What Should You Put In A Roth IRA?

I was reading a popular investment forum that was discussing the best types of assets to put in a tax shelter like a Roth IRA, and the recommendation that received universal praise was real estate. This is the conventional wisdom, and for good reason. In taxable accounts, real estate investment trust income is taxed at the full ordinary income tax rate.

In other words, a BP dividend in a taxable account will only require the payment of 15% in tax if you file single and make between $37,450 and $413,200 or file jointly and make between $74,900 and $464,850. But if you own something like Realty Income in a taxable account, the dividend tax would be anywhere from 25% to 35%. When you put a REIT in an IRA, you get to avoid those extra points of taxation and receive the full compounding of the asset while it remains in the tax shelter.

This type of tax planning wisdom does contain some merit–if you have $5,000 left to contribute to tax-deferred accounts this year and intend to invest $5,000 into the REIT W.P. Carey (WPC) and $5,000 into AT&T, you’d almost certainly end up with more wealth if you put W.P. Carey into the retirement account and AT&T in the taxable account. But you should note my assumptions for making that assessment: I assume both firms will have nearly identical total return rates, and from there, it is merely a question of tax efficiency.

My aim in challenging part of the conventional wisdom, then, is by questioning the north star that people use to make their allocation decisions. I have noticed a trend in recent tax advisement articles and forums to treat tax efficiency as the objective–it’s not. Your objective should be to go through life looking for the most risk-adjusted post-tax gains. Tax efficiency is a component of this, but really, what you should be after is what you have after the tax payment. I’d rather pay $350,000 in taxes as a necessary condition of turning $100,000 into $1.3 million than pay $50,000 in taxes to turn $100,000 into $900,000.

With this in mind, I would like to challenge the assumption that C-corp dividend stocks with low initial yields and high dividend growth rates like Colgate-Palmolive have no business being bought in something like a Roth IRA.

The advantages of owning something like Colgate Palmolive in a Roth IRA for someone that plans for decades are manifold: (1) because retirement money are usually prized for their safety, you can be sure that the enterprise will still be profitable years in the future; (2) as a related matter, if you are reinvesting the dividends, you are escalating your commitment to the stock, and you want to make sure that the chosen firm is of worthy quality rather than being the best choice due to perceived short-term tax efficiency; (3) tax loss harvesting isn’t permitted in retirement accounts, so you want to make sure you are owning something with a remote possibility of permanent capital destruction; (4) the superior earnings per share growth eventually reach a point where you would want to avoid hefty tax payments from capital growth over the decades; (5) the yield-on-cost would eventually reach such a high threshold that avoiding taxes on a “rocketship asset” would be desirable; (6) tax-advantaged accounts have contribution limits, and capital appreciation within the account can be converted into tax-free income at a later date; (7) and most importantly, you will end up with the greatest amount of post-tax wealth by focusing on net returns rather than maximum tax efficiency.

Colgate has the natural power to expand into new countries, roll out new cleaning and dental products, buy back stock, raise prices in excess of inflation, and develop a niche of high end products that generate very high profit margins. Roth IRAs didn’t exist back in 1970, but if it did, you would have been able to see a $5,500 investment grow into $8.7 million (that’s what happens when you compound at 16.5% for 45 years straight.) I’d reassess future expectations for Colgate down to the 12% range, but it’s still above some of the REITs I see people stuffing into their accounts in the name of tax efficiency in the pursuit of 8% annual gains.

If you want to maintain rationality in your capital allocation decisionmaking, you should ask yourself the question: “Which stocks do I think have the best combination of value, earnings growth, and dividends?” Then, you eliminate the ones with business models or risks that you don’t want on your household’s balance sheet. Then, you figure out what types of tax-deferred accounts are at your disposal, and select the benefits that will give you those best returns once you adjust for the tax benefits. If you find a REIT poised to deliver 13% annual returns, sure, you can pick it over Colgate in an IRA.

I do understand the impulse to realize tax savings. That’s why Royal Dutch Shell is the very first stock I covered when I started this site. I called it the perfect Roth IRA stock. That statement is even more true now. It grows earnings around 5% to 6% over the decades. It currently offers a 7.5% dividend. It’s within the realm of possibilities that it will return 12% to 13% from here, in line with past long-term historical performance. Someone that tucks Royal Dutch Shell into a retirement account for forty years will receive 160 untaxed dividend payments over the course of a lifetime.

Investments that are taxed at ordinary income tax rates, and REITs are the most mainstream example of this, do enjoy significant benefits from being placed inside of a tax-advantaged account. But you should keep in mind that this advice from tax professionals can be skewed by (1) assumed short-termism in which retail investors are talked to as if they never own a stock for more than thirty months, and (2) making an altar out of tax efficiency rather than the best risk-adjusted post-tax returns.

Instead of making decisions based on how much money you might save in tax payments, you should think in terms of opportunity costs. Your question should be “What’s the best way I can compound my money to get the highest end result?” rather than “What’s the lowest amount of money I can pay the tax man?” That second question is important, but it should be regarded as a subset of the first question rather than an objective in its own right.