Many readers of this blog contact me directly with questions and comments. While often the responses are very specific to a particular circumstance, occasionally the subject matter is general enough that it might be of interest to others as well. Accordingly, I will occasionally post a new "MailBag" article, presenting the question or comment (on a strictly anonymous basis!) and my response, in the hopes that the discussion may be useful food for thought.

In this week's mailbag, we look at two recent inquiries: 1) whether or not it's a good deal to use 1031 real estate exchanges to avoid the new 3.8% Medicare surtax on investment income that begins in 2013; and 2) some thoughts on the recent Center for Retirement Research brief about using the RMD method as a retirement income/withdrawal strategy. 

Question/Comment: The article on the 3.8% Medicare Tax quoting you is absolutely correct. However, one 3.8% tax avoidance strategy that is not being mentioned is the deferral of gain by the use of a IRC 1031 exchange.

Using a 1031 exchange (for like-kind exchanges of real estate) is certainly a valid option for deferring the new Medicare tax, but strictly speaking it is merely a tax deferral strategy, not a tax avoidance strategy. The looming Medicare tax remains if a 1031 exchange occurs. In fact, all else being equal, it increases potential exposure to the tax in the future, as the income thresholds are not indexed for inflation, and those who may not fully be exposed to the Medicare surtax now will only become increasingly exposed in the future, the longer they wait and defer taxes (for those who are high income, of course, it won’t matter, as they may always be over the line for the foreseeable future). In some situations, it may be possible to defer the real estate gains into a lower income year and thereby reduce exposure to the tax, but only if the property can subsequently be sold (at a still reasonable price) to time the transaction in the desired year.

Of course, in reality such strategies can actually be accomplished this year, by NOT doing 1031 exchanges and instead deliberately recognizing the capital gains this year, before both the potential rise in capital gains rates themselves and the imposition of the new Medicare tax as well. In general, such “capital gains harvesting” strategies are very appealing in rising tax rate environments as I've written in the past, and can actually result in more wealth than deferring gains. Notably, the capital gains harvesting strategy is particularly appealing in a real estate context with rising tax rates, as not only can the investor avoid higher capital gains taxes and avoid the Medicare tax permanently, but the investor can also reset cost basis to the current fair market value and claim fresh depreciation deductions, which will have greater economic value in the future if/when/as ordinary income tax rates rise (especially for the most affluent real estate investors, who have the greatest risk of rising tax rates).

Once we’re past 2012 and the new Medicare tax is here, though, the capital gains harvesting strategy is generally less effective, as the new tax will then be in full force. At that point, the tax rate environment is ostensibly flat again (at least until Congress does who-knows-what next!), and the value of tax deferral (including 1031 exchanges) returns, both to defer capital gains taxes themselves and the new Medicare tax.

But the bottom line is simply to note that tax deferral itself is generally not a tax avoidance strategy, but simply a deferral strategy. That doesn’t make it bad – in fact, all else being equal, tax deferral is good, which is why we so often pursue it. But it is nonetheless notable that with a looming tax rate increase, in the immediate near term tax deferral actually can be wealth destructive, too!

Question/Comment: I just read an Investment News story on the Center for Retirement Research's study indicating that using the RMD as a guide for withdrawals is better than the 4% rule.  Have you had a chance to read this report?  If so, your thoughts?

I did read the Center for Retirement Research (CRR) brief on the RMD method. But unfortunately, while I’m generally a very big fan of the CRR research briefs, I don't think they did a good job with this one. 

The primary issue is simply that they appear to have assumed ZERO volatility in their analysis, which clearly is unrealistic. And the moment you include volatility, the RMD methodology breaks down. Remember, the RMD method recalculates every year, which means any volatility in the portfolio becomes volatility in spending. So in bear markets like 2000-2002 and 2008-2009, when a balanced portfolio has a 20%+ decline, that means the client has to take a whopping 20%+ spending cut! Many people can’t handle that at all, or at least not without a major lifestyle change – which is all the more brutal when the market recovers a year or two later, after people have already changed their lifestyle!

