Thursday, February 6, 2014

Unraveling Retirement Strategies: Floor-and-Upside

In Unraveling Retirement Strategies: Systematic Withdrawals, I set up a structure for comparing the various retirement funding strategies using eight criteria: primary advocates of the strategy, spending plan, investment strategy, liability-matching, longevity risk, spending floor and upside potential for standard of living.

In a second post, Unraveling Retirement Strategies: Life Annuities, I reviewed the strategy of purchasing a life annuity contract from an insurance company. Now, let's take a look at the floor-and-upside strategy.

Floor-and-upside is a strategy that locks in a secure stream of retirement income before investing any remaining retirement savings in a risky portfolio. It is sometimes referred to as “safety first” and derives from The Theory of Life-Cycle Saving and Investing. The Floor-and-upside strategy is championed by Retirement Income Industry Association founder, Francois Gadenne.

You could build a really simple floor-and-upside strategy by using part of your retirement savings to buy a life annuity to guarantee a certain amount of income for as long as you live and then investing whatever is left of your savings in an S&P 500 index fund.

Social Security retirement benefits and pensions also provide secure "flooring".
by Retirement Income Industry Association founder Francois Gadenne - See more at:
by Retirement Income Industry Association founder Francois Gadenne - See more at:
spoused by Retirement Income Industry Association founder Francois Gadenne and it follows a very BASIC premise, but one that after seeing the worry in the faces of many middle Americans, - See more at:

Compare floor-and-upside to the previously described strategies, systematic withdrawals and life annuities. Let's say you looked at the income range of the systematic withdrawal strategy, shown here from my previous blog, and thought, “I love the upside potential for my standard of living, but I'm really not feeling the downside.”
Next, you look at the chart from my annuities blog, shown below, and think, “OK, you took away the downside, but you took away the upside, too! Can't I get rid of the downside and keep a little upside?”
The answer is, “Yes, if you have enough money.”

And here's a tip. In personal finance, the answer is always, “Yes, if you have enough money.”

In this case, you can achieve the desired goal if you have enough money to buy secure income and have some left over to invest in a risky portfolio of stocks. But secure retirement income, or flooring, is expensive (which is another way of saying interest rates are very low), especially in early 2014. Purchasing $10,000 a year of real flooring for the next 30 years currently costs about a quarter million dollars.

So we can compare strategies, instead of starting from scratch to implement a floor-and-upside strategy, let's start with an optimized stock and bond portfolio to implement systematic withdrawals, as I described in the first post of this series. Assume you have $500,000 in retirement savings.

You choose a stock/bond allocation based on how much volatility you can tolerate (or, said differently, how big a loss you could stomach in a major bear market) with the systematic withdrawals strategy. A bond allocation of perhaps 40% to 50% of your portfolio would be typical, so assume you would invest $300,000 (60%) in bonds and $200,000 (40%) in stocks for systematic withdrawals.

To create a floor-and-upside strategy, let's take money from that bond allocation and dedicate it to purchasing secure income for the rest of your life. Perhaps you will invest that money in a TIPS bond ladder or purchase a life annuity. Either can provide you with pretty secure retirement income.

Also assume that your retirement shortfall, the amount of income you will need after Social Security benefits are considered, is $15,000. As I mentioned, $10,000 of secure flooring costs about $250,000 in early 2014, so the additional $15,000 you need will cost about $375,000 that you would use to purchase a TIPS ladder or a life annuity.

You can take $300,000 from the systematic withdrawals bond allocation to cover this, but you will also need to take $75,000 from the stock portion. The remaining $125,000 will remain in the stock portfolio to provide an increase in your standard of living if your investments perform well.

Note, however, that your portfolio allocation is now 75% bonds instead of 60%, so your optimal asset allocation for total portfolio return is no longer attainable. You'll have a safer, but lower return portfolio than you would have had with systematic withdrawals. As I mentioned, if you have lots of savings you can buy even more stocks after you build the floor, decreasing your bond allocation to the 60% that optimizes portfolio return.

