Welcome, everyone, to another installment of the Safe Withdrawal Rate Series! See here for Part 1, but make sure you also check out Part 26: Ten things the “Makers” of the 4% Rule don’t want you to know for a more high-level, less technical intro to my views on Safe Withdrawal Strategies! Today’s topic is something that has come up frequently in reader inquiries, whether through email or in the blog post comments. Let me paraphrase what people normally write:
“Here’s how I can guarantee my withdrawal strategy won’t fail: I simply hold a portfolio with a high enough yield! Now the regular cash flow covers my expenses. Or at least enough of my expenses that I never have to worry much about Sequence Risk, i.e., liquidating principal at depressed prices.”
I’ve seen several of those in the last few weeks and it’s a nice “excuse” to write a blog post about this very important topic. So, what do you think I normally reply? Want to take a guess? It’s one of the two below:
A: Oh, my God, you got me there. This is indeed the solution to once and for all, totally and completely eliminate Sequence Risk! I will immediately take down my Safe Withdrawal series and live happily ever after.
B: Your suggestion sounds really good in theory but there are serious flaws with this method in practice. It will likely be no solution to Sequence Risk. And in the worst case, your “solution” may even exacerbate Sequence Risk!
Anyone? Of course, it’s option B. It sounds like a great idea in theory but it has very serious flaws once you look at the numbers in detail. Let’s take a look…
Chasing higher yields: A quick guide
If you’re familiar with the Safe Withdrawal Rate Series, you’ll notice that I focus mainly on equity and bond portfolios comprised of U.S. large-cap stocks (e.g., S&P 500 index) and 10-year U.S. Treasuries. (But make sure you also check out my free Safe Withdrawal Simulations Google Sheet where you can also simulate returns of portfolios with short-term bonds, gold as well as Fama-French style biases!) So, what I call a 60/40 portfolio would be closely matched by a portfolio holding 60% in an S&P 500 index fund (e.g., iShares Ticker IVV or State Street’s SPY) and 40% in the corresponding U.S. Treasury bond fund, something like the iShares ETF Ticker IEF. But that’s not the entire ETF universe! Here’s a table with a selection of ETFs in different asset classes: bonds (U.S. Treasuries, Aggregate, Corporate) and equities, together with their Expense Ratio and Dividend/Interest yield.
Side Note: The reason why I put the “Yield” in that table in quotation marks is that this is the 12-month rolling (backward-looking!) yield according to Yahoo Finance, so especially for the bond indices, the actual forward-looking yield is even a little bit higher.
That 60/40 portfolio would have a weighted dividend/interest yield of about 2.1% today. If you were to plan a 4% withdrawal rate, you’d be forced to generate an additional 1.9 percentage points, roughly half of your withdrawals, from withdrawing principal. Scary when facing Sequence Risk, right? So, I can understand the temptation to look for a higher yield. To that end, I also consider two additional portfolios. They are the ones mentioned on the fellow FIRE blog Millenial Revolution (MR), specifically, in their post on how to generate higher yields to guard against sequence risk. They call this method the “Yield Shield.” Here they are:
Portfolio 2 (MR baseline portfolio): 30% VTI (the ETF version of the famous VTSAX U.S. Total Stock Market Index), 30% International Stocks (Vanguard Ticker VEU) and 40% in the U.S. Aggregate Bond index and the aggregate bond index a higher yield because it also includes corporate bonds. This portfolio already has a higher yield (2.55%) than the Big ERN 60/40 portfolio: non-U.S. equities have higher yields and the Aggregate Bond index includes a significant portion of corporate bonds in addition to the relatively low-yielding U.S. government bonds.
Portfolio 3 (“Yield Shield” Portfolio): With Portfolio 2 as a starting point, let’s really juice up the yields:
- Out of the U.S. equity portion, move 10 percentage points into REITs with a pretty impressive yield of more than 5% and an additional 5 percentage points into the Vanguard High-Dividend-Yield ETF (Ticker VYM) with an also pretty impressive 3.2% yield!
- Out of the U.S. aggregate bond portfolio, move 10 percentage points into investment-grade corporate bonds (LQD) and 20 percentage points in Preferred Shares with a Dividend yield North of 6%.
