Very few insurance companies want to hold onto all of the risk that they agreed to take on. There are two ways around this problem.
The first is to commit fraud on a massive scale and laugh at the people who are filing two million dollar claims when their policies state there is neither a lifetime nor an event limit. That tends to lead to arrests of senior executives who were on the hookers and blow bonus level and the firing of the low level drones who were on the free pizza on fifth Fridays bonus level. From a self-interested perspective, this is a bad thing for the insurance company.
The other means of dealing with tail risk is to buy protection from another party. This is reinsurance. And it is quite common as a means of risk and cash flow control. Reinsurance is a contract between two parties.One party has a significant risk of all health care costs for 10,000 individual. It buys protection for the fourth or fifth standard deviation and greater level of risk to another company. Spreading the risk over the nation or often internationally allows for smaller variance costs. There are a few different ways risk can be sold and we’ll work through examples.
Cystic Fibrosis is an extremely expensive condition to treat and it is not randomly distributed as it is a genetic disorder. I’ve used this family before because I don’t have privacy concerns and it providers good numbers:
Affecting one in every 3,900 births in the U.S., CF is one of the most common genetic disorders known. Yet it afflicts too few people — just 30,000 in America…
Laura and Cate’s daily regimen of two pills of Kalydeco costs $841 per day; that’s $307,000 each year, making it one of the world’s most expensive drugs. In most cases, private insurance picks up the bill. Medicare or Medicaid may pay as well.
A reinsurance contract might be structured so that if a health plan has more than 3 CF members per 10,000 total members, the reinsurance company will pay half of the drug costs above $100,000. The health plan would have the baseline risk, but the reinsurance company would take a good chunk of the tail risk of having a family of seven CF patients sign up for insurance from a single provider. This same type of contract could apply to numerous other diagnosises and diseases.
Another common type of reinsurance is a contract where the original insurer pays the first half million dollars of claims in a year and then the reinsurer pays half of the next half million and everything above one million. Million dollar annual claim histories are very unusual events. A single insurer with a risk pool of ten thousand individuals can handle a normal load of million dollar or more claims becauase the total number of members with that claim history is miniscule. However, if a single insurer covered an electrician’s union that responded to a hypothetical Fukishima type disaester, that insurer would be wiped out even if the national level of million dollar claims barely moved.
Reinsurance most likely came into play for the Boston Marathon bombing victims as trauma care is extremely expensive and an isolated event concentrated costs into a very shallow pool.
The other type of reinsurance is between companies that self-insure (pay all the costs of employee medical care). In this scenario, the self-insured company uses the insurance company as a back-end administration and negoatiation unit. The health insurance company processes claims, handles member complaints and grievances and sells access to its network of providers (as well as the relevant discount) for a fixed cost per member per month. The self-insured company will often buy a form of insurance called “stop-loss” that is effectively reinsurance where the stop-loss seller will pick up the very high dollar claims.
Reinsurance is the black and tan of the insuranc industry.
RSA
Interesting! I didn’t know about this practice, but it makes a lot of sense. Checking online, I find that this is how the airline industry works, too (if what I’m reading is reliable). I’ve wondered in the past how a relatively small start-up airline company might survive, with any single plane bearing a huge potential cost if it crashes. So, reinsurance.
Gin & Tonic
@RSA: Indeed a very good write-up by Richard. But sort of in response to your query, reinsurance is an integral part of the property-casualty insurance business as well. Richard talks about it from a health viewpoint only, but think of insurance as really three (or more) separate industries: you have health, life and property-casualty (i.e. your typical auto or home insurance or the larger-scale commercial property/transport, etc.) Reinsurance is typically not necessary or required in life, because the risks are so well-understood and homogeneous, but it is necessary in property for events like your posited airliner crashes, vey large industrial accidents, widespread storm events, etc.
Big R
I don’t get the metaphor. Meta for what?
Richard Mayhew
@Big R: Mixing two types of insurance companies to produce something tasty
RSA
@Gin & Tonic: Thanks for the wider context, G&T; that’s a useful breakdown of an area I don’t know very much about.
Linda Featheringill
Black and Tan?
Historically, the British have one view of the B&T and the Irish have an entirely different view. Wonder which one is meant by the author?
raven
Isn’t Loyd’s of London the main insurer of everything?
Gin & Tonic
@raven: Lloyd’s isn’t actually an insurance company, it’s an exchange, where specified, limited (but large) risks can be insured on a sort of ad-hoc basis by pools of underwriters (investors) set up on pretty much a one-at-a-time basis. The big reinsurance companies tend to be based either in Switzerland or Bermuda.
raven
@Gin & Tonic: Thanks
Keith P.
I used to work on a reinsurance application for my first post-college programming job (it was written in Smalltalk for some reason). It was for reinsurance for natural disasters IIRC….SICS/nt. It may even still be in use, but I can’t read the Norwegian wikipedia entry.
