Buffer Assets, Bucket Plans, and Sequence of Return Risk
He has clearly done a lot of thinking and modeling on how to avoid running out of money in retirement. The easiest way is to have an extremely low withdrawal rate, but that could mean working additional years and shortening your retirement or making frugal choices you don’t want to (or necessarily need to) make.
Today, Dr. Graham looks at a number of ways one can use “buffer assets” to mitigate the risk of a particularly bad stretch of returns near your retirement date, such as those we had to kick off this century.
Stick around to the end; you will likely be surprised to see what might work best!
Buffer Assets, Bucket Plans, and Sequence of Return Risk
How effectively do various buffer assets protect portfolios against sequence of return risk?
Buffer assets stand between cash and equities. During the distribution phase of retirement, buffer assets prevent selling equities at a loss to pay expenses. This is especially important just prior to and during retirement.
“Reverse dollar cost averaging” (or Dollar Cost Ravaging) can have devastating and permanent effects on portfolio value. Sequence of return risk (SORR) describes this long-term effect of negative initial stock market returns on the portfolio.
What is a Buffer Asset?
Like many terms in retirement planning, “buffer asset” doesn’t have a clear definition. Wade Pfau coined the term by borrowing from the insurance industry. Dr. Pfau states he started using it after seeing cash value life insurance described as a “volatility buffer.” He specifically refers to buffer assets as those assets outside of your investment portfolio where returns are not correlated with the stock market. He cites cash value life insurance and reverse home mortgages as examples.
Other assets have the same intent, however: to immunize your portfolio against SORR and prevent reverse dollar cost averaging.
In this blog, I define buffer assets as any strategy, investment or asset that stands between cash and equities. Buffer assets provide you with the luxury to sell equities when you chose, rather than being forced to sell for living expenses. Buffer assets’ specific purpose is to avoid SORR; they are bucket 2 assets.
The Bucket Plan
Actually, 1-3 years of cash reserve is not infrequently called a buffer. In the simplest of “bucket plans,” bucket one is cash or treasury bills and bucket two is everything else. This plan has a bad rap, and rightfully so, as cash is for spending and not an effective buffer to reduce SORR. This is a cash reserve plan rather than a bucket plan.
Other bucket approach strategies are mere time-segmentation of the portfolio’s asset allocations into different mental buckets. There is a bucket for the next 5 years, then 5-10 years, and then more than 10 years. Stock allocation increases as the time until the money is needed increases. This simplistic bucket approach, while attempting to prevent poor behavior during market downturns, is just fancy mental accounting of a systematic withdrawal strategy (such as safe withdrawal rate). This is a “now, soon, and later” plan rather than a bucket plan.
A more complete bucket plan has 4 buckets. Bucket 1 is 1-2 years of cash for spending. Buffer assets fill bucket 2. The third bucket contains stocks and bonds in an appropriate ratio to maintain overall asset allocation. Bucket 4 provides lifetime guaranteed income (a “floor” for fixed/essential expenses) from social security, pensions, and annuities. This plan has “time segmentation” (buckets 1-3) in addition to “goal segmentation,” which is the flooring of fixed income needs from bucket 4.
Time segmentation, however, is commonly misconceived. The intent is not to spend sequentially in order from bucket 1 to 2 to 3, rather you refill bucket 1 from 2 OR 3 depending on market conditions. If socks prices are soaring, you sell some bucket 3 to refill bucket 1. If there is a bear market, you convert buffer assets into cash and avoid reverse dollar cost averaging. SORR is greatest 5 years before and 10 years after retirement, so the eventual goal is to spend down bucket 2 when the risk has passed.
Goal segmentation is important to consider as well. No one wants to run out of money in retirement, and the future is plagued by market, inflation, and longevity risk. Understanding future fixed/essential expenses (vs discretionary expenses and legacy goals) and matching these expenses with secure, inflation adjusted income is important for piece of mind.
Bucket 4 provides this floor, which is funded at some point in the future (say when you turn 70 and take social security, or when you consider an immediate annuity at age 80), so bucket 3 only has to last for a known number of years, rather than until some unspecified time in the future when you die. This distinction is subtle yet important: bucket 3 only has to last until you fund bucket 4 and have all future expenses covered during the “no-go” years of retirement. Let’s call this the 4-bucket plan with goal segmentation.
