From the March 2007 issue of Research Magazine • Subscribe!

March 1, 2007

Lesson 2, The Uncertainty of Death and Taxes

By now everyone has heard the widely abused clich?, first coined by Benjamin Franklin in 1789: "In this world nothing can be said to be certain, except death and taxes."

Oddly enough, although these two distasteful events are unavoidable, the entire life insurance industry is built on the actual uncertainty of the first of Franklin's certainties -- namely one's date of death. Thus, while death's existence is guaranteed, the randomness of its timing generates a need for risk protection. A family can be left destitute by the premature loss of a breadwinner's life.

Along these lines, I believe that the financial services industry must do a better job of identifying and managing the uncertainty of Franklin's second certainty: income taxes during retirement. The memorable point of this month's lesson is that perhaps the time has come to move taxes to Monte Carlo (in contrast to the people who move to Monte Carlo because of taxes.)

Allow me to explain. If someone were to ask me today to provide a forecast or predict the change in the U.S. Consumer Price Index (CPI) during the year 2020, I would reply with a range between 1 percent and 5 percent, even though it is 13 years away. This is mainly because I have confidence in the Federal Reserve's determination to keep inflation within a reasonable target. Likewise, if I were asked to predict the total return of the S&P 500 index -- or whatever it might be called then -- during the year 2020, I would select 10 percent as my arithmetic expectation plus or minus 20 percentage points in either direction. Indeed, we have over 100 years of data for such financial variables, and despite the short-term volatility I am fairly optimistic about equities in the long run.

Of course, you might disagree with my particular estimates but I'm sure you accept the basic premise that inflation, interest rates and stock returns are random. And, that's why we manage this risk by diversifying, hedging and protecting our retirement income.

However, I would argue that one of the great unexplored sources of risk and uncertainty during retirement lies in our marginal tax rate. If you think about it carefully, this randomness comes from two distinct sources. First, the codified structure of marginal tax rates is up to Congress and politicians. Second, and just as important, I am unsure as to what my particular financial situation and taxable income will look like vis-?-vis this changing schedule.

Currently, a (single) U.S. taxpayer is facing a federal marginal tax rate of 25 percent on income from $31,000 to $74,000; 28 percent on income from $74,000 to $155,000; 33 percent on income from $155,000 to $337,000; and a top rate of 35 percent above $337,000, not including the standard deduction. (All numbers are rounded to the nearest $1,000.)

Remember that a marginal tax bracket measures the income tax you will be paying on the next dollar of income you earn and is distinct from your average tax rate, which is your total tax divided by your income. One of the foundation lessons of economics is that financial and investment decisions should be based on this marginal tax rate and not the average tax rate.

Now, just as a reminder, during the 1970s and 1980s, the top marginal tax rate was over 70 percent. A taxable income of $100,000 resulted in a marginal tax rate of between 55 percent and 60 percent. Yes, we all know that taxes have declined in the last 20 years. But what will the next 20 years look like? How will the U.S. pay for its expensive entitlement programs such as Social Security, Medicare and Medicaid? Can we be certain that someone with an income of $100,000 (in today's dollars) will continue to live in a 28 percent bracket for the next 30 years of retirement? Will entitlements become means-tested, which effectively increases the marginal tax rate?

Sure, we can use the Congressional Budget Office projections, or some other organization's long-term estimates for tax rates. But everyone involved in this process will admit that these long-term tax rates are just estimates. In the words of economists, we face stochastic taxation. More importantly, marginal tax rates exhibit much greater randomness than average tax rates.

So what should we do about it, you ask?

Well, besides accounting for the possibility of random rates in our retirement income Monte Carlo simulations -- which is an important topic in its own right, something I will address in a later lesson -- financial advisors should think about adopting tax diversification and hedging strategies.

This might sound academically abstract, but at the very least this implies that there are a number of cherished rules-of-thumb related to retirement that should be re-examined.

For example, it is well documented that many Americans hold fixed-income bonds in their fully taxable accounts while at the same time holding stocks in their tax-deferred accounts. To most specialists, this is quite puzzling and perhaps irrational. After all, if you want to invest in both stocks and bonds, why not place the highly taxed bonds in the tax-deferred account and the lightly taxed stocks in the taxable account?

A steady stream of commentators continues to lament this example of poor financial hygiene. They argue that since the coupons and income from holding bonds are taxed at your marginal tax rate, they should be sheltered and held inside IRA, 401(k) or 403(b) accounts. In contrast, stocks, ETFs and mutual funds which are taxed more favorably, belong in taxable accounts. In fact, financial economists label this the "pecking order" rule of asset location.

And, while there is obviously much merit in the theory, this so-called puzzling behavior might be a good example of where the masses have got it right. If you think in terms of tax uncertainty, people might implicitly be hedging their retirement bets by gambling on a bit of everything.

Remember, capital losses are difficult if not impossible to use (as deductions) inside of a tax-deferred account, but could be netted against future gains outside of the shelter. And, while nobody invests under the anticipation that they will suffer losses, you might want to differentiate between investments with low volatility that are less likely to experience these unwanted losses versus more volatile holdings (e.g., tech stocks) where the option to use those losses is more valuable. Likewise, the uncertainty of tax rates themselves, your own financial situation and your ability to borrow money if needed might create an incentive to keep some bonds un-sheltered.

Here is another example of tax-uncertainty thinking, courtesy of Professor Bill Reichenstein at Baylor University, who is one of the leading scholars on integrating investments and the tax code. At retirement, if you have both tax-deferred accounts and fully taxable accounts, conventional wisdom dictates that you withdraw or spend from your taxable accounts first and let the tax-deferred accounts grow as long as possible. For the most part this makes sense. However, Professor Reichenstein points out that this strategy of waiting and letting shelters grow, might forcibly push you into a higher tax bracket later in retirement. Instead, a retiree might benefit by diversifying withdrawal sources over time.

Obviously, all of these situations are case-specific and there are no one-size-fits-all solutions. My limited point here is that it's worth thinking about managing your future tax risk, not just your current taxes.

In sum -- not unlike the conventional rules of investment diversification -- the diversification of tax risk implies that you adopt strategies that will benefit under alternative states of nature. Act dynamically. Keep your options open.

----

Moshe A. Milevsky , Ph.D., is a finance professor at the Schulich School of Business at York University and is the executive director o the IFID Centre in Toronto, Canada ; see www.ifid.ca

Reprints Discuss this story
This is where the comments go.