RSR
@Linda Featheringill: Indeed. I’ll take a half-and-half, though.
http://thebeerchicks.com/drink/165/please-don-t-order-a-black-n-tan
Pongo
@Gin & Tonic: My dad was an exec with a property-casualty re-insurer in the midwest. Their primary business was with regional insurers in rural areas who covered large agricultural properties. One massive storm could wipe out these smaller insurers or farmers who were self-insured (this was in the good old days prior to corporate farming taking over), so reinsurance was really critically important in this context.
Big R
@Richard Mayhew: Grazie, Signor.
jl
Thanks for the post and explanation of black and tan reference, which I also did not get. I need to pay more attention to the booze posts on this blog.
When I worked on a project on managed care regularion, I remember reading things in contracts and regulatory filings that suggested for some types of insurance plans, it’s reinsurance all the way down. I’m not sure those types of managed care organizations are still around: basically some offices and desks with Money Guys in suits who have built up a network. Individual practitioners and medical groups would have what seemed to be essentially re-insurance policies with the company that was offering to include them in an insurance network, and this was usually connected in some way with benchmarking. In other words, an individual doc has resource to a reinsurance policy for capitated covered lives, but there was a penalty for anything above a minimal amount or recourse to the reinsurance.
This always seemed to me to be troublesome wrt to maintaining reasonable standards of care for populations that have potentially very expensive chronic conditions, some of which are more common that cystic fibrosis, such as advance heart failure.
Does the RM know how common this practice still, is or has movement way from certain forms of capitation made it less common. And, is there evidence that it distorts clinical decision making.
Personally, I would rather have comparative effectiveness analysis (‘death panels’ up front and for professional and public inspection) and publicly accepted and proven standards of care and protocols (which would be more common with a uniform standard national policy). with a uniform standard benefit design (edit which is important, since copays and reinsurnace levels affect what care different patients of different means actually receive, and inefficiencies from dependencies in inadequate care actually received when different organizations have to pick up the risk created by a previous insurer’s deliver of care) It would be preferable for me to dealing, without my knowledge with a doc or a group which had been unlucky and went to the individual organizational level reinsurance pool already.
I mean, RM is talking about large numbers issues with large insurance companies. Think how it looks to a medical group or an individual doctor, or even a small hospital chain, or whatever.
I wonder how much this sort of worry, besides reimbursement arbitrage, has lead to the growth of private specialty hospitals that can select its risk pool and chances of outliers in cost by declaring its scope of practice. I think the growth of these specialty clinics and hospitals is a dubious thing, since there is evidence that they have lead to overcapacity in some affluent areas and undercapacity in less affluent areas. And from what I have read, they can fiddle with their scope of practice and case mix to get a 10 to 20 percent profit boost of reimbursement over cost.
burnspbesq
Reinsurance is the nectar of the Gods … as long as the reinsurer is solvent.
And the vast majority of them are. They’re companies that almost nobody has heard of, that have hundreds of billions of dollars of assets and are organized in countries that know how to regulate the insurance industry.
I get a little hinky about reinsurance companies that are set up in places like the BVI, where regulation is a bit, shall we say, less robust.
LanceThruster
It sounds like people dealing with the industry need some enforcers with ‘muscle’ (if youse gits my drift).
johnny aquitard
Isn’t this reinsurance thing that like credit default swaps? It sounds similar.
Richard Mayhew
@johnny aquitard: not really — the big reinsurers are heavily liquid and have massive assets (Swiss RE for instance has $200 billion in available assets for an oh-shit scenario)
And they have history on their side. They have a decent idea of how many cancers will occur in a year, how many bombings may occur in a year, how many multi-billion dollar weather disaesters occur etc. They are not modeling off of Excel spreadsheets running MBA approved copulas that can’t pass basic muster… they’re working with 100+ year risk profiles.
The Pale Scot
Come out you Black and Tans
..Come tell us how you slew
Those brave Arabs two by two
Like the Zulus they had spears and bows and arrows,
How you bravely slew each one
With your sixteen pounder gun
And you frightened them poor natives to their marrow.
..Oh, come out you black and tans,
Come out and fight me like a man
Show your wives how you won medals down in Flanders
Tell them how the IRA made you run like hell away,
From the green and lovely lanes in Killashandra.
Oh..OH.. you mean the beer
PST
@Richard Mayhew: the big reinsurers are heavily liquid and have massive assets (Swiss RE for instance has $200 billion in available assets for an oh-shit scenario).
This is a little off the topic, but of interest, I think. I do legal work for some of the giant property-casualty reinsurers. If you ask people employed there if they believe in global warming caused by rising CO2 levels, they roll their eyes and look at you with pity for asking such a stupid question. People who place hundred-billion-dollar bets on the frequency and severity of future natural disasters can’t afford the luxury of ignoring science.
Richard Mayhew
@PST: And that is why I’m convinced that the denialist will be isolated — denial costs a lot of people with money a lot of money.