If you’ve been following my blog closely this week, you know I spent the first half of the week responding to some questions about the US military’s new Blended Retirement System. Those questions and answers were largely dependent on the need to estimate future investment returns. A solid estimate was necessary to make an educated guess on whether BRS would be a good choice.
Today I’m going to focus on why I use 7% when I estimate future investment returns, specifically my retirement accounts and brokerage accounts. I specify those accounts because I do not use 7% growth for my rental property! It’s an investment, but it’s a completely different type of investment. Today I am only talking about investing in stocks and bonds.
I’m just going to apologize in advance. Normally I try to write these “finance basics” posts in a very clear and orderly way. More of a lesson than a story, if you will. Since I’m actually responding to a person, though, this is going to be a mix of the two. Some facts and some opinions, and a whole lot of side commentary necessary to explain the multiple things that got brought up. Bear with me! It may not be the clearest post I ever write, but by the end you should clearly understand why I use the assumptions I do.
More than that, hopefully you will be able to use this information to refine your own estimates.
If you are coming into this post without having read the others, I recommend reading them in this order (if you care about military retirement systems). Make sure you read the comments since that’s where a lot of the discussion happened.
- The Military Guide – “Your Military Blended Retirement System: It’s Not About The Money“
- Military Dollar – “Would I Have Switched to BRS If Given The Option? Part 1“
- Military Dollar – “Would I Have Switched to BRS If Given The Option? Part 2“
- The Military Guide – “11 Years Of Service: How I Chose the Legacy High Three Over The Blended Retirement System“
And that brings us to today!
Bottom Line Close To The Front
A common pushback both against the BRS and against investing assumptions in general is that we overestimate investment returns. Plenty of people do! But I use a 7% nominal return, which is not only not higher than historical averages, it’s well below historical averages for most asset allocations. It’s about in line with the real (inflation-adjusted) rate of return for my preferred asset allocation. If you want to see those historical averages, I recommend this allocation models page put out by Vanguard. It’s a great resource to estimate your own future investment returns.
Nominal and Real Investment Returns
Warner asked this question on my post on Tuesday.
Before I get into answering Warner’s question directly, I want to explain nominal and real investment returns.
Nominal returns are the earnings or losses on an investments before accounting for inflation, taxes, or fees, expressed as a percentage. If you hear that the S&P 500 long term average return is about 10%, that’s the nominal return.
By “inflation-adjusted” Warner was referring to real returns. Real returns are when the gross percentage realized on an investment (earnings or losses) are adjusted for changes in prices due to inflation or other external effects. If you use real returns, you are keeping everything in current-day dollars. This has the advantage of being simple to understand, but not reflecting what actually happens in your account.
How Inflation Fits Into This
Unless I say otherwise, on this blog when I suggest a rate of return I am talking about nominal rates. That’s because I personally find it easier to adjust up to future dollars than try to keep everything in current dollars. I do this for the very simple reason than that it is easier for me to control how much inflation affects my life than any other factor.
What do I mean by that? Well, there is overall monetary inflation that I have no control over – true. But I do have a ton of control over how much I allow monetary inflation to impact my life.
An example: the Consumer Price Index Inflation Calculator says $500 in June 2013 is equivalent to $541.50 in October 2018. I picked $500 because that’s how much I budgeted for groceries, entertainment, and miscellaneous purchases in 2013. Guess I how much I budget for those things in 2018? $500. Why hasn’t it gone up? Because I’ve made choices that allow me to keep the spending level the same, even though my lifestyle hasn’t changed.
So since I can control personal inflation more than I can control the stock market, I find it easier in my brain to use nominal rates of return. I then estimate inflation to calculate what my future expenses will be. When my future projected income (which is somewhat out of my control and relies on assumptions) is greater than my future projected expenses (largely within my control and more calculable), I am financially independent.
If you are going to do this (use nominal returns and estimate your future expenses) I recommend using 3.5% to estimate inflation. Why 3.5%? Easy! That’s about the long term average for inflation in the United States. You’ll find different numbers because there isn’t just one type of inflation but the numbers I usually see are 2.5%-3.8%. 3.5% is higher than inflation has been in the last 30 years, but about on par with longer periods.
Should I Have “Moved the Goal Posts?”