Optimizing portfolio return at a given level of risk isn't a goal of the floor-and-upside strategy. It focuses on safety first. It is possible to create a floor-and-upside strategy and maintain your optimum MPT portfolio asset allocation, but only if you have enough money. For most households, the cost of flooring alone will exceed total savings and there will be no additional capital for a risky portfolio at all.

You trade more longevity risk and maximum upside spending potential with systematic withdrawals for less risk of outliving your money, less risk of a decline in your standard of living, but also less upside potential for spending with floor-and-upside strategies.

Following is a chart showing the potential range of annual incomes provided by a floor-and-upside strategy. There will be less upside income potential than the systematic withdrawals chart shows, but a firm floor. That's because increases in the value of the stock portfolio are switched out of stocks into lower returning bonds to secure more flooring.

If there are no remaining funds to purchase stocks after the floor is purchased, the upside curve simply goes away. The range of possible future annual spending with the floor-and-upside strategy will look something like this:
Compare this to the systematic withdrawals chart above. Also note that if you implement the floor with annuities instead of a TIPS ladder, the red line on the floor-and-upside chart would continue to the right for as long as you live, rather than being limited to the ladder length you choose.

Fixed annuities with inflation protection were recently paying out about 3.5%, so $10,000 of lifetime, inflation-protected annual income costs about $285,714 today. A 30-year TIPS ladder, according to a December 2013 study by Wade Pfau, would cost about $247,588 today to provide $10,000 of inflation-protected annual income for exactly 30 years. As interest rates rise in the future, both will become less expensive.

Floor-and-upside is my personal favorite retirement strategy for most people, even if they can only implement the flooring part. I prefer a TIPS ladder to annuities because I prefer to control my capital. I prefer the downside protection of floor-and-upside to systematic withdrawals, even with less upside potential, because I can't think of many financial outcomes worse than going broke in old age. I like to take that one off the table.

Floor-and-upside is a strategy more often embraced by economists. The investment strategy is split. The safe portfolio is invested in the safest possible investment, TIPS bonds, or annuities, while the risky portfolio is typically invested in stocks. The risky portfolio can be quite risky because living expenses are secured by the floor.

The spending strategy is also split. If your future stock investments perform poorly, your spending will be fixed when you purchase an annuity or fund a TIPS bond ladder. If your investments perform well, you may be able to increase spending over time by "raising the floor".

Assuming long term real TIPS returns of 2% (not achievable in today's capital market), a 30-year ladder will payout 4.46% of its initial value adjusted for inflation and deplete the ladder's value in exactly 30 years. A 30-year ladder purchased today would pay out about 4%, according to the Pfau analysis.

The floor is set by the payout of the bond ladder or annuity. The floor can increase over time if stocks perform well. The retiree controls his capital when he invests in a TIPS ladder, but loses control of capital to the extent that he purchases life annuities.

Floor-and-upside provides the highest level of liability-matching. Each year of the bond ladder is purchased to provide the income that will be needed for a specific future year of retirement income. In other words, every future year of retirement is considered a separate liability.

Only the life annuity strategy provides more longevity insurance than floor-and-upside. Flooring can be set up for any number of years expected in retirement, 30 years, 35 years, or more, but annuities will continue to pay no matter how old you become.

Who would prefer a floor-and-upside strategy? Primarily a retiree who is not content with the variable nature and downside potential of income with the systematic withdrawals strategy, or its weak longevity insurance. Someone who is attracted to the concept of knowing where the money is going to come from to fund each future year of retirement. Someone who believes that secure retirement income is worth giving up some of the opportunities of a possible huge bull market after they retire. Perhaps, someone who isn't willing to hand their life savings over to an insurance company.

But, as I said, the full advantages of floor-and-upside are only available if you have saved enough money to both purchase the floor and fund the risk portfolio. Otherwise, "floor-and-upside" becomes simply "floor", but it will still be more attractive to many retirees than a life annuity.

In my next post, I'll describe the remaining major category of retirement funding strategies, the “bucket”, or “time-segmentation” strategy.


