Does it always make sense to set up or contribute to (or roll over to) a Roth in today’s tax environment?
When the SECURE Act first passed the House of Representatives by an overwhelming majority, advisors in the CPA community performed an immediate tax analysis and concluded that the provisions tilted the playing field in favor of Roth conversions. They focused on the provision requiring heirs to fully distribute out their inherited IRAs in ten years. If the IRAs were large (as they are for many financial planning clients) even distributing the proceeds out evenly over ten years would likely push the heirs into higher tax brackets, raising Uncle Sam’s percentage of the distributions. If the IRA assets were converted to Roths, especially with partial conversions in the low-income years after clients retire but before they are required to take RMDs (so the thinking went) the family could get a relative bargain on tax rates.
If the tax rate is lower for the client making the conversion than it would be for the heirs taking distributions over ten years, the logic is unassailable. But Phil DeMuth, of Conservative Wealth Management in Los Angeles, CA, thinks that there is a better way to manage the tax consequences now that the SECURE Act is becoming the law of the land.
DeMuth is the author of nine books intended for a personal finance audience, and he has been a prolific writer about investing issues. He agrees with the CPA community’s premise that the Act is going to add tax costs to families with large IRAs; in fact, he thinks most of us are underestimating the impact. “The bill forces kids into a much higher tax bracket when they liquidate,” he says. “Most kids, when their parents die, are going to be in their peak earning years. The money comes to them at a horrible time; it is effectively like another estate tax, where people are suddenly thrown into the 33% tax bracket, and the government’s share of the IRA is greatly increased.
Want proof? Downloading the Congressional Budget Office revenue estimates for eliminating stretch IRAs, and extending the assumptions out to 2050, DeMuth found that the government would collect $4 billion of extra tax revenues every year straight out of the pockets of IRA owners, above what it would have collected without the SECURE Act provision. “This is a huge matter, taxwise,” he says. “And it was passed with no discussion, no debate and very little public notice that any of this was even going on.”
Rather than accelerating money into Roth IRAs, what can investors (and their planners) do to mitigate this tax grab? DeMuth thinks that advisors are overlooking some possibilities with plain ordinary taxable accounts.
You can see his logic in a very simple chart, reproduced here, where DeMuth compares a traditional IRA, from contribution to distribution, with a Roth IRA and a taxable account that invests in a portfolio of zero-dividend stocks. This comparison is illustrated where the tax rate is the same when the contributions are made as they are when the money comes out in distributions (top part of the chart), and then when there are disparities in rates: first when the tax rate when contributions are made is higher than when distributions come out (middle of the chart), and vice versa, when the tax rate when the client makes contributions is lower than the tax rate at the time of distributions (bottom).
Why is the taxable account invested in non-dividend-paying stocks? DeMuth argues that this is the most tax-efficient way to accumulate wealth in a taxable portfolio. He explains the logic neatly on his website: “Nearly all investments generate a cash flow, leading taxes to gnaw away at your total returns. The money lost to taxes never compounds to produce greater wealth for you. The financial services industry has set up a conveyor belt that feeds your investment returns to the Internal Revenue Service.”
In California, he notes, high-bracket investors pay a 36.1% tax on dividends and capital gains year after year. The goal of investing in stocks that don’t pay dividends is to take the government’s hand out of your clients’ pockets as much as possible.
DeMuth acknowledges that there are investment risks to limiting yourself to only stocks that don’t pay dividends, but he believes that you can build a diversified portfolio that fully participates in the world’s economic growth if you’re careful and somewhat scientific about it. Since you’re buying stocks rather than funds, you can engage in yearly tax-loss harvesting, and using the losses to offset gains, continually raising the cost basis of your overall portfolio, and perhaps leaving a handful of highly-appreciated positions to leave to your heirs—and achieve a step-up in basis.
But what does this have to do with the SECURE Act? DeMuth cites research (you can find it here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=794144) showing that an investor who follows this approach can liquidate 3% to 7% a year from this taxable portfolio tax-free, assuming he/she follows the same process of harvesting losses and offsetting gains. This answers a question that many readers were probably asking themselves when they looked at DeMuth’s tables: didn’t you forget to take into account the capital gains from the taxable portfolio when you liquidated the taxable portfolio? The paper shows how, by harvesting capital losses on a regular basis, these can be entirely or largely finessed away over at least the first half of retirement, and potentially for the duration of it.
So the basic proposal here is somewhat shocking: instead of contributing to an IRA, 401(k) plan or Roth IRA, the client would invest instead in a plain old taxable portfolio that is very tightly tax-managed with non-dividend-paying stocks. Instead of taking required minimum distributions every year, or paying additional taxes to make Roth conversions, the client would take entirely or nearly tax-free withdrawals from the taxable portfolio each year. If there are capital gains taxes to be paid, they would be paid at a relatively low tax bracket.
And at death, instead of leaving a traditional IRA to heirs, with accelerated mandatory withdrawals coming out as ordinary income, clients would leave a taxable portfolio with a full step-up in basis. Any distributions that came out the very next day would be tax-free—and if the portfolio is managed in roughly the same way it was before the client’s death, the tax grab is deferred and mitigated along the way.
“By lowering the future value of retirement accounts,” says DeMuth, “I think the SECURE Act would make the taxable brokerage account more desirable for anyone who does not plan to consume their retirement plans during their lifetime. They will certainly be more difficult for the government to expropriate via confiscatory taxation.”
Let’s concede that investing in individual stocks (and especially selecting from the non-dividend-paying subset) represents more work than investing in ETFs and index funds. That work is becoming easier thanks to increasingly powerful automation (see: http://www.bobveres.com/portfolio-management/high-tech-investing/), but DeMuth’s approach is inarguably more labor-intensive than the way most advisory firms are investing today.
With that caveat, is there anything else DeMuth doesn’t like about the SECURE Act?
He REALLY doesn’t approve of the mandate that all 401(k) plans should let their participants check a box and get a quote from the annuity industry on how much income their current balance would translate into.
“I don’t understand why the insurance industry should be given special privileges in terms of their ability to advertise to everybody in the United States who has a retirement plan,” says DeMuth. “Why shouldn’t Fidelity Investments also have a box, where you can send all your money to Fidelity and they’ll manage it. I have a little advisory firm. If the IRS told me that I could have a box to check, and that would mean I would manage America’s financial retirement assets, I’d be happy to scale up and do it. Just check a box and send your money to me, and I’ll manage it.”
The logic behind that provision of the SECURE Act is that plan participants are not good at translating their portfolio balances into monthly or annual paychecks, due to cognitive shortcuts and behavioral finance issues. But DeMuth plausibly argues that these shortcuts and issues cut both ways. “People check the box so they get $5,000 a month for life, and that sounds like a good idea,” he says. “But they don’t realize that in 20 years, that is going to spend like $2,000 a month, because of inflation. They don’t understand, hey, the house needs a new roof, they can’t just write a check for more money to pay for the new roof. When the money is in an annuity, they can’t get that lump sum when they need it.”
Bigger picture, DeMuth wonders why the financial planning community is just taking these legislative initiatives lying down. “I have written every one of my clients,” he says. “I sent them an email, I sent them an article I’d written, I sent them bullet points, and I sent them the links to each of their Senators, saying: please tell them to keep their hands off their retirement assets. I don’t see why the RIAs in this country, who have all the high-net-worth clients who are affected, are not speaking out on this and other public policy issues.”
Whatever you think of DeMuth’s suggesting that taxable accounts could become superior to Roth and traditional IRA accounts, this, here, is a darned good question.