Warner brought up a good point – I was using Hank’s estimate of a $10,000 pension difference between the two pension systems. I then based my math on how close I could get to making up that $10,000 difference.
But that $10,000 difference is based on what it would be today, in 2018. By the time I’m retirement eligible, in 2024, it won’t be $10,000 anymore, at least not for an O-5 with my promotion dates.
So I should have increased that difference by pay inflation to present a more accurate picture. I was too focused on responding directly to Hank’s numbers – that’s my fault.
So let’s do that: let’s see what a $10,000 difference in 2018 means in 2024. Once again, I’ll be assuming that military pay raises will be 1.7% on average. The pension, then, will similarly increase by 1.7% per year.
- 2018: $10,000 difference
- 2019: $10,000 x 1.017 = $10,170 difference
- 2020: $10,170 x 1.017 = $10,342.89 difference
- 2021: $10,342.89 x 1.017 = $10,518.72 difference
- 2022: $10,518.72 x 1.017 = $10,697.54 difference
- 2023: $10,697.54 x 1.017 = $10,879.40 difference
- 2024: $10,879.40 x 1.017 = $11,064.35 difference
Using that 1.7% increase assumption, the difference between BRS and Legacy is actually $1064 more per year than Hank and I were using.
Of course, since the pension increases by such a small amount and the TSP can increase by much more, it will be a wash just a few years into retirement. Project forward 20 years, and the picture looks even better for BRS. But as you will see below I was looking at the numbers right at military retirement.
Why Is 7% A Good Estimate for Future Investment Returns?
Okay, so far we’ve covered that:
- I usually use 7% nominal returns when I talk about estimating future investment returns
- Inflation is usually about 3.5%
- So if I use a 7% nominal return, minus 3.5% inflation, that means you should expect a 3.5% real return, right?
Well, not quite. If nominal returns were actually 7%, then expecting a 3.5% real return would make sense. But nominal returns are usually much higher.
To estimate future investment returns accurately you need to look at what you are investing in. The more conservative your investments, the lower the long term returns (nominal and real) will be.
If someone was only invested in the S&P 500, as Warner implies I’m talking about, a reasonable estimate would actually be about a 10% nominal, or 7% real, return. That’s because the S&P 500 has returned 9.8% nominally since 1928. The S&P 500 doesn’t represent all stocks, though. Vanguard found, looking at a variety of stock indices, that a 100% stock portfolio has had a 10.3% nominal return since 1926.
But most people aren’t 100% invested in stocks. One common asset allocation, and the one Warner uses, is 60% stocks and 40% bonds. A 60/40 portfolio has a nominal return of 8.8% since 1926. Based on the typical age of my readers (over 90% are in their 20s and 30s) some of you probably have about an 80/20 portfolio, which has had an average 9.6% nominal return.
You don’t get to the 7% nominal return area until you are looking at approximately a 30% stocks/70% bonds mix. This would be a very conservative mix for my readership!
So looking at Warner’s question again…
- I use 7% as a nominal return, which is clearly a conservative nominal return looking at history
- I should’ve adjusted the difference in the pension…but once I did, you see it’s not much!
- My 7% assumption wasn’t inflation-adjusted, but even if it had been it wouldn’t be outperforming the historical average for stocks. 7% is nearly exactly in line with the historical, inflation-adjusted real return for the S&P 500.
- But even if I had been using 7% as the real return, it would’ve just meant I was looking at about a 90% stocks/10% bonds portfolio. That’s right in line with recommendations for my readers, using the 110 or 120 minus your age rules.
But if you are using a very conservative allocation, then maybe a 7% estimated nominal return doesn’t sound conservative. Warner, for instance, is in his early 30s and has a 60/40 stock/bond split. That’s very conservative for someone his age, especially if you (like me) consider your pension to be a bond-equivalent. I don’t mean to suggest he or anybody should change their allocation – that’s a personal choice. But it is objectively called “conservative” for that age range in the personal finance community.
Follow-up: Explaining the 7% Even More
After asking the above question on The Military Guide, Warner asked more questions in the comments on my blog.
Am I suggesting a 100% stock portfolio now and indefinitely?
This question falls into a straw man argument logical fallacy. I didn’t suggest a 100% stock portfolio in any of my posts or comments this week. I’ve never suggested people hold a 100% stock portfolio, either now or indefinitely into retirement. In fact, I specifically said other people probably should not buy 100% stocks 18 months ago.