An Aside for FIRE
Folks who retire early have additional considerations. Physician on Fire started a drawdown plan in early retirement chain which discusses strategies planning for 50 years or more in distribution.
Flexibility is key to FIRE. Those who are responsible enough to FIRE won’t stand idly by as their portfolio is decimated. Buffer assets are for traditional retirees but also require special considerations for FIRE.
Back to Buffer Assets
I define buffer assets as a bucket 2 asset. Assets are not just what you own, they are also advantages or resources that you have. Strategy plays a large role in SORR immunization, so any description of buffering is incomplete without consideration of additional planning options.
A classification system for buffer assets is lacking, but I propose the following classification scheme:
- Cash Reserve Strategies: As mentioned above, cash is for spending and not an effective buffer asset by itself.
- Volatility Dampeners: Frequently mentioned “ladders” of CDs, bonds, or multi-year guarantee annuities (MYGAs). One may actually consider laddering a combination of 2 or 3 of these assets depending on interest rates.
- Planned Allocation Strategies: Rising equity glide paths or bond tents are important considerations.
- Income and Expense Modification: Income need is driven by expenses in retirement. One can use human capital (i.e. working part time) or modify spending (spend conservatively or have guiderails for spending).
- Uncorrelated Non-portfolio Assets: Cash value life insurance and reverse mortgages fit the bill given low correlation to market returns.
- Additional Assets: Businesses, hobby collections or rental properties can be sold to hedge against a down market.
- Goal Segmentation: Fixed guaranteed income “floor” provided by social security, pensions, annuities or TIPS ladder.
- Other Strategies: Social security claiming strategy, income from immediate annuities or longevity insurance from deferred annuities.
- Income portfolio: Preferred stock, dividend stocks, utility stocks, REITs, convertible bonds and MLPs deserve mention, but won’t be considered further due to high correlations with the market and other issues.
Let’s look at the effect some of these buffer assets have on a fictional portfolio during an especially bad sequence of return–the years 2000-2010.
Consider a fictional couple who just retired at 64. They have $4.5M in assets which includes a $1M house, $500,000 in cash value life insurance, and $500,000 of LLC shares paying 5% a year.
Figure 1 (Portfolio)
Figure 1 shows the asset allocation of the $2.5M portfolio. They have a $1M taxable account with a large cash reserve, and US and International stocks and bonds. In their $1M rollover IRA, they are 60/40 US stocks/bonds. Their $500,000 Roth IRA is more aggressively invested. (Spreadsheet from PoF)
Figure 2 (Asset allocation compared to target allocation)
Above in figure 2, we see the combined asset allocation of the portfolio on the left which is similar to a model Moderate portfolio of 60/40 seen on the right. They are overweighed in US equities and cash.
Social security will provide $3000 a month at age 65. They plan on spending $150,000 a year, a 6% withdrawal rate from their investments or a 3.3% withdrawal rate from combined assets. Assume inflation is 2.5% except for health care which is 5%, and health care will cost an additional $10,000 a year.
Figure 3 (Assumed sequence of returns for this scenario)
Figure 3 shows the sequence of returns from 2000-2010 against which we will stress various scenarios. Note from year 11 until the end of plan, return on stocks is 6%, bonds 3% and cash 2%. These may be low returns on average, but are nonetheless interesting to look at as an added stress.
Results of the Plan
Monte Carlo simulation of this plan demonstrates an 82% success rate using standard returns on financial planning software. Stress testing reveals concern for a market crash, inflation, and longevity. Net worth at year 90 (end of plan) is $5.8M with their portfolio worth about $3.8M.
When we stress this plan with return assumptions from figure 3 (actual returns from 2000 to 2010), however, there is a zero percent chance of success and their portfolio expires at age 84.
Now, let’s examine some different options for immunizing their portfolio against SORR.
Delay Social Security
Figure 4 (Delay in social security)
What if they delay claiming social security? The dark green demonstrates portfolio value over time if they delay to 70. The lighter green shows portfolio value if they claim immediately. Over time, we see the investment accounts expire at age 84 with either strategy.
The cross over point where income from social security is higher as a result of a delayed claiming strategy is age 79. It is interesting to note they actually do worse with a delayed claiming strategy as they have a higher distribution rate from the brokerage account prior to claiming, and this occurs during some very negative years of market returns.
Therefore, to prevent SORR during years where you delay claiming social security, buffer assets are needed to draw on during years of negative stock market returns.
