buying ACA insurance are buying almost entirely on price. We know that these enrollees are relatively healthy/low cost compared to the rest of the ACA pool. We know that these enrollees likely generate a big risk adjustment payable obligation as they don’t use many services. We also know that most enrollees are like this. We know that insurers that barely miss being the least expensive plans in a metal level don’t get a lot of enrollment. We have a winner take most market.
If an insurer is mainly competing for this market segment, they want to drive down premium as far as possible below the benchmark silver plan. Assuming an insurer is constrained to price in a way that makes them break even or better, the three big options are to reduce the benefit value (low acturial value/high expected cost-sharing plans) and/or get great provider pricing (usually by having a narrow network) and/or reduce utilization (either lots of prior authorization and gatekeeping or value based arrangements to reduce dumb spending). An insurer can do all three things. Most try to do all three things.
However if we relax the constraint that the insurer needs to make a profit in a given year, we can tweak the parameters. For the same premium (partially paid for by capital reserves or VC/IPO cash), the insurer can offer better benefits, or a bigger network or less restrictions. Or for the same network, benefits and network, the insurer can take a bit off the premium. Winning the bottom price slots is really valuable in terms of head count.
So why would you want head count that you’re losing money on in the first year?
There are three legitimate business case reasons why a VC backed insurer might want to lose money in the first year to get a lot of new membership.
Provider pricing is a chicken and an egg problem. Good deals happen because the insurer can move lots of members to a particular hospital that gives a good price. Lots of members are needed for good deals, but good deals with good prices are needed for lots of members. Losing money on membership to get good deals could set up an insurer for long run success.
The second is if you think that the enrollees are “sticky.” Year one might be a loss leader, but years 2,3,4,5… could be profitable as the insurer might be able to raise rates to cover costs without losing likely to be profitable membership because members have high inertia unless they get hit with an attention 2×4 to the side of the head. Increasing prices and using automatic re-enrollment rules to move last year’s enrollees to higher premium plans for this year while offering new, lowest cost, subsidized by capital losses plans to attract new membership might lead to profitability.
The big assumption is that the ACA market will have a lot of repeat customers AND the repeat customers aren’t looking around for the best price. The second assumption is mostly true, but the median enrollee in the ACA lasts about a year in the market. So this is iffy.
The final reason why a tech focused, VC backed insurer might want a lot of money losing membership is that these insurers have a strong belief that their technology and information parsing/meaning creation software is a comparative, competitive advantage. However for machine learning/AI/lots of logistic regression to actually be useful and better than whatever United HealthCare or Aetna or the local Blue is doing (as they are doing a lot of data mining!) there needs to be enough training data to produce useful predictions that produce a lot of 2% and 3% comparative efficiency wins. Buying membership and losing money on that membership can be justified as a means of making the AI/Machine Learning/lots of logistic regression better.
The challenge here is scale. One of the three VC backed insurers might, just might, have a million covered lives in a year and might, just might have ten million person-years of data that went through their claims system. These data might be in 10 or 20 states. This is not small data. It is also about the size of the data sets that I played with when I worked at UPMC Health Plan‘s Medicaid division for a single state (actually for about half of a single state). A big Blue or United or Aetna etc. can drown these data sets in both breadth and depth.
If a VC backed insurer thinks that its machine learning is its secret sauce, spending a lot of money to get both patient reported inputs and outcomes as well as lots of proprietary claims to feed the machine learning would be justifiable.
So those are the obvious business cases where a VC backed tech focused insurer could justify intentionally losing money to get customers.
The other explanation is that some of these insurers lit money on fire because they were grossly incompetent and did not realize that not every incumbent in the insurance industry was grossly incompetent and a bunch of MBAs coming in cold with great press seeking abilities might not know everything….
Take your pick.
Ken
There’s a fourth possibility, which is that the VCs can generate a lot of hype, then cash out by selling their stake to whatever greater fools are attracted. This has been happening a lot with crypto startups. Is there any evidence Oscar/Bright/Friday might have been trying this? For example, the “AI data mining” claims might be part of the hype.
Barbara
If you are in a tech field and you look at health care spending it might seem like low hanging fruit, until you actually look more closely and realize that it’s the kind of low hanging fruit that is enormously resistant to being picked by machine.
I don’t know how many programmers are employed by the big insurers, but I do know that more than 20 years ago I represented one of the Goliaths and they had an entire programming team based in Ireland who did nothing but build analytical software to make sense of data. And in health care, even when you understand what’s going on, that doesn’t mean you can do — or make other people do — whatever you want based on your findings on what might be optimal.
And even if there might be some secret sauce to be developed by these unicorn magicians, how do you develop that sauce by enrolling the healthiest people in the population? The ones who don’t generate reams of useful claims data?
We pay too much and we have poorer outcomes, but the most significant reasons for that state of affairs won’t be solved by technology.
Barbara
@Ken: Yes, certainly, and this was very obviously the case with many MA start-ups after 2008. Scale up fast and cash out. But no one is going to buy you until you can show some kind of successful business vision somewhere. As it is now, enrollment is going to competing plans without any of them having to pay for it.
Fraud Guy
I’ll take the latter option.
