Philadelphia Reflections

The musings of a physician who has served the community for over six decades

Related Topics

Obamacare: Examination and Response
An appraisal of the Affordable Care Act and-- with some guesswork-- its tricky politics. Then, a way to capture major new revenue, even paying down existing Medicare debt, without raising premiums or harming quality care. Then, an offering of reforms even more basic, but more incremental. Finally, the briefest of statements about the basic premise.

Health Savings Accounts, Regular, and Lifetime
We explain the distinction between Health Savings Accounts, Flexible Spending Accounts, and Lifetime Health Savings Accounts. Sometimes abbreviated as HSA, FSA, and L-HSA. Congress should make it easier to switch between them. All three are superior to "pay as you go", health insurance now in common use, only slightly modified by Obamacare. It's like term life insurance compared to whole-life. (

Proposal Number Seven: Buy-In Prices at Different Ages

If we aim for lifetime (or "whole life") health insurance, using a Health Savings Account, provision must be made for the vast majority of people who do not buy it with a single premium at birth, as we use for a simple example. If we know the lifetime goal and the expected average rate of return, it is easy to project the average growth of accounts by any future year. A simple table of such projections becomes useful for displaying the "buy-in" costs for any age. Naturally, it incidentally underlines how costs increase for late-comers since essentially the same costs are distributed among fewer remaining years. Conversely, compound investing while you are young is very attractive. These are important selling points and are valuable lessons to learn. But no strong argument is improved by exaggeration. This is not a way to reduce the cost of medical care, it is a way to pay for some unnecessarily added costs inherent in choosing a "pay as you go" design. The use of compound income does indeed give the initial appearance of something for nothing but should be viewed as a more efficient insurance design.

If it does nothing else, lifetime health insurance clarifies where this money is coming from. Under Medicare, for example, a person contributes all his life but gets no income on that contribution. At some advanced age, he spends that money at prices which reflect the Federal Reserve's 2% inflation of the money, but he pays no taxes on the gain. Meanwhile, some younger person pays his bills at the inflated rate, and in due course inflates it again before he spends it. In the case given, the last of three generations are spending money which includes eighty years of inflation at 1-2% tax-free. However, he has an opportunity cost: the money could have been invested in index funds which Ibbotson has shown would have grown at 12% per year over the same time period, tax-exempt if he put it in a Health Savings Account. And it wasn't even mostly his own money at work. It really doesn't sound impossible for that amount of compound earning to pay for a great deal if not all of the lifetime cost of one person's healthcare, providing he does not pay excessive investment costs to do it. And since this conclusion is based on considerations which have almost nothing to do with the cost of medical care or increasing longevity, it is nevertheless impossible to make precise predictions. Any investment outside the insurance plan will result in a considerable revenue gain, probably a big gain, and possibly pay for all medical costs. Fundamentally, this money is generated by obtaining a higher return on the money, and not sharing much of it with financial agents, or other contestants for your wealth, like the health industry, or like the luxury goods industries. In that sense, it's just like any other wealth.

It also should not matter much whether the planning design aims for the account to terminate at death or when Medicare takes over. Naturally, the same lesson applies to the terminal end as to the buy-in date; the more you delay withdrawals, the more time there is for growth of the principal. With a growing tendency for costs to cluster around the last year of life, many "young old" retirees have only minimal medical expenses for ten or more years after retirement. Since invested money at 7% will double in ten years, there would be a considerable advantage for latecomers in transition. If the latecomer intends to terminate deposits at an attained age of 65, he will need about $40,000 to see him out. If he intends to terminate at death, he will need about $300,000, but the buy-in price at any age before 65 should be the same. As experience gathers, there probably will emerge some distinctions matching health status changes, but curiously the costs seem to decline after age 85. After the preliminary layering by age, the annual lump sum can be broken into installments more realistically matching the income and health practicalities of individuals. Each annual step of a table would represent the "buy-in" price at that age, which is also the average account size achieved by a lump-sum at birth by that age, without withdrawals. The point about "without withdrawals" should be seriously considered as a reason to substitute out-of-pocket payments for trivial expenses. With the passage of time, it can be made more precise by experience. But it should be kept in mind that a regular HSA only hopes to make as much income as possible, while a lifetime HSA seeks an average lifetime target. As experience accumulates, it may be found that required balances actually shrink with advancing age, as the individual lives past the time when expenses had been expected but not experienced. However, without experience, the best conservative assumption is that expenses steadily rise with age.