  1. Dirk-
    I always enjoy your posts. I agree 100% with you that the Floor and upside model but have also come to the conclusion that if you can afford to adequately fund this model you don't really have a retirement income problem. In your example above, to fund $10,000 of expenses using an inflation indexed annuity would cost $285,716. This is a 3.5% withdrawal rate. Using TIPS would cost $247,588 which is a 4% payout. If I have enough money to fund that and still fund a $100,000 stock portfolio then I am really only looking at a SWR of 2.6% with the annuity and 2.89% on the TIPS ladder. Both of these rates are virtually guaranteed not to fail over the long term. I think that regardless of the funding method, the problem arises when the retiree tries to maximize their withdrawal to meet lifestyle needs by taking 4, 4.5 or 5% of their portfolio every year instead of 3 or 3.5%. The frugal thrifty retiree who saved enough over their lifetime won't have a problem because she has always lived below her means. The spendy retiree needs to take the maximum because he doesn't have enough to begin with.

    1. Mike, thanks for your kind words.

      This is a very-well thought out comment and I have to say that I mostly agree with all of it. However, I’d like to make a few additional comments.

      "I always enjoy your posts. I agree 100% with you that the Floor and upside model but have also come to the conclusion that if you can afford to adequately fund this model you don't really have a retirement income problem.”

      Floor-and-upside is quite expensive because you have to fund 30 years or so of retirement with low-yielding Treasuries. No strategy is cheap. Systematic withdrawals has inherent costs resulting from the need to hold large reserves. But, I disagree with your implication that it is intended for people who have a retirement income problem. Its purpose isn’t to fix a retirement income problem, but rather to prevent one.

      The only strategy that can “fix” a retirement income problem is systematic withdrawals, which has an equal probability of making it worse.

      "In your example above, to fund $10,000 of expenses using an inflation indexed annuity would cost $285,716. This is a 3.5% withdrawal rate. Using TIPS would cost $247,588 which is a 4% payout. If I have enough money to fund that and still fund a $100,000 stock portfolio then I am really only looking at a SWR of 2.6% with the annuity and 2.89% on the TIPS ladder. Both of these rates are virtually guaranteed not to fail over the long term."

      But only virtually.

      I personally don’t care how safe SWR might be with 2.6% withdrawals; it isn’t as safe as floor-and-upside. If I can secure income for the rest of my life AND have plenty of equity for upside income potential, why would I choose to risk any of my standard of living in the stock market? My risk tolerance and resources, of course, may be very different than yours. Otherwise, we’d probably just have one strategy.

      "I think that regardless of the funding method, the problem arises when the retiree tries to maximize their withdrawal to meet lifestyle needs by taking 4, 4.5 or 5% of their portfolio every year instead of 3 or 3.5%. The frugal thrifty retiree who saved enough over their lifetime won't have a problem because she has always lived below her means. The spendy retiree needs to take the maximum because he doesn't have enough to begin with.”

      Agree, except for the value judgment. I know many low-income, frugal people who aren’t able to save enough for retirement. They rarely spend, let alone overspend. But yes, 4-4.5% withdrawals is probably the upper limit of “adequately funded”.

      Excellent comment and thanks for reading!

  2. Dirk-
    Another excellent post. I think some discussion (definition) of what "income" level should be 'floored' would enhance the applicability of this post for a wider audience. What I mean is that the floor could be designed for 'essential' expenses only or for 'total' expenses (essential + discretionary). Although you don't distinguish, there seems to be an inference to floor 'essential' expenses. I would personally agree with this approach. However a! Some retirees might not want to tolerate any income variability, making the approach not right for them. Would like to hear you/other's thoughts.

    1. Excellent point.

      I didn't intend to imply that a retiree should secure only non-discretionary expenses.

      Some retirees might want to do that to provide more upside potential in the stock portfolio. Some might do it because they can't afford to secure more than those essential expenses.

      I would secure my current standard of living, including the discretionary expenses, because I can. I recently recommended a client secure more than his current spending because he was underspending, given his savings.