And while I personally am buying 100% stocks right now, I don’t hold a 100% stock portfolio myself. My portfolio has been about 7%-10% bonds since I started investing.
I bring this up because Warner makes some assumptions about my beliefs and recommendations in an (intentional or not) effort to discredit or undermine my words. That’s pretty normal…bloggers have to deal with that kind of thing a lot. Usually, and probably in this case, it’s because the person doesn’t have the larger picture on what the blogger thinks so they are filling in the holes with assumptions.
But I try to be incredibly upfront with you guys about what I do and why. If you ever have questions, just ask! I’ll definitely explain.
Specifically addressing the “indefinitely” part of the question, I never suggested that my investment calculations went beyond the point of retirement from the military. In the case of the “Would I Have Switched to BRS” Part 1 and Part 2 posts, I was doing calculations from the beginning of my career (age 21 in my case) through potential retirement (age 41 if I left at 20). While I doubt my investment allocations will change much between now and age 41, they will probably change a fair amount by the time I reach age 60, or 65, or 80.
Point being: I don’t personally think I in any way recommended people have 100% stock now. But I know for sure I didn’t recommend they have them “indefinitely into retirement.” I think you should have the asset allocation that allows you to grow your money as much as possible while allowing you to sleep at night.
The nice thing about that is that you can have a healthy bond allocation and still expect 7% nominal returns. Or much higher!
My Long Term Estimate For Future Investment Returns
As I’ve hopefully shown by now, my use of a 7% nominal return isn’t based on me expecting “stock returns of the past decade, which have exceeded historical averages, to continue.”
My TSP investment returns over the last decade have been way more than what I estimate. That’s because the stock returns from 2008-2017 have indeed been good, despite the 2008, 2011, and 2015 returns. Using my personal TSP returns, a $100 investment in January 2008 would have turned into $238.47 by December 21, 2017. Meanwhile I’ve been averaging 2% annual personal inflation, meaning $100 spending would have turned into $119.51. My return was slightly higher than the S&P 500 return over the same period.
Do I predict my long term returns, nominal or real, will be that high? NO! I estimate them based on long term (90+ years) averages, and then I trim some more from there. Specifically I think my personal returns will be about 10% nominal and about 6.5% real. But I calculate my estimates based on a conservative 7% nominal estimate, which means 3.5% real.
So Warner’s second question in this group of questions also uses a straw man argument. I don’t expect the returns of the last decade to continue, and I haven’t said I do. I actually frequently warn people that they shouldn’t use the last decade to estimate future investment returns. The last decade has been good for returns, even including the three (yes three) bad years.
What I do think is reasonable is to take the historical returns for your preferred allocation, subtract a couple points from that, and then use the remainder to estimate future investment returns. If you have an 80/20 split, with a long term 9.6% nominal return, then estimating 7% nominal returns is conservative.
Summary: Estimating Future Investment Returns
Hopefully this shows clearly why I use a 7% nominal return, and why I think that’s a conservative estimate. And hopefully it prompts any of you that don’t have a clear reason behind your assumptions to refine your thinking. There are many ways to think about this, but (if you haven’t figured this out by now) I think using history is a good jumping off point.
I’ll leave you with some tweets from friends about their own estimates!
For a post tmrw:
– what estimated rate of return do you use for projections?
– is that gross or net of inflation?
– appreciate any thoughts on why you use that!I know it’s #Thanksgiving but consider this a good thing to escape to when convo with family gets to be too much 😁😂
— Military Dollar (@Military_Dollar) November 22, 2018
Sorry 8% rate of return, less 3 percent inflation, for a 5% real rate of return. Other times I use 7
— Anne @ Unique Gifter (@UGifter) November 23, 2018
I plot a range from 4 to 8%. My projected assets aren’t adjusted for inflation, but my projected expenses are. I’m also trying to think how best to incorporate a hypothetical mega-crash into projections.
I’ll never have just one projection. Spreadsheet sheets are cheap. 🙂
— Debt Ascent (@DebtAscent) November 22, 2018
(Debt Ascent gets me)
I use 7% gross return and 3% when I adjust for inflation. 7% gross is more realistic and I tend to be more conservative. HTH.
