December 1, 2009
But, assuming he has an estate goal, to model this we will bump up his savings to an $8,500 Roth contribution, a lifestyle-equivalent $12,274 tax deferred contribution, and we will add a goal of a $1 million estate in today’s dollars ($6.4 million in actual dollars). This gives the Roth 78% confidence of exceeding his income and estate goals after income taxes (but before estate taxes).
Under this scenario, we can be 86% confident that the traditional 401(k) would exceed his $1,000,000 estate goal, but this is before income taxes on the amount remaining in tax-deferred assets. For the traditional 401(k), the 78th percentile outcome leaves his estate with $615,489 in taxable assets1 with a stepped-up basis and $850,878 in tax-deferred assets that would be subject to ordinary income tax. If we assume the beneficiaries immediately pay the income tax on this, their tax rate would have to be about 55% on the tax-deferred assets to be at a disadvantage to the Roth. Tax rates could go this high, and if they did the Roth conversion would look attractive. Of course, most people would apply some common sense and would opt to stretch out the tax bite using either the five year deferral option or the lifetime distribution option.
Within the distribution of outcomes, there are advantages to Roth for this combined income and estate goal circumstance. For example, at the 50th percentile outcome, the Roth would leave an estate that is 2.8 times his estate goal (assuming, of course, that the client wouldn’t use the extra two million dollars for his lifestyle). The tax-deferred account would leave an estate that is 3.3 times his estate goal, but a significant amount of that would be in tax deferred balances with income taxes still due. At the 50th percentile, the breakeven tax rate on the tax-deferred assets left for the heirs would be only 23% to equal the wealth of the tax-exempt Roth. Of course we have no idea what the tax rate will be for heirs or how long they would defer or stretch out the distributions. We also don’t know whether the client would end up spending more than we are assuming if he had a couple extra million dollars lying around, completely changing the analysis and our future advice.
What we do know is that as confidence increases (lower investment return scenarios) deferring taxation in the traditional 401(k) or IRA makes it reasonably unlikely the future tax bite heirs will pay will be high enough to justify the Roth conversion (i.e. 55% tax rate at the 78th percentile). We also know that if returns are higher and investors are thus significantly exceeding their goals, it becomes increasingly likely that the Roth tax treatment would have a wealth advantage over tax-deferred treatment, albeit at levels of wealth that are already at multiples of the client’s goals (for example, at the 50th percentile heirs’ income taxes at more than 23% would show an advantage to the Roth but the estate would already be 2.8 times more than desired).
This should dispel some of the common claims made about the advantages of Roth. It just isn’t as easy as assuming a flat tax rate on all scenarios, as taxes will vary from year to year based on uncertain market returns. The higher returns are, the more likely there will be a wealth advantage to the Roth but also less confidence one can have in achieving those higher returns.
What if you have accumulated significant tax-deferred assets?
The industry paper I reviewed showed another example that is perhaps more germane to the decision many high earners now face: whether they should convert now after already having accumulated significant tax-deferred assets.
The paper showed several examples to make the case that if one has enough taxable assets to pay for the conversion, there would be a high probability of an advantage to the Roth. In fact, the paper claimed that if tax rates don’t decline in retirement for someone with $1 million in tax-deferred assets and $450,000 in taxable assets that they use to pay for the conversion then that person would have a 97% probability of benefiting from the switch.
Silly assumptions in the calculations caused this ridiculous conclusion.
Replicating the analysis in our system, I take the same 65-year-old couple that is faced with making this decision, the same 20-year time horizon (they used normal life expectancy), the same present account values and even the same 60/40 asset allocation assumption. Instead of assuming the same flat tax rate regardless of the market simulations, I used our dynamic tax assumptions that not only calculate the differing year by year tax rates and forced RMDs as previously mentioned but also assumes postponing withdrawals from tax deferred assets until RMDs kick in or there is a need to make tax deferred withdrawals to meet spending goals.. This is more akin to how we would actually manage our own clients’ affairs. To keep their tax rates on the IRA distributions high, I included a taxable retirement pension income of $60,000 starting at age 65 and assumed an after-tax inflation-adjusted spending need of $130,000. (In our process, we are concerned with the lifestyle of what our clients can actually spend, so our engine forces larger distributions and taxes them sufficiently to meet the net spending desired.) Our assumption of the $130,000 spending need is what is materially different from the erroneous analysis of the paper.
1 Taxable assets were accumulated by forced RMDs above his income need that were taxed at ordinary income tax rates and then invested in a taxable account with 20% turnover and 20% short term capital gains rates, about twice what we would safely assume with our portfolio management approach.
Would you like to send this article to a friend?Remember, if you have a question or comment, send it to .