Side note: I took the liberty of replacing their proposed Vanguard BND and VTC funds with the iShares AGG and LQD, respectively. They are very similar funds but have a longer return history. Especially, the VTC only starts in 2017 so it would be impossible to simulate how the yield shield portfolio would have performed during the Great Recession of 2007-2009.
Wow, now we’re getting somewhere! A yield of 3.69% in our portfolio! And again, that’s likely a little bit underestimating the true expected yield. Keep this portfolio and maybe a really tiny “cash cushion” to finance the gap between the 4% withdrawal rate and the 3.69% yield for a few years and we’re done and never have to worry about Sequence Risk, right? Wrong! Let’s look at how the three portfolios would have performed over the last decade or so:
- The simulation period is May 2007 to December 2018. This is the longest time horizon for which I can gather returns for all 9 ETFs used in the 3 portfolios.
- The returns come from PortfolioVisualizer.com and they are the monthly total returns, i.e., including dividends. If you like to check the returns for yourself, see this permalink with my inputs and the returns of all underlying ETFs and the three portfolios.
- I calculated how a $1,000,000 portfolio would have performed using the three different ETF allocation. This is assuming a 4% withdrawal rate, i.e., with the first month $3,333.33 withdrawal and subsequently adjusted for inflation.
See the results below. Uhm, bad news for the dividend chasers. The low-yield portfolio (60% IVV and 40% IEF) would have performed the best and ended the simulation period roughly where we started (in real terms). Portfolio 2 would be down about 25% and the Yield Shield portfolio would be down almost 30% by 2018. Clearly, this Yield Shield Portfolio doesn’t eliminate Sequence Risk. It doesn’t even lessen Sequence Risk. The Yield Shield actually exacerbates Sequence Risk! Look at how far the portfolio fell at the bottom of the 2009 bear market trough!
I could just leave it at that and be done for today. But that’s not a thorough and comprehensive exercise. I like to mainly understand why this yield chasing strategy would have failed so badly!
So, let’s look at the potential problems of this method:
Problem #1: Efficient Markets
At first glance, the higher yield strategy makes perfect sense. If you assume that you can edge out a higher yield, then, all else equal, you should be better off, right? Well, the emphasis is on the phrase all else equal. Unfortunately, that “all else equal” assumption doesn’t work so well in the real world. So, the first objection is that if this was so easy then everyone would do it and the advantage of the higher yield is arbitraged away, see below:
Case in point, Vanguard’s high-dividend yield index fund VYM. How did it perform during the simulation period? Not very differently from the VTI (U.S. Total Stock Market Index) or the IVV (S&P 500 Index). See the cumulative return chart below. Not only did the VYM fail to outperform the broader indexes by the gap in the dividend yield, the total return (= price return plus dividends) even underperformed a little bit. Thus, if you had banked on the higher dividend yield you would have lost all of it – and more – in the price return. Bummer!
Side note: I’m not here to bash the dividend investing crowd. I’ve even grown quite a bit more sympathetic to the dividend growth investing approach. I even used some “play money” to built my own little dividend portfolio replicating the “Dividend Aristocrats” ETF (Ticker NOBL). I did that in my M1 Finance account (affiliate link) where I can buy fractional shares and M1 would automatically reinvest my new contributions and dividends to rebalance to the target weights. But even I have to admit that the higher dividend yield is not all alpha. I’m OK if I lose some of the extra yield through a lower price return. Hopefully not all of it!
Problem #2: Confusing Nominal and Real Returns
When computing how much yield you need to pull this off, keep in mind that many retirees assume that their withdrawals will grow (roughly) in line in with inflation. It’s probably fair to assume that equity dividends grow in line with inflation over the long-term (and often much faster than inflation). But the same is not true for bonds and preferred shares. If you buy a bond with a principal value of $100 and a 4% yield then you may generate enough income to sustain your 4% withdrawal rate in the first year. But in subsequent years, your income will be eroded by inflation! So, for bonds, you’d need to target a yield of not just 4% but 4% plus your CPI projection! The same is true for Preferred Shares. Normally, Preferreds are structured as essentially “perpetual bonds” with a notional value of $25 and a certain fixed yield. There is no growth in your dividend income, neither explicitly nor implicitly, so you’d have to target your withdrawal rate plus CPI. Otherwise, you’d erode your purchasing power!
