«TIPS, the Triple Duration, and the OPEB Liability: Hedging Medical Care Inflation in OPEB Plans By Michael Ashton, CFA Managing Principal Enduring ...»
TIPS, the Triple Duration, and the OPEB Liability:
Hedging Medical Care Inflation in OPEB Plans
By Michael Ashton, CFA
Enduring Investments LLC
Morristown, New Jersey
Copyright 2011 by the Society of Actuaries.
All rights reserved by the Society of Actuaries. Permission is granted to make brief excerpts for a
published review. Permission is also granted to make limited numbers of copies of items in this monograph for personal, internal, classroom or other instructional use, on condition that the foregoing copyright notice is used so as to give reasonable notice of the Society’s copyright. This consent for free limited copying without prior consent of the Society does not extend to making copies for general distribution, for advertising or promotional purposes, for inclusion in new collective works or for resale.
Abstract The adoption of FAS 158 forces sponsors of post-employment health benefit plans to consider how to manage the volatility that changes in medical care inflation create in the other postemployment benefits (OPEB) liability. By choosing carefully how the nominal discount rate used for the liability is decomposed into real rates and inflation, I illustrate that the true exposure to an OPEB plan is to the spread of medical care inflation above (or below) the overall inflation rate. The implication is that an effective immunization strategy exists that can eliminate most of the volatility in the OPEB account.
Page | 2 Introduction One of the more critical dilemmas facing sponsors of post-employment health benefit plans is the question of how to manage the volatility that changes in medical care inflation create in the other postemployment benefits (OPEB) liability. This dilemma has grown more acute since FAS 158 required that OPEB funding surpluses or deficits be brought on-balance-sheet by year-end 2007. GASB 43 and 45, the equivalent guidance for public plans, have less-onerous provisions that nonetheless are leading to considerable energy being spent on structuring the liability accounts so as to make the impact on public sector finances (and on the financial statements of public sector entities) less dramatic. However, if public sector plans only act to immunize themselves against the accounting impact and not the economic impact of fluctuations in medical care inflation, they are not addressing the full problem; and so the conclusions drawn herein are applicable as well to sponsors of public sector plans.
In 2004, Siegel and Waring published a landmark paper titled “TIPS, the Dual Duration, and the Pension Plan.” The authors argued that the main sources of volatility in the funding status of pension plans—changes in the interest rate used to discount future liabilities, coupled with fluctuations in the asset portfolio designed to provide the means to pay those liabilities—can be addressed by reallocating the pension assets to Treasury Inflation-Protected Securities (TIPS) and Treasuries. 1 It may seem that the facts that medical care inflation is not currently tradable and that OPEBs are generally poorly funded or unfunded offer little hope of a similarly optimistic outcome, but this is not the case.
This paper discusses some notional approaches to addressing the problem of open-ended medical care inflation exposure, and proposes some practical steps that can be taken to ameliorate this risk. It leverages off the work of Siegel and Waring in doing so. Prior to that, though, I will discuss medical care inflation generally and how it affects the OPEB liability.
1. Trends in and Possible Causes of Medical Care Inflation
What is most striking about medical care inflation is its durability. Over the last 20 years ending in December 2010, the broad category of Medical Care consumer price index (CPI) has outpaced “headline” CPI 2 by 1.79 percent per annum. 3 Over a 40-year period, this subindex has exceeded headline inflation by 1.87 percent per annum. Putting this in perspective, one dollar today, broadly speaking, buys between one-fifth and onesixth as much as it did in 1970, but only one-eleventh as much when used to purchase medical care (more poignantly, two-elevenths versus one-eleventh). Expressed another way, $1 invested in the one-year Constant Maturity Treasury (CMT) rate at the end of Siegel and Waring. 2004. “TIPS, the Dual Duration, and the Pension Plan.” Financial Analysts’ Journal September/October, pp. 52–64.
Headline CPI is typically represented by the year-on-year percentage change in the Non-Seasonally Adjusted Consumer Price Index for All Urban Consumers.
All CPI price index data is sourced from the Bureau of Labor Statistics; these data may be accessed at http://www.bls.gov/cpi/home.htm.
Page | 3 1970 and rolled annually would have nearly doubled in real terms 4 broadly, but in terms of medical care that investment would today buy almost exactly as much medical care as it did previously—the risk-free real return, in terms of units of medical care, was essentially zero over the last 40 years because medical care prices have outpaced overall inflation so handily.
The Bureau of Labor Statistics breaks Medical Care CPI into various subindices.
Here is how those subindices have grown over the last 20 years, outright and in real terms. Column 3 is merely Column 2 minus aggregate headline inflation.
Confronted with these figures, one’s first reaction is to ask, “Why?” Why has inflation in medical care so consistently beaten generic headline inflation? An explanation tied to the demographic bulge associated with the baby boomer generation is enticing, since it is reasonable to postulate that effect as an important one now as the boomers start to retire, but this is clearly insufficient. The boomers could not have been pushing on the system for the last 40 years, and as noted above the premium to medical care inflation has been that durable.
Obviously, the state of medicine and the quality of medical care has increased exponentially over the last few decades, but the BLS indices are supposed to “hedonically” adjust for this tendency of goods to cost more simply because they have improved. However, the possibility exists that the hedonic adjustments made for medical care are insufficient. 5 Source for year-end one-year CMT rates is the Federal Reserve H15 report. Actually the increase is 85 percent.
