April 2019 Newsletter

In a recent Journal of Financial Planning article, I analyzed the costs and implications of selling investments during downturns. By my estimates, and others in the field, the cost to everyday investors is on the order of 100 to 150 basis points, annualized. Even if you aren’t likely to panic during a downturn yourself, though, the potential for panic by others affects your investment environment because of how the risk of emotional selling is built into common practice in asset allocation. Let’s take a look at that research, and what it means for investors and investing.

By focusing one’s asset allocation squarely on ones goals and what is needed to achieve them, and using behavioral tools to mitigate emotional challenges, we can find a more effective answer than current risk-preference based allocations.

How Bad Is It?

We’ve all heard horror stories of people who panicked during a downturn, went to cash, and then lost out when the market recovered. Or of people who simply became uncomfortable with a sector- or investment-specific dip and dumped an existing investment only to see it promptly recover (and outperform the “new” investment they chased after).

In this research, I start by analyzing the baseline: What would emotional selling do to a sample set of investors? I developed a simulation model of investing over time, similar to one in my recent paper on easing the retirement crisis, and explicitly modeled the probability of selling during downturns due to emotional triggers. Depending on each person’s risk preferences, they had a threshold at which they would sell; those with lower risk preferences panicked more quickly than those higher preferences. I then put them through a range of historical scenarios to see how they would behave over a 10-year period.

These baseline results came back as a loss of between 100 and 150 basis points, which aligns the range of estimates other researchers have found. The most well-established estimates look at the costs of unwisely trying to time the market, of which emotional selling during a downturn is believed to be the most pernicious part. Researchers Friesen and Sapp (2007), for example, estimate that it decreases investor returns by about 150 basis points per year. Vanguard also places it at about 150 basis points (Bennyhoff and Kinniry 2013). Morningstar’s own Russ Kinnel, in his annual Mind the Gap study has shown how it has varied considerably from sector to sector and year to year, from over 150 to a current average of 26 basis points.

That’s what happens to people who panic. But what about everyone else?

Emotional Investing Shapes Asset Allocation

One of the key lessons from the analysis is that the manner in which investments are often matched to individuals (often using a risk-tolerance questionnaire) is shaped by the potential for emotional selling. When one develops a long-term financial plan, at least two competing demands are placed on the asset allocation: taking on risk to help you reach your financial goals and avoiding excessive risk, in part to avoid discomfort and negative emotional reactions that might lead you to abandon the plan.

To manage these competing demands, the field of financial planning generally applies two approaches: a risk capacity approach which focuses on goals and generating the required returns, and/or a risk preference approach that seeks to avoid panic. Unfortunately, in isolation or in combination these two approaches can fail to meet either of the two demands–failing to help investors effectively reach their goals because the portfolio was positioned too meekly, and failing to forestall panic because the portfolio was intolerably volatile.

In this paper for example, I found that using a risk-preference approach to place more sensitive investors into lower volatility investments reduced the damage from emotion-based selling only modestly, decreasing the losses for investors from 100 to 150 basis points to roughly 80 to 130 basis points. The problem is that when a downturn hits, even lower volatility investments can drop enough to trigger emotional selling among risk-intolerant investors (and similarly for risk-tolerant investors in higher volatility investments). To mitigate this problem, one needs to address the behavioral challenge head-on.

A Different Approach

Yet, it’s also possible to address the risk of emotional selling without changing the asset allocation. Investors and, as appropriate, their advisors can use behavioral tools to help prepare for and respond to volatility when it comes. For example, potential techniques from the literature include:

• To alleviate loss aversion, investors (and advisors) can avoid frequent price updates (Larson et al. 2016).

• To decrease the likelihood of selling low, one can look for long-term investments that are explicitly “set it and forget it”; this approach learns from the positive behavioral outcomes that target date funds have achieved (Holt and Yang 2016).

• Investors and advisors can become more educated about how all people suffer from common behavioral biases, like confirmation bias and availability heuristic, that lead to predictable and avoidable mistakes (Perttula 2010). Ideally, these lessons should be reviewed right before a trade, not as a general investor education or financial literacy program (Fernandes, Lynch and Netermeyer 2014).

Measuring the Impact of Investor Panic and a Combined Approach

In the final part of the analysis, I looked at the impact of a combined approach: behavioral tools to mitigate emotional selling, and an asset allocation based solely on one’s needs (risk capacity, not risk preferences). The analysis couldn’t analyze the specifics of each of the behavioral tools above–much more work is needed there–but it could determine what would happen, in aggregate, if we can forestall emotional selling overall. In aggregate, investors focus on needs and avoid panic receive a net increase of 17-23% in assets over 10 years, or roughly 170 to 225 basis points per year in returns. Some of that return comes from avoiding panic, which supports the analyses by Vanguard and others. The rest comes from freeing up asset allocation to better serve the financial needs of investors. That removes the tension between the competing demands above: achieving an investor’s financial goals and avoiding uncomfortable volatility at the same time.