— Confessions of FI (@FiConfessions) November 23, 2018
I’m using 6% after inflation. But as a weirdo, I’m forecasting zero raises between now and FIRE, so that’s hopefully a little conservative.
— Josh Overmyer (@Jovermyer1) November 22, 2018
7% post-inflation, but my spreadsheet has different models. I track my average monthly contribution, then model different market returns and different average monthly contributions
— Zero Day Finance (@zerodayfinance) November 22, 2018
I generally use 8% without inflation. My reasoning is the S&P has retuned about 12% over the past 30 years on average. However, most portfolios include a mix of other lower interest investments. I would lower the projection if shorter term rate (<10 yrs).
— Financial Pilgrimage (@Financialplgrm) November 22, 2018
I would use 6% net of inflation, but I don’t even use simulations because there’s too many unresolved variables in my life right now to know my post retirement expenses.
Honestly, I’m not even sure the rate of return is all that important compared to sequence of returns.— Seonwoo Lee (@Seonwoo__Lee) November 22, 2018
5% after inflation. Unsure if that’s too high.
— Leigh (@leighperfin) November 22, 2018
The moral of the story is, you Personal Finance bloggers really get into the nitty gritty about your finances.
Wait. Am I supposed to be one of those personal finance bloggers too?? I guess in my defense, we are a good decade or more away from FI, have a good deal of our investments in real estate – cash flow (vs stock market), and don’t expect to fully retire anytime near when we hit FI. I DO mean to have better spreadsheets eventually, but I’m not there yet. For now, saving a good chunk of money is good enough for me.
I see that you have a lot to say here but I can’t get beyond “you Personal Finance bloggers”
🤣🤣🤣🤣🤣
¯\_(ツ)_/¯
In my opinion, the longer your time horizon, the more comfortable I am with 7% as a conservative estimate for the nominal return. As you point out, who knows how a decade will turn out. However, I think it’s more than reasonable to use 7% as a conservative estimate over a long timeframe. Combining that with your chosen 3.5% inflation rate, the real return of 3.5% is low for what I use for estimates.
As you mentioned above (and thanks for the shout-out), I’m estimating between 4-8% nominal returns, but my working assumption is 2% inflation (for most expenses), assuming my real return will be in the 2-6% range, though optimistically skewed toward the latter.
In the end, it’s my opinion that there’s absolutely nothing unreasonable assuming a 7% nominal return with a >50% holding in equities over a long time period.
Yeah, that “long timeframe” but is so important. I wouldn’t recommend this to anyone looking to stash house down payment money, for instance, not with a high stock percentage at least. But for retirement investments? I think 7% is very reasonable.
So in the end, your 3.5% expected real return is right there with my 3%. Then why did you lead me to believe that you were talking about a real return (and then insult my argument as being a straw man)? I get the sense that you never really thought about this until I brought it up.
I’d like to point out a few other oddities:
1. The official BRS calculator still seems to treat the investment return input as a real return, and it still defaults to 7% (part of why I initially thought you were doing the same). So I hope people aren’t using that default value for all the reasons we’ve given (i.e. it very much implies a 100% stock portfolio).
2. You choose to bake 3.5% inflation into your investment returns, but only inflate the pension at 1.7%. The 1.7% number comes from the current period of historically low inflation, and the 3.5% number is an average over the past ~100 years. Why choose to be inconsistent here? For the purposes of Hank’s question, $10,000 inflates to almost $20,000 after 20 years at 3.5%.
3. On one hand you talk about how unlikely it is that anyone will complete 20 years to vest for the pension at all, so they should choose BRS. On the other hand, you talk about treating the pension as a bond to justify a more risky portfolio, presumably because the pension is so safe(?).
Frankly, I think you are confusing yourself and others with this switching back and forth between real and nominal dollars. You certainly confused me.
Well it seems I was right that you wouldn’t be happy with my responses, Warner. Let me try again.
I don’t think you are going to like this answer very much either, but my answer would be “I didn’t lead you to believe I was talking about a real return…you made that assumption yourself.” I said your arguments were straw man arguments because you used straw man arguments.
Straw man fallacy:
1. A straw man is a common form of argument and is an informal fallacy based on giving the impression of refuting an opponent’s argument, while actually refuting an argument that was not presented by that opponent.