      Ultimately, the floor you choose depends on what you can afford and how much standard of living you are willing to put at risk for the opportunity to improve it. I discussed this a year ago in a post entitled Your Retirement Savings: Will You Take the Bet?

      Thanks for reading and for taking the time to comment.

  3. Dirk-
    I'd like to see some discussion in this series about variable withdrawal strategies. Didn't see it in your original list of four strategies, and think it's well worth it's own blog.

    1. Variable withdrawal strategies are systematic withdrawal strategies. The system used to calculate the withdrawals may change, but the basic characteristics of systematic withdrawal strategies do not. They provide an unpredictable, variable spending amount depending on market performance, they have the most upside spending potential of all strategies, and the worst downside. Longevity risk is greater than other strategies.

      Withdrawal systems do not provide a firm floor below which spending cannot drop. As William Bernstein is fond of saying, there is no risk fairy that will offer you market returns with no downside.

      All rational systematic withdrawal strategies are variable. I don't believe anyone says anymore that you need not lower you spending when your portfolio declines significantly. (Walter Updegrave at Money used to say that all the time, but he changed his tune after the last crash.)

      Wade Pfau wrote a recent piece on variable withdrawal strategies entitled How Much Can Clients Spend in Retirement? at Advisor Perspectives. Guess I can't leave a link here, so you'll need to Google it.

      Thanks for the comment!

    2. hi Drik, not a blog reader but taken to this article. thanks for addressing variable withdrawal. but my question is what is this worst down side. is it refering to not keeping up with inflation? i would have thought this variable withdrawal have a lower longevity risk because the rules will keep more money

    3. I don't think there is a downside to variable withdrawals based on remaining portfolio balance when compared to fixed withdrawals. I think dynamic withdrawals are better than either, though, and I'll be writing more on this in the next couple of weeks.

      Some might argue that the downside is that the amount you can spend with SWR-variable is unpredictable compared to fixed withdrawals, but if your portfolio declines in value and you keep spending the same fixed amount from it, you may be taking far more risk than the initial 95% success rate you think you have. If you have less money, you probably need to spend less.

      Thanks for writing! Hope you become a regular. Please check out my posts for the next few weeks for more explanation.

  4. Dirk--
    I really appreciate your clear analyses. Thank you.

    I wonder whether or not it might be important to think about a possible ninth criteria for judging withdrawal strategies -- flexibility. For example, if the retiree experiences unexpected but necessary ongoing expense (medical treatment) that raises the needed "floor" the annuity strategy falls short. This example also illustrates how thinking that "if you can adequately fund a floor you don't really have an income problem" can be an error because the adequately funded floor that you thought to be in place suddenly needs to be a few stories higher. At any rate, given a potential need for flexibility building a floor with TIPS seems a wiser choice than floor by annuity...

    1. Thanks for writing, William.

      You make an excellent point, but I believe you are merging two or three different retirement plan issues.

      First, I wouldn't recommend that anyone pour all of their savings into an annuity or any other strategy without first mitigating other risks, any more than I would recommend a retiree take every penny of her savings to pay off her mortgage. We need some liquidity after we retire. We should set aside a significant amount of money in cash equivalents to cover emergencies. That addresses liquidity risk.

      Next, we need some sort of protection against major health crises and the possible need for long term care. That's difficult to do because LTC insurance doesn't work, but we need to account for the risk in a comprehensive retirement plan.

      Actually, I do believe that "if you can adequately fund a floor you don't really have an income problem". That doesn't mean, however, that you don't have a liquidity problem, a health care risk problem, an LTC risk problem, or any number of other risks that you need to address.

      Floor-and-upside addresses longevity risk and provides a predictable, secure income. You need other strategies in your retirement plan for other risks.

      It is also important to note, given a few comments on the topic, two different "income problems". If you don't have enough savings when you retire, you have an income problem. Floor-and-upside doesn't address underfunding retirement. It can only prevent an income problem developing after you retire.

      Thanks for reading!

  5. Since the strategy already secure the floor, it might make sense to invest even more aggressively by long call option instead of stock.