Investing is full of uncertainty, and investors are broad and diverse. This research provides an initial look at how one can combine a new set of tools from the behavioral science community, with thoughtful asset allocation, to help some investors avoid emotional selling during downturns. To see the full research paper, please check out “Using a Behavioral Approach to Mitigate Panic and Improve Investor Outcomes” Journal of Financial Planning 31 (2): 48–56.

^{Source: Steve Wendel, September 21, 2019}

Run an internet search for “Medicare for all and seniors,” and you’ll see plenty of posts explaining just how harmful single payer health care would be for retirees.

The Medicare for All debate is just starting, with many questions yet to be answered. But don’t be too quick to assume that Medicare for All would take away something that seniors enrolled in the current Medicare program now enjoy. Just the opposite: The proposals circulating in both the House and Senate would improve Medicare by eliminating the program’s most glaring coverage gaps for long-term care insurance, and for dental care, vision and hearing. The bills also would eliminate cost-sharing, meaning that one of traditional Medicare’s weakness would be gone–the absence of a cap on total out-of-pocket spending.

Here’s a rundown of the risks Medicare enrollees currently face for each of these five areas, in rough order of their relative importance. As a proxy for Medicare for All, we’ll use the Medicare for All Act of 2019, sponsored by Rep. Pramila Jayapal, D-Washington, which is widely seen as the most pure Medicare for All plan circulating in the House. Sen Bernie Sanders, I-Vermont, has long been a proponent of Medicare for All, and his Senate legislation is expected to be similar in most respects to Jayapal’s bill.

Long-term Support and Services

The Jayapal bill would cover long-term care, including nursing and medical services, rehab services, and services to support daily living. Services would be provided in a home or community-based setting, unless the patient requests otherwise.

The proposal aims to replace what just might be the most dysfunctional private insurance market in the U.S., and a key area of unprotected risk facing seniors. It is crystal clear that Americans just don’t have much interest in buying private long-term care insurance. The American Association for Long-Term Care Insurance reports that fewer than 60,000 traditional (health-based) long-term care insurance policies were purchased in 2018, down from 700,000 in the peak years of 2000 through 2002. (Roughly 300,000 hybrid policies combining life insurance and long-term care protection were sold, the organization reports.) The latest sign of the industry’s struggles came last month, when one of the industry’s largest underwriters–Genworth–announced that it would stop selling individual policies through brokers and agents.

The flagging sales stem from a plunge in the number of carriers underwriting new policies, and negative headlines and spiking premium rates for existing customers. Confusion about the need for insurance may also be a cause: a recent survey by Bankers Life found that 56% of middle-income boomers think Medicare does pay for ongoing long-term care; in reality, Medicare pays only for the first 100 days in a skilled nursing facility after a hospitalization.

Yet the need for long-term care remains one of the biggest unknowable risks retirees face. A study by Rand Corp. in 2017 found that 56% of people now age 57 to 61 will spend at least one night in a nursing home during their lifetimes. People currently in this age group run a 10% risk of spending three years or more in a nursing home and a 5% chance of needing more than four years of care.

Those odds may worsen in the years ahead. Howard Gleckman of the Tax Policy Center noted in a recent post for Forbes that the widespread use of statins and other life-prolonging drugs are reducing the growth of medical costs for seniors–but may also help them live longer. That increases the odds that they will survive long enough to ultimately require long term support and services.

“To oversimplify instead of dying of heart attacks at 60, more of us will live to 85, when we will get dementia,” he writes. “That’s why we need to shift resources from medical care to long-term support and services.”

The costs can be staggering. Last year, the national monthly median cost of a private nursing home room was $8,365, according to the Genworth Cost of Care survey. And in high-cost states, the expense can be far greater. In New York state, for example, the monthly median cost last year was $12,189, Genworth found.

I have long argued that at least part of the long-term care solution must involve social insurance. An automatic payroll deduction for all workers will be the only way to get everyone contributing to a system that will provide risk protection affordably, mainly due to the enormous risk pool that would be created. This was one of the acknowledged conclusions of several well-regarded reports released in 2016 developed by a nonpartisan consortium of researchers. Those reports called for a hybrid of publicly financed risk pools to cover basic care, supplemented with optional privately-offered policies.

Out-of-Pocket Spending

Currently, traditional Medicare has no cap on total out-of-pocket costs. The two big risks for seniors here are hospitalization and prescription drug costs.