Since a Health Savings Account tries to serve several purposes, a particular account may not have enough deposit content to match the deductible, for example, because the cost of an individual's illness has little to do with his investment history. The manager of an account will create rules designed to protect his own position, and for example might have a minimum designed for investment purposes, which happens to be considerably short of what the high-deductible insurance company needs as a deductible for a particular age-group's sickness experience. For another example, a gift from a grandparent at birth may be adequate to cover lifetime expenses, but not at first. Only after it has multiplied several times will it be enough to meet the deductible. Some managers impose fees on the first $10,000 of deposits rather than reject the account, and it really only becomes an attractive investment a decade or more later. At present, accounts have an annual limit of $3300 for deposits, so it takes three years to reach a suitable size, and perhaps ten years if only a single deposit is made. Investors must learn to be highly resistant to brokerage fees, especially in new accounts. True, the regulations are likely to be highly changeable in the first few years, so investors much learn to pay fees from outside sources, in order to protect the tax advantage when it is most needed. Unfortunately, educating young investors about complex new regulations can be expensive for the investment advisor, so it is, unfortunately, true that the interests of broker and client are not well aligned in the early years.

The rules should be adjusted to recognize this problem, even though many people would find it unnecessary. A subscriber may have enough savings to make a single-payment deposit but is hampered by the $3300 rule. Finally, an account might once be large enough to be self-sustaining, but be reduced below that level by one or two depletions to pay deductibles. Generally speaking, the conventional HSA does not need to concern itself with such issues, which only become a serious problem if lifetime Health Savings plans are contemplated.

Consequently, the regulations should be modified in the following ways:

1. A family of tables is prepared, showing the deposit required to equal the total average future health cost for each yearly age cohort of life, from now until the average death expectancy, using various extrapolation assumptions. It is possible to reach the same goal with almost any investment assumption, or almost any time period, or almost any starting deposit, but only so long as the other variables are adjusted to conform to that specification. A family of tables would show several investment levels of compound interest reaching the same goal, let us say $40,000 at age 65, at ten percent; seven percent; and five percent. Obviously, a higher interest rate gets you to the goal sooner. Attention should be focused on achieving $40,000 at age 64.

Any subscriber should be allowed to buy into the lifetime Health Savings Account for a one-time deposit of $30,000 at age 48 assuming 5% return, at age 55 assuming 7%, or at age 57 assuming 10%, merely as a rough example. The subscriber may not have savings of that size, of course, and a separate calculation should be made for time payments, also reaching the same goal. If such tables are displayed on a computer terminal, it should not be difficult to make a selection, but "user friendliness" is often more difficult to achieve. As are later modifications, if life circumstances change. The relentless mathematics will soon demonstrate that the more money deposited, and the earlier it appears, the more attractive the investment becomes.

2. Salesmen for HSA should be required to carry their illustrations out to the goal of $40,000 at age 64, supplying achievement benchmarks along the way. So long as the account contains less than the buy-in amount for the subscriber's age, the manager of a fund should be allowed to wager a guaranteed band of investment results; let us use an example of 7% and 10%. If the funds make more than 10%, the manager keeps the excess. If the fund achieves less than 7%, the manager must make up the difference, either by reinsurance or by offering to be at risk for it. If fund results fall between 7 and 10%, the investor retains it all. This mandatory arrangement may (not must) terminate when the fund reaches a buy-in level, so long as it resumes if illness depletes the account.

The actual limits should be set after consultation with managers in active practice since the purpose is to create incentives to get the funds to self-sustaining levels without kickbacks to investment vehicles. The tension is between a subscriber who has investment choices to make, and a fund manager who must cover his expenses. In the long run, it is to everyone's advantage to maintain a steady view of the risks and rewards of income compounding, while serving the goal of paying as much as possible toward everyone's health care costs in a free-market system. It is best if the limits are realistic, and they should be chosen to drive all the participants toward the highest safe level of performance. Ultimately, the subscriber should recognize that an unsafely high level (with leverage, for example) will make paying his health bills more difficult, not easier.

3. The easiest definition of the top limit is the running cumulative average of the stock market, so total-market index fund results are ideal for the purpose. However, index funds differ in their results, so transparent competition should even be able to squeeze out somewhat better results than the 10% compounding which has characterized the past 90 years. Curiously results significantly worse than the market is not a sign of safety. Consequently, freedom to change managers should be as unhampered as possible, comparative results and costs widely available, and exit fees discouraged.

4. In general, single premium policies are rarely used. However, since starting an HSA at birth adds 26 years to the compounding, and while the amounts at first are so small they are somewhat unattractive to HSA managers, they could nevertheless become an important feature of their future. Throughout childhood, the paradox will be common that the necessary deposits in an account for lifetime coverage of health costs are nevertheless far too small for a deductible which is sensible for children with such low health expenses. Therefore, provision should be made for supplemental policies which cover this gap in childhood between the amount in the account and the amount of the deductible, which is often set with an eye to the parents' situation, not the child's. Intuitively, such insurance supplements ought to be quite cheap.

Originally published: Thursday, April 10, 2014; most-recently modified: Monday, June 03, 2019