My Theory of the "Two Retirements"

I have had, over the last two or three years, an ongoing "discussion" with a friend about whether I am retired: her-yes, me-no.  She's more or less right of course, but not just because she tends to be right.  On the other hand I don't think I have ever successfully gotten my idea (i.e., "not retired") articulated exactly right or understood very well, and again, it is not just because she tends to be right, which she usually is. 
The thing is, it's probably more about the simple process of the ideation of one idea in two different minds and the subsequent weakness of language to bridge the gap once the ideas have formed, diverged and settled. That, and maybe the conservative impulse to protect one's settled worldview.  It is also probably a little bit about the (innate?) differences between what I'll call "A thinking" and "B thinking" where A is kind of a "binary or black and white or 2D or common sense or logical or category X&Y" type of thing.  B, on the other hand, is more grey tones (Shane Parrish at FarnamStreet Blog calls it "grey thinking") or non-linear or 3D or continuums or even sometimes some pretty paradoxical stuff.  Think of Special Relativity for a second and you'll see what I mean.  Both truth and paradox can coexist. Some people are not comfortable that two contradictory things can both be true at the same time (or neither true, I guess, or maybe just one thing that is both true and not true) or that there may be a constant greyness in things depending on the day or the person or the point of view; Me? I'm ok with it. B thinking.  

"…truth always lies somewhere in between, and the discomfort of being uncertain [or even contradictory] is preferable to the certainty of being wrong." -- Shane Parrish on 'grey thinking'

 “How wonderful that we have met with a paradox. Now we have some hope of making progress.”
― Niels Bohr

 So yes, I'm retired (but yet, not really).  It's A vs. B and A seems to win…but then there's B again.  So why is B on my mind? Probably because I've had this theory that there are two retirements (early retirement vs. traditional or, in my case, what I call not-retired vs. retired). When I have tested this concept out in conversation, quite a few folks seem to disagree with me and usually they do it pretty quickly. They tell me that it's all the same, or one thing, that there are just slightly varying degrees of risk, which is completely true of course.  Except that my personal experience of early retirement (ER) seems sufficiently different from a traditional retirement in enough ways that it makes it hard for me to square my experience and my theory of "the two" with all the head-patting "isn't that cute that you've over-thought this thing" stuff.  ER seems harder, longer, riskier, less certain, less safety-netted, scarier…by a lot.  That and I seem to have been working at it full time so it doesn't yet feel quite so retire-ish to me.  While this point of view might be true too of course, it's not clear that I've demonstrably or convincingly added anything here to the conversation.

On the other hand, at least I've had a little bit of corroboration on this stuff lately.  First, I was reading a paper by David Blanchett the other day (SimpleFormulas to Implement Complex Withdrawal Strategies, 2013).  In the paper he attempts to create some simple formulas for withdrawal rates using a base of more complex underlying analysis. The theory here, an intuition well understood and beginning to be well quantified in the retirement finance world, is that updating retirement income strategies over time will generally improve outcomes vs. a static head-in-the-sand approach.  The theory, unfortunately, is also that, as Blanchett states it: "many financial planners and engaged retirees may find a dynamic withdrawal strategy difficult or impossible to implement given the sometimes complex software, tools, or processes that are needed to adjust portfolio withdrawal amounts at some regular interval."  So, Blanchett leans on the underlying "relative efficiency of a far more complex methodology" to create a couple of much more simple and usable "paper equations [that] capture 99.9 percent of the relative efficiency" of that other more complex model -- they are working proxies in other words.  My emphasis here is on "two," by the way, because just one equation didn't cut it for David. That's because he found two retirements in the data and that "two" helps me make my point.  Here is Blanchett again:

"After running different types of regressions over different periods, two equations (or models) were created that can be used to approximate the initial portfolio withdrawal percentage over two different periods, for either periods greater than 15 years or less than 15 years. Equation 1 is optimized for retirement periods that are 15 years or greater, where the withdrawal rate is determined given a particular target distribution period (Years), equity allocation (Equity%), target probability of success (PoS), and fees (Alpha). Equation 1 is called the dynamic formula…. The formula does a very good job of approximating the withdrawal rate that would be determined through solving a Monte Carlo simulation for the target parameters."

…For periods shorter than 15 years, using equation 2 is suggested. Equation 2 is called the RMD approach, and it is based on the IRS’ required minimum distribution (RMD) rule. The following variables have been excluded from equation 2 because they do not materially improve the explanatory power of the model: (1) equity allocation (Equity%) (2) target probability of success (PoS) and (3) fees (Alpha)… although the RMD approach results in an estimated withdrawal that is relatively similar to the actual withdrawal for periods of 15 years or less (R² of 95.10 percent), it is materially less useful for periods of 15 years or greater, with an R² of 45.76 percent.

So right there in my hands I finally found explicit, visible math from a leading researcher that splits retirement into two pieces: long and short, early and traditional, complex and simple, Equation 1 and 2.  This should match intuition, too, if you think about it enough…except that it is still pretty hard to convey sometimes.  For example, I had a conversation once with a recently retired 69 year old, a guy with past and future heart disease, a more than decent pile of assets, and very modest longevity expectations…say 10 years or so, maybe more...hopefully more.  He looked at me with my retirement anxiety and my theory of two retirements like I had two heads if not three (in this case we were talking about a retirement starting around age 50 with a longevity expectation that could stretch at that time beyond 40 years, maybe as far as 50, and assets that might or might not put one on a razor's edge given the unforgiving math of a long time-frame) . For him, it was all simple, easy and riskless -- which it probably was.  I, for my part, while I thought I understood his view of the world, I could not satisfyingly explain my theory of "the two" to him...or the feeling of risk and complexity.  We clearly had two totally different sensibilities about retirements but I had no good language to bridge the difference (or at least from me to him).  In retrospect I now know I was equation 1 and he was 2. 

The second written version of this idea that I ran across recently was in Darrow Kirkpatrick's new book "Can I Retire Yet? How to Make the Biggest Financial Decision of the Rest of Your Life."  In that book, not more than 12 pages in, he nailed it to the wall for me:  "In my experience, there are two distinct classes of retirement: early retirement and traditional retirement [his emphasis] ." The corroboration could not be more explicit; thank you very much Darrow.  He went on to say "There is no official definition, but the following table of relevant factors might help clarify what I mean by those terms." A table that looks like this is provided:

Retirement	Age	Length	Pension	Health Insurance	Could Work?
Early	50's or younger	40+ years	No	Purchased	Yes
Traditional	60's +	20-30 years	Yes	Medicare/benefit	No

After a comment on the potential irresponsibility of retiring early without sufficient capital, he goes on to conclude that "The range of possible investment returns, inflation rates, and health care costs over the long time spans involved in early retirement, magnify [my emphasis] the already unknowable variables in most retirement equations."  So just as we started this post: we have both two retirements and one. Except that it (early retirement) really is different. 

So, who's right? My friend or me?  Well, of course she's right.  But in the end, I think that by imagining myself as "Blanchett's equation 1" or "Kirkpatrick's Row 1" -- with all the magnified risk that that implies -- it means that it looks like there will be no small amount of work ahead of me and many years to go before I can settle into anyone's more quiet, more traditional definition of equation 2/Row 2.  So, I call "not retired." Or maybe retired/not-retired are both true at the same time. Or maybe not-retired is both true and not true.  We are making progress.