The 2019 Part A deductible is $1,364 and the daily coinsurance charge for longer hospital stays (61 to 90 days) is $341. In the unlikely event of a very long hospital stay, traditional Medicare covers up to 90 days of inpatient hospital care for each instance of care; enrollees also have another 60 days of coverage, known as “lifetime reserve days.” These are like a bank of days that can be used once. (The Part B deductible this year is $185; you can view all Part A and Part B premiums and cost-sharing information for 2019 here.)

Currently, there are two ways to protect against high out-of-pocket hospitalization risk. Enrollees in traditional Medicare can purchase a Medigap supplemental plan, which carries an annual cost ranging from as little as $2,000 to $7,000 for the most comprehensive plans. Or, they can enroll in Medicare Advantage. These plans do put an annual limit on your maximum out-of-pocket expense, not exceeding $6,700 per year. (The average ceiling last year among all plans was $5,185, according to the Kaiser Family Foundation.)

For prescription drugs, even people who have bought Part D insurance face significant out-of-pocket risk in the event of a serious illness requiring specialty medications. A recent report by the Kaiser Family Foundation found that Part D enrollees are exposed to thousands of dollars in out-of-pocket costs for drugs that treat cancer and other serious illnesses. The researchers studied expected annual out-of-pocket costs this year for 30 specialty drugs used to treat four conditions: cancer, hepatitis C, multiple sclerosis, and rheumatoid arthritis. Median out-of-pocket costs ranged from $2,622 for Zepatier (for hepatitis C) to $16,551 for Idhifa (for leukemia).

Dental Coverage

Many retirees are surprised to learn that traditional fee-for-service Medicare does not cover dental care, despite its critical role in preventive healthcare. Research shows clear links between poor oral health and chronic disease such as diabetes, as well as pain, chronic infection, and reduced quality of life.

Many seniors simply pay for dental care out of pocket; the average out-of-pocket expense among Medicare enrollees who needed dental care in 2016 was $607, but expense can run much higher if you need a crown, bridge, or root canal, for example.

A new research brief by the Kaiser Family Foundation finds that almost two thirds of Medicare beneficiaries do not have dental coverage and many go without needed care–almost half of all Medicare beneficiaries did not have a dental visit within the past year (49%). One in five spent more than $1,000 out-of-pocket on dental care in 2016.

Traditional Medicare will pay for dental care only in very limited circumstances–it must be deemed necessary as part of a covered procedure; for example a tooth extraction needed in preparation for radiation treatment. Many Medicare Advantage plans include some dental coverage, but usually cap annual benefit payments at $1,000 to $1,500. And cost-sharing is higher for surgery, restorative services and periodontal procedures than for preventive services.

Hearing Costs

Hearing loss has been linked with increased risk for cognitive problems and dementia and higher risk of falling. Federal data shows that nearly 25% of those ages 65 to 74 and 50% of those who are 75 and older have disabling hearing loss. But many seniors don’t use hearing aids, probably due to cost–a device for just one year can cost upward of $2,000.
Yet traditional Medicare does not cover hearings aids (some Advantage plans cover hearing exams and hearing aids). The Jayapal bill would provide full coverage for vision and audiology services.

Vision Care

Traditional Medicare does not cover routine eye care, such as exams, with some exceptions, such as people with diabetes or those who are at high risk for glaucoma. Medicare also covers some surgical procedures to correct chronic conditions, for example cataracts or glaucoma. The program also covers certain surgical procedures.

Many Medicare Advantage plans offer some level of vision care, typically routine exams and eyeglasses.

^{Source: Mark Miller, April 8, 2019}

We know this: Affluent retirees love to hate their required minimum distributions.

Taxes are the major reason. The amounts that people who are age 70 1/2 or older must begin withdrawing from their tax-deferred accounts can push them into higher marginal tax rates than in the pre-RMD period, and can also increase the percentage of their Social Security benefits that are taxable.

But there’s another complaint I’ve heard about RMDs: that these mandatory withdrawals could lead retirees to prematurely deplete their assets if it turns out that their RMDs were too rich, especially if they live way longer than the “average” life expectancies that underpin the IRS’ tables for RMDs. Required minimum distributions start at a comfortable 3.6%, but ramp up to 5.3% at age 80 and are at 6.8% at age 85.

I have a few responses to that concern. First, just because the RMD rules require you to get the money outside of a tax-sheltered wrapper and pay taxes on it, there’s nothing saying you have to spend that withdrawal. I’ve written about how to reinvest unneeded RMDs in a tax-efficient way: If you find yourself with the high-class problem of your RMDs exceeding your living expenses, you can keep that money working for you in a taxable account or even in a Roth IRA if you or your spouse has earned income.

