Hindsight 8: The mis-allure of the constant spending plan

This is neither a technical or an analytic post nor is it a dissection of the 4% rule, a dissection that has been done a million times before by others better than me.  And don't get me wrong, given the chance for a lifetime of constant income/spending adjusted for inflation I'd be in.  I just don't have that in my future.  Anything I make of what I have at this age is all of my own design and decisions.  The idea of any constant set-and-forget constant approach is appealing but deceiving and I've had enough time looking at retirement finance to see the mis-allure a little better.  Not perfectly, just better.  The point of the post is to provide some slightly impressionistic thoughts on why I look elsewhere for my retirement solutions.




1.  If I "retired" in 2009, I certainly don't live there today. Right now I live in 2017.  I view my retirement as a continuously re-starting enterprise each day I get up.  Constant spend or income approaches would still be stuck in 2009 rather than what I called in a past post the continuously unstable present. When things change, as they will, spending is part of the "team" and needs to be open to change just like everything else.

2. Constant spending programs are easy and simple. But, as in a past post, where I made a probably similarly rude comment, there is a small distance between easy-simple and lazy.  One might make a fetish of easy-simple because it's easy-simple but there are other things to fetishize. Ruin risk is one.   Years between ruin and retirement-end might be another.  Maintaining the ability of the portfolio to lock in a life annuity before it hits a free boundary might be a third. I'm sure there are others.  Simple and easy is just one objective among many and it seems like only the constant spend satisfies that, an approach that has it's flaw when retirement objectives are opened up a little bit.

3. It doesn't really work unless it is insured. Without some kind of pooled longevity insurance (SPIAs DIAs) the constant spend is pretty much guaranteed to fail at some point unless it (spend) is really low and the market cooperates.  Hedge out 85+ and the constant spend can probably make a pretty good run but you'd have to give up some liquidity to do it.  Live to 105 and try a constant spend without the insurance of mortality credits and you are playing another game, one that eventuallywill put your family in a hard place.

4. One can devise constant income plans that are either self-managed or insured but the real trick is spending. Spending is not constant it varies in at least several ways: it is a random variable, there are spend shocks, and there are trends in spending.  The income plan has to anticipate the spending plan and its variance.  As Dirk Cotton has pointed out "The expense side will always be uncertain and our goal in retirement isn't really to secure income but to secure non-negative cash flow."  I don't feel like constant income plans are realistic and they leave some risk on the table in cash flow terms. Constant spending plans are a pipe dream

-------

Here, just for fun is an image of a constant spend plan within a series of spending games I played earlier this year.  Each line is the arc of a spending game in PV terms over a pseudo-lifetime (not a stochastic analysis, btw). All of them started with $1M, spent according to some set of rules, and evolved as the portfolio reacted to a fixed return and inflated or rules-adjusted spending over time.  The exact details are mostly irrelevant. The red line is a constant spend. This was part of an analysis that also included certainty equivalent math, ruin risk, and a free boundary test, none of which is shown here.

My take away from the illustration?

- barring superannuation a constant spend probably works in this fake world I played
- superannuation makes constant-spend a dead letter (if that's the right way to put it)
- all of the plans are more or less a trade-off between spending either sooner or later
- You can't see it but the constant spend is bad a preserving the option to annuitize
- Hedging out late age spend makes the constant spend more appealing
- optimizing a spend plan in a continuous present is an impossible whack-a-mole game

In the end I never really fully made perfect sense of all of this to myself but I do at least agree with the following quote of which the chart, if you squint hard enough, is an illustration.

"Early retirement behavior matters most. The first five to 10 years of retirement have an outsized impact on long-term success. Spending less during these years offers a better likelihood of positive outcomes over the entire retirement horizon…A flexible drawdown strategy leads to better retirement outcomes. The best simulated results came from strategies that increase the drawdown percentage later in retirement." Savings to Income by Luke F. Delorme, Research Fellow AIER

Hence my subjective hindsight that constant spending is like the devil, beautiful to look at but a deceiver.