Get new articles by email:

Oblivious Investor offers a free newsletter providing tips on low-maintenance investing, tax planning, and retirement planning.

Join over 20,000 email subscribers:

Articles are published every Monday. You can unsubscribe at any time.

How Much Do I Need to Save Per Year?

A reader writes in asking:

“I’m 26, and I recently enrolled in my company’s 401k. I know to contribute at least enough to get the maximum match. But how much should I contribute to be on track to retire comfortably? I’ve read numbers anywhere from 10% on up.”

This is one of the most frustrating questions I ever receive. It’s frequently asked, and it’s super important, but I don’t have a very good answer.

The reason the question is so difficult to answer is that it spans such a lengthy period of time. It’s like the already-difficult-to-answer “how much money do I need to retire?” question, but with an additional 3-4 decades of uncertainty tacked on at the beginning of the analysis.

Safe Savings Rates

The best research I’ve seen on the topic is Wade Pfau’s study of “safe savings rates.” Pfau analyzed U.S. historical data (starting in 1926) to determine what percentage of salary an investor had to save per year in order to meet a certain income goal in even the worst-case historical scenario.

For example, Pfau found that for an investor who:

  • Uses a fixed 60% stock, 40% bond allocation,
  • Has 40 years to save,
  • Expects retirement to last 30 years, and
  • Wants to replace 70% of his/her pre-retirement income,

…a 12.27% savings rate would have gotten the job done in each historical scenario. (Table 1 of the article I linked to above shows the “safe savings rate” for various sets of inputs.)

Applying This Concept to Real Life

Unfortunately, applying this “safe savings rate” concept to your real-life finances is a bit tricky.

One complicating factor is the fact that safe savings rates are simultaneously:

  • Probably higher than necessary (because in all historical U.S. outcomes except one, they resulted in excess savings), and
  • Potentially not high enough (because the historical worst-case scenario is not in fact the worst-case scenario).

This is not a fault of Pfau’s work in any way. It’s simply the nature of using historical data to answer questions about the future. But how does one deal with such uncertainty? Should you adjust your savings rate upward (relative to the historical safe savings rate) just to be on the safe(r) side? Or should you plan for a more historically-normal scenario and adjust your savings rate downward?

In addition, we must remember that Pfau’s calculations involved a number of simplifying assumptions.

For example, he assumed that the investor’s inflation-adjusted income is constant throughout his/her working years. In real life, there are raises, promotions, layoffs, career changes, etc. For most people, the constant-real-income assumption is a conservative one — most people’s inflation-adjusted income increases over time. But it’s certainly possible that a particular investor could have the opposite experience. So, how should you adjust the “safe savings rate” to apply it to your own life? Again, there’s no easy answer.

And the calculations ignore taxes completely too. In real life, taxes will play a role in determining:

  • The net rate of return that your investments earn (if you’re investing in a taxable account), and
  • The amount that you’ll have to withdraw from the portfolio per year in order to have a given level of after-tax income.

Of course, it’s difficult to predict with any meaningful degree of certainty how your personal tax rate will change over the next 6-7 decades.

Pfau’s research is helpful in that it gives us something to work with. But once we account for all the different types of uncertainty involved (e.g., investment returns, tax rates, changes in your income over the course of your career, changes in Social Security, the age at which you’ll retire, and the age at which you’ll die) it’s simply not possible to get any more than a very rough idea of how much a young person should be saving for retirement each year.

While I admit it sounds like a total cop-out, I think the most honest answer I can give to somebody early in her career is, “save what you can (hopefully 10% or more), and make sure to reassess the situation every few years.”

New to Investing? See My Related Book:


Investing Made Simple: Investing in Index Funds Explained in 100 Pages or Less

Topics Covered in the Book:
  • Asset Allocation: Why it's so important, and how to determine your own,
  • How to to pick winning mutual funds,
  • Roth IRA vs. traditional IRA vs. 401(k),
  • Click here to see the full list.

A Testimonial:

"A wonderful book that tells its readers, with simple logical explanations, our Boglehead Philosophy for successful investing." - Taylor Larimore, author of The Bogleheads' Guide to Investing


  1. Mike,

    I don’t think it is a cop-out- some questions can’t be answered exactly without knowledge of the future. While the 12.27% isn’t exact I think it is a pretty good goal for a minimum savings rate. Does anyone want to end up with the bare minimum needed to survive in retirement? A few extra percent savings could do a lot toward making retirement much more enjoyable.