Problem #3: More yield may imply less diversification
That would be the scenario below: Not only is the price return lower when you chase yields but it’s so much lower, that your total return lags the low-yield portfolio! How is that possible? Well, the reason people want to hold bonds in their retirement portfolio is not exclusively to generate income but also to diversify equity risk. Holding higher-yielding bonds may expose you to more downside risk the next time a recession hits. In other words, for every dollar you generate in extra yield you may lose much more than that dollar in the price return.
So, let’s look at the correlation matrix for the nine ETFs in the three portfolios, see the table below. I sorted the nine ETFs by their correlation with the S&P 500. And guess what, the three bond alternatives to the low-yielding IEF all have higher correlations with the equity indexes. Not only that, moving to higher and higher yields (IEF to AGG to LQD to PFF) you increase the equity correlation in exactly that order!
So, it’s no surprise that some of the ETFs proposed for the 40% bond share had a very underwhelming performance during the simulation period, especially during the 2007-2009 period when Sequence Risk hit your portfolio. Look at the cumulative performance below:
- PFF (Preferred shares) with the highest yield among the four ETFs had the worst performance overall and a 50%+ drawdown in 2009, even worse than the S&P 500 equity index. Bad news from a Sequence Risk perspective!!!
- LQD did perform on par with the IEF if you look at the endpoint only. But don’t be fooled! This is the cumulative return chart without withdrawals. Because the gray line spent so much time below the blue line, LQD would have done worse due to Sequence Risk!
Problem #4: Dividends can be cut!
Part of the poor performance of the yield-chasing portfolio is due not only to poor price performance. One ugly little detail that the Yield Shield fans seem to ignore is that even your dividend income will take a hit when the next recession strikes. It certainly did during and after the Great Recession. This is true for the broad market indexes (e.g., the S&P 500) as I have alluded to in a previous post, see the chart with the drawdowns from the peak of real dividends per share:
And it gets even worse; some of the popular dividend boosters – REITs and Preferred Shares – experienced even worse dividend payment streams during and after the Great Recession! Let’s look at the dividend payments per share of the Vanguard REITs ETF (VNQ) and the iShares Preferred ETF (PFF) since 2008 (the peak in per-share income). (Side note: I couldn’t find the USRT dividend stream in Yahoo Finance, so I used the VNQ for this chart!) In the chart below are the per share dividend payments, both in nominal terms and adjusted for inflation. REITs had a sharp drop in dividends (-40%, much worse than the S&P 500!) almost immediately. It took until 2016 to recover to the pre-recession level in inflation-adjusted terms (much longer than the S&P 500).
Preferred Shares kept up their income a little bit better initially but eventually got hit. Really badly! And Preferred Shares look like their dividend income is now down pretty much permanently! More than 40% from their peak in 2008, see below!
Well, the lesson here is that dividends are in no way contractually guaranteed. They can be cut at a moment’s notice. Also, be really careful with preferred shares. They are frequently (mis-)interpreted as a bond alternative, see the Millenial Revolution post. But there’s an important distinction between corporate bonds and preferred stocks! Here is the kind of language you’ll frequently find in a Preferred Stock prospectus (Source: Goldman Sachs Group Inc 6.20% Non-Cumulative Preferred Stock, Series B (GS.PRB) Prospectus, retrieved through PreferredStockChannel.com):
“In the event dividends are not declared on Series B Preferred Stock for payment on any dividend payment date, then those dividends will not be cumulative and will cease to accrue and be payable. If we have not declared a dividend before the dividend payment date for any dividend period, we will have no obligation to pay dividends accrued for that dividend period, whether or not dividends on the Series B Preferred Stock are declared for any future dividend period.”
In other words, the issuer has the option to skip dividend payments. Not only that, there is no obligation to ever make up for that in the future (i.e., the phrase “Non-Cumulative”). This is a stark contrast to a corporate bond where a missed payment would usually be considered a default. So if someone tells you that you can avoid Sequence Risk with Preferred Shares I’d be really, really cautious!
