Tax-Efficient Asset Location
Introduction
Often when new clients come to me with portfolios that have been managed by other advisors, I am aghast when I see what they have endured from my competitors. These include indefensibly high advisory fees, opaque kick-backs from mutual fund companies and custodial firms, and poorly understood commission-based “products” that saddle clients with onerous costs for years. The conflicts of interest in the retail financial advice-giving world are rife.
I also notice some less obvious problems that indicate a lack of attention or understanding. One of these has to do with “asset location” or “asset placement” as it is sometimes called. The basic idea is simple: minimize long-term taxes by placing the least tax-efficient assets in tax-deferred accounts (IRAs and 401Ks) and the most tax-efficient assets in taxable accounts. It appears that this basic principle of responsible wealth management is seldom practiced, based on what I have seen from incoming accounts.
The Definition of Tax-Efficient Asset Location
William Reichenstein, CFA, is probably the leading academic authority on the subject of tax-efficient asset location. In one of his many published articles, he defines optimal asset location as follows:
“Individuals should locate lightly taxed securities in taxable accounts and heavily taxed securities in tax-advantaged retirement accounts to the highest degree possible, while maintaining their risk–return preference.”
The most “lightly taxed” or most tax-efficient securities for most U.S. investors are municipal bonds and stocks. As discussed in a previous article, municipal bonds are likely to be attractive only to taxpayers in the highest tax brackets. For most investors, that leaves stocks as the most tax-efficient investment.
However, stocks are not always tax-efficient. It depends on how they are managed. Actively managed mutual funds tend to be highly tax inefficient because of their high turnover. Passively managed (index) funds, especially exchange-traded funds (ETFs), are usually quite tax-efficient, on the other hand. (Because of their tax-advantaged structure, ETFs are often able to avoid distributing capital gains to their investors even in a rising market.) Individually-owned stocks and passively-managed stock funds share the following tax-minimizing characteristics:
- Capital losses can be realized to offset capital gains (or to offset up to $3000 of ordinary income).
- The realization of capital gains can be controlled and may be delayed indefinitely.
- If gains are never realized, inheriting beneficiaries can take advantage of a “step-up in basis” thereby eliminating the capital gain altogether.
- Realized long-term capital gains (on stocks held longer than 12 months) are taxed at preferential rates of 15% or 20%.
- Dividends paid on most stocks are taxed at the same preferential rates as long-term capital gains (15% or 20%).
By comparison, interest income from bonds and certain “non-qualified” stock dividends (from REITs, MLPs, BDCs, and some non-U.S. stocks) are taxed at “ordinary income” tax rates, which can be as high as 37%. This makes them heavily taxed securities.
The wider the spread between lightly taxed securities and heavily taxed securities, the more important asset placement becomes. Thus, according to Reichenstein, “asset-location strategy is most important to high-income taxpayers who manage stocks passively.”
Harvesting the tax advantages available to owners of stocks requires a strategy of very low turnover. That excludes most actively managed funds. In fact, the tax benefits of passive management itself probably outweigh the benefit of smart asset location. For example, Vanguard research indicates that “[actively managed] domestic stock funds lost about 1 percentage point annually, on average, to taxes” over a 15 year period. And there is ample evidence that index funds outperform actively managed funds even before taxes are considered. Including taxes in the comparison makes passive management even more compelling.
Why It Matters
Astute asset location can add a lot of value. According to Asset Location Methodology, a paper published by Betterment: “Asset location is widely regarded as the closest thing there is to a “free lunch” in the wealth management industry.” An often-cited paper in the Journal of Finance says that “the asset location decision is “crucial to the wealth accumulation and welfare of investors over their lifetimes.” According to Michael Kitces, a widely-followed financial planning blogger, speaker, and commentator, “recent research has shown effective asset location strategies can add 20-50+ basis points of "free" value to annual returns.” Vanguard uses an even higher estimate of the value of asset location: “75 basis points (bps) of additional return in the first year.” That’s a lot, and it’s worth going to some trouble to get it.
Asset Location in Practice
Most of the experts agree with the basic principle of allocating passively managed stock funds to taxable accounts and bond funds (other than municipal bonds) to tax-deferred accounts, such as IRAs and 401Ks. In fact, if a client is willing to borrow, then “the optimal asset location policy involves allocating the entire tax-deferred account to taxable bonds,” and all of the taxable account to stocks, even if the taxable account is less than 100% of the amount that would ideally be allocated to stocks, according to the paper in the Journal of Finance. Investors would then either borrow or lend in the taxable account to achieve their desired asset allocation.
Most investors are not comfortable with the idea of borrowing to implement an ideal asset placement strategy. Barring that, the next most optimal strategy is to allocate taxable bonds first to tax-deferred accounts and stocks first to taxable accounts, and then mix taxable bonds and stocks as necessary in one or the other to achieve the target asset mix. Thus, optimally, investors would either have minimal stocks in their tax-deferred accounts but no taxable bonds in their taxable accounts, or minimal taxable bonds in their taxable accounts but no stocks in their tax-deferred accounts.
