The Efficient Frontier Theory

Submitted by on Sunday, March 29, 2015 - 07:01

Newtown, CT – March 28, 2015

In 1952, Harry Markowitz published a paper on the "Efficient Frontier," which is a line that defines the optimal possible return for the investment in a portfolio – The basis for the theory is relatively simple, highlighting that any number of securities, each with its underlying risk profile, can be assembled into a portfolio and yield a potential return. That return is based on the difference between the purchase and future expected sales price, and any dividends or other income that can be generated from the securities. The risk is based on the intrinsic volatility of the security. Because the theory presumes the existence of a market in which the securities are traded, the purchase price creates the limits of the total potential return. In other words, if a security is very risky, then the buyers will likely demand a premium, which would translate into a low enough purchase price to justify the potential risk associated to holding that security.

What this means to you – As providers take on increasing financial risk, the Efficient Frontier theory may hold important lessons. If you substitute the security in Markowitz's model with an episode of care (or patient), you can assemble these episodes (or patients) into portfolios (or populations). Each episode has an intrinsic volatility which is a function of the observed variability in costs and quality outcomes. Some of that volatility is inherent to the episode – highly complex conditions that are difficult to manage and for which the treatments might vary due to patients' genotypes; some of it is inherent to the providers – lack of skills or processes in the management of a condition; and some of it might simply be due to variability in patient compliance. Providers, therefore, should expect a return for taking on the portion of the risk that is outside of their control and inherent to the episode. And the greater that risk, the greater the premium. Of course, that would suggest that providers actually understand the differences in the risk profiles of the episodes and the potential return for taking on that risk, but, for the most part, that's simply not the case. So what's happening today is a completely inefficient allocation of risk, without distinction of the return for the different combinations of risky episodes/patients that are part of provider's portfolio/population, and reminds me of my first exposure to Wall Street investment banking decades ago. Bond traders on the sales floor were busy selling to state and municipal retirement funds, and I was observing what a day in a possible future career might be. As one trader hung up after a significant sale, he high-fived his colleague and self-congratulated himself for having unloaded a pile of junk. He knew the inherent risk of the security, and he had gotten someone else – far less sophisticated – to pick it up, and pocketed a nice margin. Since then, the securities markets have become far more transparent and far more efficient. Health care has a long way to go, and while payers aren't high-fiving themselves for unloading risk on providers that don't understand what they're "buying", we are very far from the Efficient Frontier.