Barbara
@Fraud Guy: They have a lot in common with crypto currency mavens, except that they cannot avoid financial regulation — the assumption that “old guard” in the industry — banks, insurers — can be bypassed by magic beans. There is room for innovation, for attacking current structures and arrangements, but there is no silver bullet that only visionary outsiders can see.
narya
The only way I could see it working would be ALSO cherry-picking a medical condition that would bring a lot of adjustments to them, to offset. Get really focused on one thing, do it really well. That said, I can’t think of many conditions that would work this way; hypertension and diabetes tend to bring a host of other issues, for example.
dirge
There’s another story line here: hubris about reforming an inefficient industry.
Having written software for many startups in a lot of industries (finance, logistics, marketing, etc…), I’ve seen a ton of variations on the theme.
If you’re smart, you quickly realize that building software in an industry also makes a map of all the players, their relationships, and decision making processes. You’ll observe, correctly, that the tangle of petty fiefdoms, rife with uninformed decision makers, duplicative efforts, and ill-served stakeholders, is wildly inefficient. If you’re a little smarter, you’ll figure out that if you can get enough people using your applications, then you can evolve them to adjust those relationships and processes, presumably for the better. If you’re not too smart, this looks golden; everybody will want to work with you, or buy your stuff, because you can massively reduce errors, delays, and administrative costs.
Now this can work, and when it does the payoff is huge, but…
Pretty soon, you’re going going to discover just how fiercely those petty fiefdoms resist change. It’ll dawn on you that because you’re trying to shift relationships and processes, each change requires three or four interest groups to simultaneously reach consensus on adoption (and get budget from their respective finance departments), which is seriously hard even if they all want to. But every proposal you make is a threat to someone, somewhere, and one well placed and determined enemy can burn down months of negotiations and man-decades of development.
That IT director from the obscure medical records company, who seemed helpful but you weren’t sure who’d invited him to the meeting? Yeah, it turns out everyone trusts him or needs him, he talked to everyone after, and now they won’t call you back.
At some point, you capitulate, and adjust to the same client and vendor relationships as everyone else. It follows that you’ll need a lot of the same internal roles and organization that goes with that. Having been beaten into the shape the industry expects you to be, you’re now just another insurer, with a massive overinvestment in incrementally improved tech that does mostly the same thing as everyone else’s.
I interviewed with Oscar something like ten years ago, and left with the distinct impression that they were headed down that road.
Anonymous At Work
@Fraud Guy: Tell me when/where another VC group is getting into the game, so I bring the marshmallows, graham crackers and chocolate bars.
TF79
Thanks! To the extent VC firms (and perhaps others) are making some sort of dynamic profit optimization, does that suggest the Risk Adjustment process could/should be more dynamic instead of true-ing up annually? (Of course, if the reason is that these guys just don’t have a clue, that’s no help.)
Barbara
@dirge: Yes, I totally get where you are coming from, but the difference in the cost of health care in America versus the rest of the world is mostly attributable to high unit costs of providing health care services and the explosion of chronic conditions that can be materially attributed to low social and safety net spending.
As for streamlining — there is a lot of friction in the system and eliminating that would result in savings as well as a fairer system, but that again is mostly a failure of political will not technology.
I have never understood how Oscar (or anyone else) is going to wring enough efficiencies out of administrative processes through the application of technology to offset costs and friction inherent in what we have now. Not that technology couldn’t be useful — for instance in helping doctors to get to more accurate diagnoses. Things like that just aren’t going to materially move the needle on cost.
Another Scott
@dirge: Thanks for the story.
I’m reminded that a hard part of coming up with something really new is coming up with something really new. Electronic television. A computer hypervisor. Heavier than air flight. A hot air popcorn popper. Intermittent windshield wipers.
But coming up with a new thing isn’t enough. Humans are great at copying things. Once someone shows that something is possible, someone else will find a way to figure out how it works, copy it, make it better / stronger / cheaper / faster.
A lot of the behemoths stay big because of the patent and copyright laws in the US, not because they’re better at inventing or faster at innovating. The government-sanctioned control keeps competition down.
Just having a good idea is never enough. Fast followers often win, especially when trillions of dollars are at stake, and even-moreso when Uncle Sam isn’t protecting your business processes with iron-clad patents and trademarks.
Thanks.
Cheers,
Scott.
dirge
Well, I suppose the thread is long dead, but to clarify, I did not mean to say, “our brilliant technology could have fixed everything, but for those meddling Luddites!” My criticism is mostly aimed at a certain common type of tech industry leadership, who somehow remain blissfully unaware of political constraints, institutional inertia, and human nature. Technology can help you with those problems, sometimes in important ways, but it’s just a tool. I’d guess we probably mostly agree on the political issues here.
I do, however, think you underestimate the amount of friction in the system that’s not directly attributable to the political and regulatory environment, and which could, theoretically, be ameliorated by changes within the industry. Political action would make it all easier, but with or without that, the opportunities for improvement are ubiquitous and huge. Just really hard to execute on.
not a luddite
@dirge: Appreciate your insightful comment. Do you have any interest in working or consulting with a profitable health-tech startup trying to avoid those very fiefdoms? How can I get in touch if so?