A More Conservative Allocation
Figure 5 (20/80 Asset Allocation)
What if they were more conservatively allocated? Figure 5 shows the effect of a 20/80 portfolio, where they have 80% of their assets allocated to fixed income. Although they suffer from less volatility especially during the down years, and have an extra $500,000 after the poor sequence, the initial drawdown in the portfolio value is eventually eaten away by inflation and runs out at year 87.
A rising equity glidepath or a bond tent would improve this scenario. These allow programmatic increases in equity exposure without “market timing.” Unfortunately, however, I’m not able to demonstrate these changes in portfolio allocation with current planning software. This is a good reminder that a financial plan is not set in stone and requires periodical re-evaluation.
Bond and CD ladders are popular for SORR planning. For a large upfront investment, you can guarantee 5 or more years of income.
Figure 6 ($75,000 a year bond ladder)
In this scenario, $500,000 is used to purchase an 8-year treasury bond ladder paying out $75,000 a year. A Bond/CD Ladder Toolkit helped model the purchase price.
As seen above, this bond ladder fails. Note that at year 9 there was a large drop in market prices, so perhaps a 10 year bond ladder may have performed better. CD ladders or use of MYGAs are not modeled, but also suffer from current low interest rates.
Given low current interest rates of about 2.4-3%, treasury bond or CD ladders are not all that exciting. Historically, when interest rates were closer to 6%, a similar bond ladder costs about 20% less.
What about Selling a Business?
Figure 7 (Sale of the business)
They do have assets outside of their portfolio. Figure 6 shows income sources for retirement. As seen above, when they are 65 years old, they sell the LLC and receive $500,000 in income (though they lose the 5% returns for the rest of the plan). In this scenario, they still run out of money at age 84. A poorly timed infusion of cash is not beneficial during these harsh sequence of returns. The income peak at age 90 is the life insurance policy paying out upon death.
What About Annuities?
In this scenario, the fictional couple really didn’t need income, so models of single premium immediate annuities (SPIAs) are not shown here–they didn’t help against SORR in this scenario anyway. Let’s look at differed income annuities (DIAs) and their kissing cousin QLACs, which is a qualified longevity annuity contract — a DIA in a qualified account — and see if they help.
Figure 8 (10-year $500,000 DIA)
Income sources are shown above with a DIA. Assume $500,000 from the brokerage account purchases DIA payments of 14.2% after a 10-year deferral period. Note that we have social security and income from the LLC. At 10 years the annuity starts to pay, as well, but their investment accounts run dry at age 83. The large withdrawal from the brokerage account raises the withdrawal rate from the portfolio higher than can be tolerated during SORR, leading to this poor result.
Figure 9 ($130,000 QLAC)
A QLAC is more successful than a DIA as seen above in figure 8. Even though it is a smaller sum of money (QLAC maximum contribution is $130,000 currently), it is taken from the IRA rather than the brokerage account. This QLAC only utilizes 13% of the IRA, and although the income is fully taxable, there are benefits in maintaining funds in the brokerage account, which are spent down first. A QLAC still fails in this scenario but I have high expectations that QLACs will be part of a successful SORR plan among 401k Millionaires.
Get a Job
What about working in retirement to increase income? Given their high spending levels, they would have to earn $60,000 a year for 10 years in order for Monte Carlo analysis to be above 50%. They are already in high tax brackets from other income, and it is difficult to out-earn their high spending rate.
What Actually Works?
Let’s now move on to demonstrate effective ways to avoid SORR in this scenario.
Figure 10 (Decreasing spending to $120,000 a year)
Variable spending strategies are important for retirement income planning. Logically and emotionally, this is what most retirees will do during down markets, though a good bucket plan allow planned spending during the “go go” years of early retirement.
Note that the portfolio is not much larger even after the second major decrease in the stock market, but the tail end stays much higher showing that likely they could have increased spending again after the sequence of risk evaporated (usually about 10 years).
Cash Value Life Insurance
Figure 11 (50k loan a year from cash value life insurance)
You can “borrow” the cash value of a permanent life insurance product tax free. In this example, $40,000 a year is borrowed from the cash value of the policy for 10 years. Even though this is a fraction of their spending, the results are impressive, and this plan does not fail. Cash value life insurance is not a popular topic among DIY investors, but clearly it is a consideration for those worried about SORR. You can pay back the money you borrow from your life insurance or it can be deducted from the death benefit when the time comes.