Nor should you be worried about too-high RMD amounts if you also have substantial assets that aren’t subject to RMDs. After all, it’s total portfolio withdrawals that matter to portfolio sustainability, so if you’re concerned your RMDs are too high, you can hold down your total portfolio withdrawal rate by withdrawing limited sums from your non-RMD-subject accounts.

In addition, your withdrawals should ramp up as you age and your life expectancy declines, especially if you don’t have a goal of leaving substantial assets behind for heirs or charity.

Finally, it’s worth noting that the assumptions underpinning the distribution periods for RMDs on the Uniform Lifetime Table, which is what the IRS requires most people to use to calculate their RMDs, are conservative. Understanding that might help allay fears about RMDs being too rich.

A Closer Look at the RMD Formula

If you’ve looked closely at the Uniform Lifetime Table, you’ve probably observed that the distribution periods seem pretty long. For a 70-year-old just starting RMDs, the distribution period is more than 27 years. At age 80, the distribution period is almost 19 years. (As with all actuarial tables, the longer you live, the longer you’re expected to live.) Those periods are longer than life expectancy: For example, the average life expectancy for a 70-year-old male in 2015 was 14 years, according to the Social Security Administration’s website, and 16 years for women. At age 80, the average life expectancy for men is eight years and 10 for women.

Why the disconnect? For one thing, the Social Security figures are based on a single person’s life expectancy, whereas the Uniform Lifetime Tables are based on joint life expectancies. As Morningstar director of policy research Aron Szapiro writes, RMDs are designed to ensure that money comes out of tax-deferred mode–and that taxes on those distributions are paid–during the account owner’s lifetime. But many account owners are also concerned about leaving assets for their surviving spouse. The distribution periods for the Uniform Lifetime Table seem relatively long because they’re meant to account for two lifetimes.

In addition, the formula now used to determine an IRA account owner’s distribution period makes the very generous assumption that the beneficiary spouse is 10 years younger than the account owner, even though that may well not be the case. (There’s a separate table for IRA owners whose spouses are more than 10 years younger; the distribution periods on that table are longer still.) The net effect of this assumption, especially for single people or spouses who are close in age and have similar life expectancies, is that RMD-based withdrawals are quite conservative.

Take, a 75-year-old husband and a 72-year-old wife, each of whom owns a $1 million IRA. The 75-year-old is using a 22.9-year distribution period, whereas the 72-year-old is using a 25.6-year distribution period. The 72-year-old’s spouse is actually three years older, but her distribution period is calculated with the assumption that he’s 10 years younger. Even if both partners have reason to believe their life expectancies will exceed the averages, there’s still plenty of wiggle room built into their withdrawals. The 75-year-old would have an RMD of $43,668, or 4.4% of his balance, whereas the 72-year-old’s withdrawal would be $39,063, or 3.9% of her balance. Given their ages, those withdrawal rates easily pass the sniff test of sustainability. In fact, their withdrawal rates are arguably too conservative, especially if they’re not aiming to leave assets behind for family or charity.

For spouses who are further apart in age, the Uniform Lifetime Table is a more accurate depiction of their actual life expectancies. (As noted above, there’s a separate table for use by IRA account owners whose spouses are more than 10 years younger; it takes into account both partners’ actual life expectancies.) If one or both such partners believe they’ll exceed average life expectancies, it’s reasonable to reinvest a portion of RMDs as a safeguard against premature asset depletion.

For example, let’s assume a 75-year-old account owner and a 66-year-old spouse; the older spouse’s IRA account makes up most of the couple’s assets. They don’t have a more than 10-year gap, which would necessitate use of the separate table for calculating RMDs, but it’s still a significant age discrepancy. The 75-year-old account owner’s distribution period on the Uniform Lifetime Table is 22.9 years, and the average life expectancy for a 66-year-old woman is about 20 years. If the younger spouse believes that she’ll live longer than average, that’s a good case for reinvesting a portion of the RMDs into the portfolio as a guard against prematurely depleting assets.

Takeaways

To sum up, similarly aged partners shouldn’t be too concerned about premature asset depletion due to RMDs unless they’re hoping to leave substantial assets behind or one or both partners expects to dramatically exceed average life expectancies. The same is true for singles, whose RMDs are automatically based on a distribution period that’s longer than their life expectancy. That’s especially true if they have substantial non-RMD-subject assets that they can withdraw at a slower pace. They can even use required minimum distributions as a guide to ensuring that their withdrawals tie in with their portfolio’s value and sync up with changes in their life expectancies over time.

On the other hand, married couples with a significant age difference have good reason to be more conservative about their RMD amounts, reinvesting a portion of their withdrawals back into their accounts.

^{Source: Christine Benz, March 4, 2019}