Pricing Risk: What Margin Should Risk-Bearing Healthsystems Reasonably Expect?

There is an old adage in the business world: reward follows risk. In other words, in an ideal market, investors are supposed to balance the potential returns of an investment against the possibility that they’ll lose their money.

4376727123_8fc3fb172d_zI’ve thinking a lot about risk recently as our organization came to the end of our budget process.  Philosophically, the question was this: what should the financial rewards be for a healthsystem that decides to take on medical risk? Are budgeted margins appropriate for the degree of exposure that groups assume when they migrate from fee-for-service?

Here are some high-level thoughts that I’ve had.   First– as caveat– I’m a doctor and not a macroeconomist… these ideas are traditionally applied to financial investing, not medical management… and there are nuances to be sure.  But, at a high level, here are some basic assumptions:

Rule 1: First, for every investment, there should be a risk/ return index. It’s the ratio between the amount of risk an investor assumes and the potential return he/she can make on the investment.  The idea is that as risk increases, reward should follow. It’s why government bonds yield 1% and junk bonds yield 10% or more. Investors are paid for exposing themselves to market fluctuations.

Rule 2: Reward should be compared to a “risk free” investment– in investment circles, buying Treasury Bills.

Rule 3: Risk is often considered to be the degree of volatility in a business, or the amount of “movement” around the expected outcome. The greater the volatility, the greater the chance that the return isn’t what was expected.

Economists have developed a ratio of reward to risk, called the Sharpe Ratio. It’s designed to evaluate the returns on an investment, by normalizing out the “riskiness” of an investment. The Sharpe ratio should tell you whether the return you are seeing is due to the inherent “riskiness” of the investment versus sound business decision making.


Screenshot 2014-12-30 10.22.56




Economists argue that a Sharpe Ratio should be above 1—and better investments have ratios of 2 or 3.  These investments generate high returns with low risk. Bad investments, conversely, generate low returns at high risk and have low Sharpe indices.

In the world of healthcare there are two basic ways to get paid: You can take fee-for-service revenue, which isn’t particularly generous (and has been falling each year) but is fairly predictable (relying only on volume and negotiated rate).  

Or, as I’ve written about before, you can accept risk payments which provide lump payments to cover an episode of care—or at their extremes, as in the case of global capitation, all the medical services a patient needs while enrolled as a patient of the system.

Systems like the one I work for take risk because 1) we believe that there is an opportunity to generate a better margin than FFS assuming we do well with medical management and 2) being paid on risk contracts gets us off the “RVU treadmill” and enables us to develop creative, effective and crowd-pleasing healthcare programs that wouldn’t be covered under FFS.

But, taking on risk is not without… well… risk. Plenty of things can happen over the course of a year which drive up medical expense. There can be a bad flu year, or a new drug, or a rash of patients who need things like liver transplants. And, the system is on the hook. Surely there should be some upside for systems who elect to take on risk: what should they reasonably expect?

We could use the Sharpe ratio to guide the pricing of risk here.  I presume Dr. Sharpe didn’t design it to price medical risk-bearing, but I like the concept in that it tells us that there needs to be a certain margin to justify the exposure.

Using Sharpe ratios, I’d argue that at a minimum systems taking risk could anticipate making a margin equal to the margin they’d make under FFS (which is effectively the “risk free” margin), plus an additional amount that reflects the volatility of the risk (medical expense standard deviation to budget) that the company is exposed to.  This gets you to a Sharpe Ratio of 1. It should be more than this for higher Sharpe ratios.

Screenshot 2014-12-30 10.23.34




Thus: Margin in Risk= (volatility of medical expense x Sharpe ratio)+ Margin in FFS

At a high level, after the budget factors in revenue and expense, the margin in a risk setting should equal the margin you’d get in FFS plus an added percentage, that equals the “risk profit” you pay yourself as the reward for your exposure.  For groups with stop-loss medical expense reinsurance, the variability is theoretically limited to the limits of the policy.

Now, according to the AMGA, most medical groups are breakeven this year, with a few in the black. For the sake of the calculation, lets assume that the FFS margin is zero at most places.   At a Sharpe ratio of 1, groups should anticipate risk margins equal to the variability of the medical expense they assume. A group with a standard deviation of 5% should budget a “risk profit” margin of 5%. The variability of medical expense is what should drive margin expectations.

(This is obviously hard to do for a small system, since the smaller you are the greater the percentage of year-to-year medical expense variation around the mean. Which is one reason why risk is increasingly driving consolidation in healthcare… you need to be large enough to absorb random variation in medical expense without blowing up).

For groups that aren’t close to where they need to be there are two possible culprits. 1) there is somehow a performance issue or 2) there are fundamental challenges to the business model itself.

Even though many folks might argue that the upside to medical systems taking risk is nearly unlimited (the margin lies in managing patients well) I’d argue that risk is a tough business that starts out lean. After all, insurance companies keep up to 5-20% of premium revenue as overhead and margin (it’s limited to 20% by the Affordable Care Act) so medical groups start with revenue that’s close to the minimum medical cost ratio (the amount that is spent on healthcare) plus a bit.

There isn’t a lot of capacitance available in case of a bad year. To mitigate some of their exposure, medical groups buy medical expense reinsurance, again drawn from their pass-through dollars from the insurance company. At the end of the day, the margin opportunity lies in managing medical expense, which is hard work- risky, complex and challenging for groups that haven’t made insurance-company-level investments in medical management.

All that said, I don’t see the move toward risk slowing anytime soon. The problem is across the country FFS margins are approaching negative territory for most physician groups—even after the $120,000 subsidy that the AMGA is reporting that 1/5th of groups receive from affiliated health systems.

Here is the 2014 AMGA report showing operating margin per physician nationally, by region from 2009- 2013.  Most are predominantly paid FFS.

Screenshot 2014-12-30 10.21.01

So, despite the threats of being in risk, it seems to be the only way to generate a margin– even if that margin seems relatively underpriced relative to the exposure groups take on. 

It’s not too different from the financial markets, where risky investments don’t seem to be generating the kinds of returns that they should.  A recent article in the Atlantic noted that investors are undervaluing risk given the meager returns available for “safe” investments.  The article notes:

As happened in 2005 and 2006, investors are searching desperately for good returns on their investments—a natural reaction when returns on savings accounts and high-quality bonds are so miniscule. As a result, risk is once again being badly mispriced by the debt markets. The yield on junk bonds—representing the credit quality of the riskiest companies—is typically about 10 percent. These days junk bonds yield about 6 percent, a sure sign that too many investors are chasing dicey securities and not getting properly paid for the risks they are taking. Given these favorable conditions, junk-bond issuances have soared: at the time of this writing, new junk bonds totaling $290 billion had been sold in 2014, on par with 2013 and nearly on pace with the record issuance of $345 billion in 2012.

Medical risk-bearing has nothing to do with sub-investment grade bonds.  But, the challenge of pricing risk appropriately resonates here.  What reward should risk-bearing groups expect?  Looking forward, do they have a better option?



Photo: via flikr, creative commons license, and AMGA