2. Substituting a person’s actual position or argument with a distorted, exaggerated, or misrepresented version of the position of the argument.
If you can find an instance of me doing the things you accused me of that I called straw men (suggesting people use a 100% stock portfolio not only now but indefinitely and expecting the returns of the last decade to continue) I’d happily admit my mistake, as I did with not accounting for inflation on the pension. But I don’t think you are going to be able to find that. Hence, straw men. You are arguing against a position I never took in the first place, and saying that I did.
As for not thinking about this until you brought it up, I think the clearest sign that I have indeed thought about it before is the multiple links I included to prior posts of mine where I addressed exactly those things. One is 18 months old. I assure you I didn’t predict this conversation 18 months ago and prep the battlefield. I’ve been consistent on these points. I don’t expect you to know everything I’ve ever said, though, so that’s why I linked to old posts to show that I’m not saying them for the first time.
To the other points:
1. The BRS calculator defaults to presenting the information in future dollars, so I disagree with your assessment that it seems to treat the investment return input as a real return. That’s a pretty good indicator that it is using nominal returns. In fact I’m willing to bet most of the people using it have never heard of real vs nominal returns, and only know things like “Use my recommended mutual funds and you’ll earn 12%” (Dave Ramsey) so I’d guess most people also use the calculator thinking in terms of nominal returns…which means the people creating it probably intended it to use nominal returns. But I don’t know that for sure (I can’t find it written on the site) and I don’t expect you to agree with that.
–1.a. Again, 7% actually doesn’t imply a 100% stock portfolio if it’s a nominal return…which I clearly believe it is. I’ve discussed this at length above.
2. Actually I used 1.7% as my assumption for pension growth because if you use the retirement calculator, that’s what it shows (although it says it’s using 2.1%). So since the pension calculator assumes 1.7% pay raise growth, I did too. Here are two screenshots of the High-36 calculator, showing what the pension would be in 2018 as an O-5 retiring at 20 vs the same situation but retiring in 2024. If you do the math it’s actually 1.75265% average annual raise if you want to get incredibly specific.
–2.a. No, I don’t know why the calculator says 2.1% is the Anticipated Annual Active Duty Pay Raise when it’s actually using 1.75265%.
—-2.a.1. If you check the years individually, it doesn’t increase by 2.1% or 1.75265%. It seems to use a different amount each year. I used what it was for my years, because I wrote the post about me.
—-2.a.2. I don’t know why it doesn’t use a standard amount.
—-2.a.3. I’m not going to do infinite calculations for this. There are better uses of my time. But I’m not just making things up to prove a point. Please just ask why I’m doing things, I’m happy to explain but the embedded accusations just really aren’t helpful or fair.
–2.b. If you want me to redo the numbers on the pension to exactly match 1.75265% I can do that too. But remember I said I was being conservative, so I rounded down.
–2.c. The military pay raise, and thus the pension starting point, isn’t tied to inflation so I wasn’t being inconsistent by not tying them…they simply aren’t tied. The military pay raise is currently tied to the Employment Cost Index, which is influenced by but not the same thing as CPI inflation.
–2.d. It’s true that $10,000 inflates to almost $20,000 after 20 years of compounded growth, but we weren’t looking at 20 years of compounded growth. We were looking at 6 years of compounded growth (from today in 2018 to when I’m retirement eligible in 2024). In that case $10,000 compounded by 3.5% growth becomes $12,292.55. But again, it wouldn’t make sense to use 3.5% because that’s the estimated inflation of expenses, not the estimated growth of military pay or the pension.
3. Again, I know you think all the bloggers like to say everybody should choose BRS but you’d be hard pressed to find me saying that. At best you’ll find that I’ve said I think BRS is the better choice for the vast majority (and it clearly is, since the vast majority don’t retire from the military). So let’s ignore that straw man and move on to treating the pension like a bond. Yes, I personally (like many others) treat the pension like a bond while I am in the military. The pension is safe! If you get it. If you don’t get it, then investing aggressively while you are young will help set you up for greater TSP amounts when you get out that will help you offset what you “lose” by not getting the pension. Not that I’m telling anybody how to invest, that’s a personal choice! (I’ve said that repeatedly, too). But that is my thought process, yes. Aggressive investing is what’s allowed me to reach lean financial independence already, and what will have me totally FI in a few years even if I don’t get the pension.