In the real world, using the CRR’s RMD methodology would lead to ridiculous results like telling a client in March of 2009 “sorry, you can’t afford your house anymore, you have to move” and then a year later saying to the same client “well, the market recovered, I guess you can go buy your old house back now! I hope the new owners will give it back to you.” It’s clearly impractical and unrealistic, especially when the safe withdrawal rate research has already shown us how to craft safe and STABLE spending paths for clients that DON’T require this kind of absurd spending volatility.

It’s also notable that the CRR brief is using the baseline safe withdrawal rate methodology from the research that came out almost 20 years ago, and NONE of the enhancements to the methodology that have been written about over the past decade or so, including material by Guyton, Klinger, Blanchett, and more, where you increase spending using guardrail approaches, ongoing mortality adjustments, updated Monte Carlo success rates, etc. If the CRR brief used some of the more updated methodologies, the safe withdrawal rate and RMD approaches would have come out very similar in their research (and the safe withdrawal rate would come out superior once volatility is accounted for). In fact, my Weekend Reading column just this past weekend highlighted an article by Blanchett showing how the "old" constant dollar safe withdrawal rate approach is inferior to the RMD method, but a Monte Carlo updating and/or mortality updating method is superior. 

But the bottom line is that practitioners rejected the RMD methodology as being radically unstable and impractical a long time ago, and the safe withdrawal rate methodologies have gone far beyond the outmoded version the CRR brief used. In fact, some of the most robust and successful safe withdrawal rate methodologies out there today are the ones that have built RMD-style mortality adjustments into the existing safe withdrawal rate model - basically, taking the best aspects of the RMD model into account but leaving the rest of the problems behind!


  • Wade Pfau


    Good to hear your thoughts about the CRR article. It does have a lot of similarities with the recent article by Blanchett.

    Just a couple ideas...

    First, in my view, the RMD approach is mortality updating. Perhaps not precisely, but my understanding of it is that the point is to spend based on your continued survival and updated life expectancy.

    Unlike the Blanchett article though, it is not 'constant probability of failure' updating.

    Also a pure RMD approach may have you spending too little early in retirement.

    About the variable income, that is absolutely right. It can never run out of money, but income is completely unpredictable and that is a big problem.

    On the other hand, the guardrails and decision rules provide more stable income because leave open the possibility of portfolio failure.

    There needs to be some balance here.

    I know you are not a big fan of this, but one related type of approach I tests was to annuitize basic needs and then use an RMD approach to get additional income for discretionary expenses.

    • Michael Kitces

      Ironically, my fundamental issue with the RMD approach is that it simply provides a far-too-variable and uncertain floor. As I've written in the past, I consider a safe withdrawal rate approach a flooring approach in application, as are some annuity approaches. The RMD method simply doesn't appear viable without some kind of floor tied to it.

      I suppose there's something to be said for a more hybridized approach as you suggest, except that as I've commented before, I'm very skeptical that most clients in the real world can effectively separate discretionary and non-discretionary goals - and more directly, that if their non-discretionary goals that are still important to them are that unsecured, that they'll often just keep working/saving (if able) until they ARE secure, which again obviates the usefulness of the RMD approach.
      - Michael

  • Avi Oren

    It's interesting that the AAII in it's magazine of this month, published the CRR study as written by its authors. Mike's analysis of the study (see MailBag) and his comments that the report neglected to consider the last decade of progress in withdrawal strategies (Bill Bengen's name is missing) in its recommendations, is another proof as to the value of advice from smart financial planners. IMO, the AAII is misleading retirees by just publishing this CRR report without adding such considerations as Mike mentioned.

    Avi Oren, retired (ex) CFP.

Blog Updates by Email

@MichaelKitces Twitter

Out and About

Wednesday, April 23rd, 2014

Understanding the New World of Health Insurance Evaluating Existing Annuities: What Every Adviser Needs to Know @ FPA New Jersey

Thursday, April 24th, 2014

Keynote @ Shareholders Service Group

Monday, April 28th, 2014

Safe Withdrawal Rates: Mechanics, Uses & Caveats @ St. Louis Estate Planning Council