The problem with borrowing from your cash value life insurance policy is, of course, it must be set up decades before it is needed. In this example, the life insurance is counted as an asset, but of course this is an asset outside of the portfolio and not correlated with stock returns, which is why Wade Pfau finds them so useful.
Figure 12 (Reverse mortgage of 40k a year for 10 years)
Reverse mortgages have become slightly more popular in the last few years, mostly thanks to a book by Wade Pfau. He suggests a Home Equity Conversion Mortgage (HECM) to be used early in the planning process, rather than as a last resort (as home mortgages are considered in popular culture).
These are FHA loans with mortgage insurance written into the deal. HECMs are non-recourse loans, which means as long as the property is kept up and property insurance and taxes are paid, the homeowner keeps title and cannot be evicted.
Loan-to-value amount is based upon the owner’s age (you must be at least 62) and current interest rates. The lending limit is based upon a maximum value of $726,525 and is usually 40-60%.
HECMs have a line of credit option which is of most interest. You can open the account and only use it if needed. Also of note, this line of credit grows larger with time.
In order to combat a negative sequence of return, draw on your line of credit in years when equities are down in order to avoid reverse dollar cost averaging. Money from the line of credit is tax free, which may be especially important if you are getting income from your fully taxable IRA, or if additional ordinary income would increase taxation of social security, or cause increased Medicare surcharges.
So, Which Buffer Assets Would You Choose?
In summary, only three strategies perform well against a repeat of the market foibles from 20 years ago.
Spending less money, or even a variable spending plan, is an important consideration. The 4-bucket system, however, is set up so that you don’t have to reduce spending during go-go years even if you get stuck with a bad sequence of returns.
Bucket 2 buffer assets fill the cash account during the bad years, and bucket 4 ensures a floor lifetime of income. Bucket 2 assets are spent down as needed over the first decade or two of retirement simulating a rising equity glidepath while providing inflation protection.
Other effective assets to consider are reverse mortgages and cash value life insurance. Obviously, cash value life insurance must be set up decades in advance (and has other downsides), but is a consideration for high-income individuals who already have large tax-deferred retirement accounts.
While reverse mortgages can be set up anytime, the amount you are able to borrow from actually increases with time if you set up your mortgage early (though of course after age 62). The key to these assets: they are not correlated with stock market returns, and equally important, the “income” is tax free.
I would hate to be accused of fear mongering. The decade from 2000-2010 was not typical, though not infrequently used by insurance salesmen masquerading as financial advisors to pitch equity indexed annuities or life insurance that “have no downside” and are “guaranteed not to lose value.”
Thus, it does take planning to make sure you don’t crash and burn due to SORR in early retirement. That actually is the point of retirement planning. No single buffer asset is perfect for everyone. Likely a combination of investments, products, and strategies will be needed in order to spend freely during your early retirement all the while planning to live a long and comfortable life.
David Graham, MD is an Infectious Diseases Physician and Registered Investment Advisor in the state of Montana. He blogs at FiPhysician.com
[PoF: The results surprised me, initially. I did not expect two products you don’t hear much good about, cash value life insurance and reverse mortgages, to protect against a poor sequence of returns.
However, it makes perfect sense when you think about it. Both a home and a life insurance policy represent a stockpile of money that’s set aside for a different purpose (living in and for next of kin, respectively).
When you take either of those stashes of money and repurpose them to instead pay for your retirement lifestyle, you’ve got more money available to spend. Yes, there will be less for heirs, but accessing the value in them during your lifetime can help guarantee you don’t run out of money.
I think there may be better, less costly ways to make that money accessible. In the case of cash value life insurance, not buying any in the first place (and using term life insurance when it’s needed) will give you a larger bankroll to play with from the beginning.
Rather than a reverse mortgage, you could sell outright and move to a smaller home that is purchased with a down payment or you could choose to rent. Such a move would require more disruption, but it might be the best choice, and it avoids the high fees that gives reverse mortgages a bad name in the first place.
I shared these thoughts with the good doctor, and he wanted to point out that a big reason these two strategies work is the tax-free nature of the “income,” which is really just a return of your principal.
Thanks again to Dr. Graham for the insightful post and giving me a lot to think about!]
Do you have “buffer assets” to fill a bucket in your retirement plan? How do you think a poor sequence of returns is best counteracted?