I’m sorry you are confused, but I didn’t switch between real and nominal dollars. You just assumed I did. If you go back and read my comments and posts you will see this is true.
Listen, I know you feel like the bloggers are leading people down the wrong path on this one. We’ll see soon enough whether more people should have switched (spoiler alert: more should switch while they still have the chance!). But please stop accusing me of saying things I haven’t said. I’ve written thousands of words on BRS, I’m happy to write more if you simply ask. No need to try to put words in my mouth. I’m not your enemy.
Wow, that image did not come out well. Sorry.
The stated pension for an O-5 retiring in 2018 at 20 years is $52,138.00.
The stated pension for an O-5 retiring in 2024 at 20 years is $57,866.70.
If you multiple $52,138 by 1.0175265 six times (to represent the result in 6 years) you get $57,866.70.
Year Growth Pension Amount
2018—$52,138.00
2019—$52,138.00 x 1.0175265 = $53,051.80
2020—$53,051.80 x 1.0175265 = $53,981.61
2021—$53,981.61 x 1.0175265 = $54,927.72
2022—$54,927.72 x 1.0175265 = $55,890.41
2023—$55,890.41 x 1.0175265 = $56,869.97
2024—$56,869.97 x 1.0175265 = $57,866.70
Just to chime in here for a second:
The military pension COLA is calculated the same way as the Social Security COLA and the VA Disability Compensation COLA, by taking the third quarter’s inflation data and annualizing it. It does not always match the CPI because that’s measured differently.
The Consumer Price Index is calculated year-round and might be adjusted retrospectively. It can also be tweaked for excessively cold weather, spikes in gasoline prices, and even the Wal-Mart/Amazon effect on corporate productivity & pricing. Then there’s the chained CPI, which is a whole different can o’ worms.
Military pay raises were indexed to the Employer Cost Index a number of years ago (1990s?) but Congress is not required to hold to that index… any more than they’re required to maintain the intended coverage of the Basic Allowance for Housing.
I’ve been suggesting for over a decade that military servicemembers (and retirees) should invest in asset allocations which are very high in equities. (Personally, these days we’re around 95% equities and the rest in cash.) Of course military families need an emergency fund like everyone else, but it only needs to account for prolonged unemployment if you’re leaving the service. While you’re in uniform I think it’s perfectly reasonable to expect continuity of income, and a military pension is the rough equivalent of the income thrown off by a portfolio of I bonds. Admittedly that’s an imperfect analogy but it works well enough for asset allocation decisions.
Anything less than 70% equities (especially when combined with a military pension) is excessively conservative and will keep people in the workforce longer than necessary. Of course it’s also important to sleep comfortably at night, but becoming comfortable with volatility seems preferable to working longer for a more survivable portfolio.
Thanks for reading, carry on.
Thanks Doug. The Annual Pay Raise page (https://militarypay.defense.gov/Pay/Basic-Pay/AnnualPayRaise/) does say pay raises have been tied to ECI (although they can go higher) since 2007. Did that change again?
And you know I totally agree on viewing the pension as a bond portion!
Thanks again for the thorough response. I think we’ve both made our points and can leave it at that. I hope someone else finds this discussion insightful.
I hope so too.
i gave up on those calculators with the assumptions awhile ago and just decided to remain nimble. we just raised our cash equivalent position to 17-18% from around 3 as we’re older than you average reader (50’s). i would say that equates to 3 cushy years or 4-5 leaner years expenses and we still have about 1.3 incomes so we’re not drawing down. i skipped any bond allocation for the past 8 years with the 10 year below 3% much of that time. my thinking is the s+p yields about 2% with the upside of equities so why settle for barely more than that in bonds except that people like hard and fast (set it and forget it) rules. they want to be told what to do versus digesting diverse information and deciding what’s best individually.
i strongly agree that if you’re getting a substantial pension then fixed income is all set. right now we’re at 67% stock/ 15% preferred etf’s/ 18% cash,stable value. i think part of that mix depends on how well you have invested and have a nice cushion if you have done well into your 40’s. my hope is that people at least take an interest in their own money and keep learning.