Side note: I don’t want this to sound like I’m bashing Preferred Shares. We hold a small amount in the PFF ETF and a six-figure portfolio of individual Preferred Shares. I also enjoy the impressive yield but I am aware that this is a reward for taking on a risk premium!
Oh, my, I got carried away! I have some more material but we’re already pushing 3,000 words. I’ll do a follow-up soon (see Part 30, published on 3/4/2019). Anyhow, what have I learned over the last few days putting together this post? What can we do about Sequence Risk? Nothing can really solve the problem unless you come up with a market timing algorithm that avoids the occasional equity drawdown altogether. I have studied a few legitimate techniques that alleviate (not solve!) Sequence Risk:
- Guyton-Klinger (see SWR Series Part 9 and Part 10)
- CAPE-based withdrawal rules (see SWR Series Part 18)
- Glidepaths for the equity/bond shares (see SWR Series Part 19 and Part 20)
- A “cash bucket” (see SWR Series Part 25)
But none of these will ever completely shield you from Sequence Risk. And the “Yield Shield” method is even inferior to that. It’s a bit like prescribing Marlboro Reds for enhancing pulmonary health. It ain’t working! In other words, the Yield Shield has the potential to make Sequence Risk even worse, as evidenced in the last recession. Of course, we can’t really tell how this asset allocation will work in the next bear market, but the previous failure of the Yield Shield was probably not just a fluke. As long as you believe that in future recessions dividends will be cut, preferred shares will get into hot water, inflation erodes some of your fixed-income yields, etc. – all pretty reasonable assumptions – then there’s a pretty good chance that this strategy will fail again! I have consistently recommended to my readers to stay away from this strategy. But if you want to be the guinea pig and try this in real-life, best of luck in the next bear market!
Also, make sure you check out the two follow-up posts:
- The Yield Illusion Follow-Up (SWR Series Part 30)
- The Yield Illusion (or Delusion?): Another Follow-Up! (SWR Series Part 31)
All right, hope you enjoyed today’s post! Please leave your comments and suggestions below!
- Part 1: Introduction
- Part 2: Some more research on capital preservation vs. capital depletion
- Part 3: Safe withdrawal rates in different equity valuation regimes
- Part 4: The impact of Social Security benefits
- Part 5: Changing the Cost-of-Living Adjustment (COLA) assumptions
- Part 6: A case study: 2000-2016
- Part 7: A DIY withdrawal rate toolbox (via Google Sheets)
- Part 8: A Technical Appendix
- Part 9: Dynamic withdrawal rates (Guyton-Klinger)
- Part 10: Debunking Guyton-Klinger some more
- Part 11: Six criteria to grade dynamic withdrawal rules
- Part 12: Six reasons to be suspicious about the “Cash Cushion“
- Part 13: Dynamic Stock-Bond Allocation through Prime Harvesting
- Part 14: Sequence of Return Risk
- Part 15: More Thoughts on Sequence of Return Risk
- Part 16: Early Retirement in a low return environment (The Bogle scenario!)
- Part 17: Why we should call the 4% Rule the “4% Rule of Thumb”
- Part 18: Flexibility and the Mechanics of CAPE-Based Rules
- Part 19: Equity Glidepaths in Retirement
- Part 20: More thoughts on Equity Glidepaths
- Part 21: Mortgages and Early Retirement don’t mix!
- Part 22: Can the “Simple Math” make retirement more difficult?
- Part 23: Flexibility and Side Hustles!
- Part 24: Flexibility Myths vs. Reality
- Part 25: More Flexibility Myths
- Part 26: Ten things the “Makers” of the 4% Rule don’t want you to know
- Part 27: Why is Retirement Harder than Saving for Retirement?
- Part 28: An updated Google Sheet DIY Withdrawal Rate Toolbox
- Part 29: The Yield Illusion: How Can a High-Dividend Portfolio Exacerbate Sequence Risk?
- Part 30: The Yield Illusion Follow-Up
- Part 31: The Yield Illusion (or Delusion?): Another Follow-Up!
Picture Credit: pixabay.com