    1. If you have a secure floor, you can certainly take more risk with your stock portfolio.

      Jason Scott recently wrote a paper along the lines of your comment entitled "The Floor-Leverage Rule". He suggests a 3x leveraged stock portfolio instead of long calls, but I have the same problem with both strategies.

      While it is unlikely that you will lose everything invested in a diversified stock portfolio, it is entirely possible to lose your entire investment in either a highly-leveraged portfolio or call options. Your diversified stock portfolio might recover some day, but if calls expire worthless or a leveraged portfolio fails after you retire, you are essentially out of the "upside potential" game forever.

      Thanks for writing!

  6. This comment has been removed by the author.

  7. thanks for these articles Dirk, they are very helpful.
    One thing I dont understand is about the return on TIPS. You say a 2% TIPS would pay out 4.46% (inflation adjusted). I would have thought 0% coupon would give a 3.3333% return (100 /30years) and so 2% coupon would therefore give a 5.3333% return. So my math is somehow very screwed up. Is there a formula for working this out?

  8. sorry Dirk. please ignore my previous question. I googled bond yield formulas so I should be able to work it out from what has come up. Thanks again for all your helpful articles.

    1. Derek, please see my response below.

    2. Derek, this question is asked often, so it is probably worthwhile to address it again.

      You don't need a bond formula, you need a present value formula because you will be spending both interest and principal. The percent of your initial portfolio value you can spend each year is the reciprocal of the present value of an annuity paying $1.02 a year for 30 years with a 2% discount rate.

      The value of that annuity is $22.42 and the reciprocal is 4.46%. That is the percent of initial portfolio (bond ladder) value you can spend each year to consume all interest and principal in exactly 30 years.

      As I recall, there is more explanation in the Sharpe and Scott paper, "A 4% Rule--At What Price?"

      It is easiest to see, I believe, with a spreadsheet. You can find one I posted at (Sorry, you will have to cut and paste the link.)

      I suspect your confusion results from the fact that you are also spending principal as each bond of the ladder matures. 4.46% is a payout amount, not an interest rate.

      Thanks for commenting!

  9. This comment has been removed by the author.

  10. The first link in this article does not work for me. I believe the correct link is:

    1. Thank you! I repaired the broken link and, yes, you identified the correct one. Appreciate it!

  11. Dirk,
    You don't mention one ancillary benefit of floor-and-upside. It makes management of the upside part of the portfolio much easier; as you don't have to be congnizant of a rebalancing requirement as between the risk components and the buffer components of a fully "probabalistic" based portfolio. You can populate the upside portfolio with a diverse set of high beta or high sharpe ratio asset classes - rebalance among them - but otherwise let it run.

    1. Correct. Or, said slightly differently, you can invest the upside portfolio more aggressively if you are not trying to optimize it for both spending stability and growth, as a SWR portfolio tries to do. These are conflicting goals.

    2. I think these are slightly different but overlapping concepts. If you have a 60/40 SWR portfolio, you still might want to have the 60 part be as rationally aggressive as possible. Say with factor tilts, emerging markets, etc. But you are forced by the (risk dictated) asset allocation to sell off risky assets in a bull market to rebalance. In the upside only portfolio, you can just carry the momentum of a bull market as far as you choose. And maybe just rebalance among the risky asset classes.

    3. I think they are vastly different concepts.

      With a floor-and-upside strategy, we allocate assets between a market-dependent upside portfolio and a market-independent floor portfolio consisting of Social Security benefits, annuities, innoculated TIPS bond ladders and the like.

      The upside portfolio allocates resources among capital market-dependent assets with hopefully low correlation to dampen portfolio volatility. Once you have decided on a 60/40 split, I believe you will find that it dominates a sub-allocation of the equity portfolio.

      Rebalancing is a separate issue and hopefully we don't do that often in a bull market.

      "Carrying the momentum of a bull market as far as you choose" implies market timing, a proven poor strategy. Besides, if your floor adequately covers your standard of living, there is little need to sell from the upside portfolio, period.