This is not, however, what actually seems to be happening in practice. One study reports that 48.3% of investors who own taxable bonds in taxable accounts also own stocks in tax-deferred accounts and that 41.6% of investors who own stocks in tax-deferred accounts also own taxable bonds in taxable accounts. Other studies find that a large proportion of investors have substantially more stocks in their tax-deferred accounts than in their taxable accounts!
Some Math
Some investors and even some financial advisors (who should know better) operate under the mistaken impression that stocks should be allocated to tax-deferred accounts because their higher expected returns will allow them to compound in a tax-sheltered manner. While it may be optimal to allocate stocks to tax-free accounts (Roth IRAs and Roth 401Ks) because of their higher expected returns, it is far less likely that allocating stocks to tax-deferred accounts (Traditional IRAs and Traditional 401Ks) will be optimal. The time period required for that to happen tends to be too long.
While the rule of thumb described above (stocks in taxable accounts, taxable bonds in tax-deferred accounts) will apply in most circumstances, it is possible to mathematically show that in extreme circumstances it may be violated. For example, for extremely long time periods, the opposite allocation rule will be optimal.
The graph below shows the after-tax cumulative values of a $100,000 investment in stocks held in either a taxable account or a tax-deferred account using the following assumptions:
Rate of
Return Tax Rate % of Total Tax %
Dividends 1.5% 15% 18% 2.6%
Capital Gains 7.0% 9%* 82% 7.3%
Weighted Avg 8.5% 100% 10.0%
MTR in Retirement (applicable to RMDs from IRAs) 24.0%
*Assumes an average holding period of 20 years, or a turnover rate of 5% per year, taxed at a rate of 15%.
For these calculations, I am grateful to Chris Reed, who provides links to an Excel workbook with many calculations in his paper “Rethinking Asset Location.”
As shown, the after-tax value of these two investments shows that, unless the expected holding period is extremely long (about 35 years in this example), the lower tax rate applied to owning stocks in a taxable account makes it the preferred placement for stocks.
Of course, the example shown in the graph below is derived from the assumptions above. The most sensitive assumptions are 1) the turnover rate (5% per year is assumed in the example) and 2) the marginal tax rate in retirement (24% is assumed in the example). Higher rates of turnover would drive the capital gains tax rate upwards. That would make holding stocks in tax-deferred accounts a bit more attractive. The expected holding period at which point it becomes more attractive to own stocks in a tax-deferred account rather than in a taxable account shortens to about 23 years if the tax rate for stocks rises from 10% to 15%. Likewise, lower marginal tax rates would make holding stocks in tax-deferred accounts somewhat more attractive. The expected holding period at which point it becomes more attractive to own stocks in a tax-deferred account rather than in a taxable account shortens to about 30 year at a marginal tax rate of 22%.
On the other hand, lower stock turnover rates and/or higher marginal tax rates in retirement would lengthen the crossover time period and make holding stocks in taxable accounts even more compelling. Even assuming that capital gains tax rate is always 15% (i.e., 100% of stocks are sold after holding them for a year and a day) and that the marginal tax rate in retirement is 22%, the expected holding period for an IRA would have to exceed 20 years in order for holdings stocks within the IRA to become more attractive than holding them in a taxable account. Very few wealth management clients will have a marginal tax rate lower than 22%. In 2021, the 22% tax bracket starts at an income of $81,051 for married taxpayers filing jointly and at $40,526 for single taxpayers.
Most of my clients are in or near retirement. Therefore, their time horizon is not long enough (35+ years) to make investing stocks in an IRA rather than in a taxable account an optimal choice. Visually, it is difficult to tell just how sub-optimal that would be from the graph above. For example, if the IRA would be liquidated in 10 years, preferring the taxable account would result in a 22% higher after-tax balance. After 15 years, the difference is 17%. Even after 20 years, the taxable account is still 12% higher. And bear in mind, if the average length of retirement is 30 years, and required minimum distributions will force the liquidation of IRAs over that time, the average time horizon for a dollar in an IRA is not much more than 15 years.
Summary and Conclusions
- Experts overwhelmingly recommend allocating tax-efficient stocks first to taxable accounts and taxable bonds first to tax-exempt accounts (IRAs and 401Ks).
- This becomes more important the higher the spread between the preferential tax rates on stock dividends and long-term capital gains (15% or 20%) and marginal tax rates (up to 37%). Thus, asset location matters even more for higher-income taxpayers.
- This assumes that stock investments are tax-efficient, i.e., with very low turnover.
- Asset placement has been described as a “free lunch” in wealth management, adding .20% to .75% in additional after-tax return right from the first year.
- The amazing thing is how rarely tax-smart asset placement is implemented, even by financial advisors who should know better.