One way to test this hypothesis might be to examine how the weighting of Medical Care CPI in the overall CPI has changed over time, and thus measure the rate of substitution between medical care and other goods (if the demand for medical care is perfectly inelastic, the increase in weighting would match the relative price change). By comparing this elasticity measure with other measures of the demand elasticity of health care, it may be possible to test the hypothesis that the hedonic adjustments are adequate.
Page | 4 Two other possibilities suggest themselves. The first is that, as with so many goods, the increasing involvement of government as a purchaser of medical care might work to increase prices in the sector. A monopsonistic purchaser should work to push inflation lower, but this is much less likely when the purchaser is not driven by anything approximating a profit motive. More significantly, some important parts of what the government buys are excluded from Medical Care CPI since the government is not, after all, a “consumer” in the usual sense of the word. Medicare expenditures, for example, are excluded from CPI. It may be that health care providers raise prices in the part of the market not dominated by a monopsonist (the U.S. government) in order to compensate for the lower price demanded by the monopsonist.
10% 5% 4% 5% 3%
-5% 0% Another possibility is supply-side in nature. The increasing cost of malpractice insurance and of tort and patent defense, driven by the ballooning tort-lawsuit industry, may have shifted the supply curve for medical care to the left, resulting in higher prices.
The fact that one of the biggest drivers of medical care inflation (as in Table 1) is Hospital Services is suggestive.
The purpose of this paper is not to answer the question of why medical care inflation has been such a problem in the past, nor whether it is destined to remain so in the future. The purpose here is to address the problem, given that medical care inflation has been historically high (and variable), that such a phenomenon poses for sponsors of OPEB plans that include health benefits.
2. Medical Care Inflation in the OPEB
Pension and OPEB plans accumulate liabilities on the basis of promises they have made to covered employees. In the case of a pension plan, the promise is in the form of a cash flow or series of cash flows: “We will pay you 60 percent of your final salary.” The present value of such a promise, ignoring the possibility that the benefit may be frozen or reneged upon, can be arrived at by calculating (a) the expected final salary, which is a function of the expected time to retirement, the current salary, and the rate of wage Page | 5 inflation 6; (b) the payout schedule given a retirement date, which is a function of expected mortality and the particulars of the plan (employee or employee plus spouse, nominal or inflation-adjusted, etc.); and (c) the discount rate curve, which may be a riskfree or more typically a corporate (aka “risky”) curve. 7 Siegel and Waring demonstrated that a “typical” plan had certain exposures to real rates and nominal rates that flow naturally from the modeling of sensitivities of this sort of plan.
In the case of medical care inflation, the promise is in the form of services: “We will provide you with health care coverage at a certain level.” The process for figuring the present value of future health care liabilities is reasonably similar except that the actuary must consider medical care inflation rather than wage inflation and also model how a retiree’s consumption of health care might vary over time, since benefits are not promised in terms of a certain dollar value of health care but in terms of a certain breadth of coverage or quality of care. For example, an employer might offer to extend the employee’s preretirement health care plan into retirement, perhaps with modifications to co-payment and out-of-pocket maximums. For any given employee, this could result in a wide range of actual expenditures based on the retiree’s health, his/her family’s health, disincentives (or a lack of disincentives) to consume medical care, and so on. Unlike a pension claim, which is reasonably estimable for any employee expected to reach retirement age with the company, the OPEB claim for a given employee is frighteningly unpredictable. It is only when these claims are aggregated for large numbers of employees that the resulting distribution becomes passingly actionable.
Consider the following hypothetical plan covering 10,000 retirees and with 10,000 current employees eligible to participate upon retirement:
The wage inflation of the company in question, that is, for employees who will still be employed as they reach retirement. This is a conditional rate of inflation that is subject to many biases. For example, it will tend to be higher for high-productivity industries and for companies that are successful at retaining quality employees; it will tend to be lower for companies subject to adverse employee selection (ones that get poached a lot or that offer early retirement plans that tend to favor the re-employable). Et cetera.
Technically, if the pension fund is a liability of the company and secured with the general resources of the firm, the discount rate should be a rate reflecting the riskiness of the sponsor’s credit. However, this would lead to the curious effect that a firm could reduce its pension fund (and OPEB) burden by becoming less creditworthy (which reduced burden, in turn, could increase its creditworthiness!). With pension funds, which are required to be secured by actual funding, a “market” corporate credit curve is generally used but with the typically unfunded OPEB plan, this is a more poignant question. Public plans under the Governmental Accounting Standards Board (GASB) actually are supposed to use a discount rate linked to the expected return of the assets, which produces the perversity that the more crazy the risk the plan takes, the lower are the liabilities.
For simplicity, we will assume the plan closes to new employees at this point, but no one ever quits or is fired. The plan therefore runs down only as employees exit the plan by dying, and the entire plan has roughly a 50-year average life. 8 In the base case, I assume somewhat typical costs and distributions of plans between Employee Only, Employee plus Spouse, etc., retirement age at 65, normal mortality experience, and medical care inflation that declines from 8 percent to 5 percent in four years (and then remains at 5 percent indefinitely). Under these assumptions, as the chart below shows, total medical care costs rise for about 16 years before finally declining; a 1 percent rise in the trend rate of medical care inflation, assuming no other change in the inputs, increases the peak by about $22mm per year. Even in present-value terms, this is an adverse change of $186mm in the OPEB liability—and this assumes the plan is no longer available to new employees.
Obviously, considerable actuarial critique can be leveled at this model, but I make these assumptions merely to illustrate the scale of the OPEB liability and its sensitivity to health care inflation.