  2. You can’t ever get away from having to make assumptions about rates of return and inflation. I have a spreadsheet which maps everything back to present value that I use to see if I’m on track. It allows me to change my retirement dates and see what that means. I discovered, for example, that I can’t afford to retire on the first day that I can claim my pension.

    For those less inclined to do the math (though I’d be happy to explain it), I’m pretty fond of Elizabeth Warren’s “All Your Worth” approach. Keep your ‘needs’ under 50%, your ‘wants’ around 30%, and your ‘savings’ at 20% of your after-tax income.

  3. Mike,

    Thanks for covering my article. Well done! Just to be clear, the safe savings rates (based on the black curve in Figure 5) are relatively lower because they account for valuations at the retirement date and allow for higher withdrawal rates in some circumstances. For someone aiming to use the 4% rule no matter what at retirement, savings rates are connected to the blue curve in Figure 5. These tend to be much higher.

    Also, partly in response to your commenter Larry, I wrote a follow-up study last October which applies the same framework for people in mid-career. The original safe savings rate paper assumes you are at the start of your career with nothing saved.

    Thanks and best wishes, Wade

  4. Wade,

    I remember that being something that (while, again, not a flaw) made me personally uncomfortable. Even if stock market valuations were low at the time of retirement, I’m not sure I’d personally be comfortable withdrawing a large portion of the portfolio per year.

    One thing I’ve wondered about is how the analysis would look if the assumed retirement plan was not a “withdraw x% per year, adjusted upward with inflation” strategy, but rather buying an inflation-adjusted lifetime annuity replacing the desired portion of pre-retirement income. (Though my understanding is that we wouldn’t have very many years of data, as inflation-adjusted annuities haven’t been around for very long.)

  5. Mike,

    This is another study you may wish to look at by the folks at Morningstar:

    They are using Monte Carlo simulations to study the savings rates needed to buy an inflation-adjusted lifetime annuity.

    I’m afraid this study will replace my study as your new go-to study 🙂

  6. Wade,

    Thank you for sharing that study. I had not come across it at all before. I’ve only glanced through it and read the summary so far, but it looks quite useful.

  7. Wade,

    Having now read the study, I don’t think it replaces your study as my new go-to. I appreciate the setup in which they choose to buy a lifetime inflation-adjusted annuity, but the various assumptions made make it no more useful for providing a precise answer.

    For example, for their MC simulations, they use an average real return for stocks of 8.46%. (10.96% minus 2.5% expected inflation.) That’s higher than most people I know would prefer to use for their analyses.

    And they explicitly include Social Security benefits in the analysis. That’s not a problem per se, but I know many people would like it separated out so they can choose to what extent they want to rely on Social Security. (Also, why assume retirement at 65, claiming Social Security right away, and annuitizing all at the same time? It’s almost always preferable to delay Social Security all the way to 70 before purchasing any lifetime annuities in the private market.)

    Also, I’m not really sufficiently knowledgeable to evaluate the quality of their assumptions regarding the hypothetical annuities purchased (explained at the bottom of page 52).

  8. Isn’t deciding via % of current income a little bit backwards? I’m thinking it should be decided based upon figuring out how much money you want in retirement, picking a conservative compounded annual growth rate and calculating the raw value (not % of income) of how much you need to be saving per year to hit your target.

    Its like the saying that you want X months of salary saved — you really want Y months of living expenses saved.

    By my calculations (I’ve got something that will graph the different investing options), if you assume 6% CAGR (low to account for inflation and more conservative investments towards the end of your saving lifetime), you need ~ $7k/year to hit $1 million, ~ $13k/year to hit $2 million, and a little under $20k/year to hit $3 mil.

    If I assume you want to have $100k left after 40 years of retirement (just to be safe) and 4% CAGR: At $1 million retirement savings, you will be able to spend ~$49k/year; $2 mil -> $100k/yr; $3 mil -> 150.5k/yr. Of course these per year values are in distant future money. In 40 years, money will be roughly 1/3rd as valuable as it is today assuming a 3% inflation rate.

    I can send csv files of these calculations and many in between points easily…if anyone is interested.

  9. David,

    I agree (and I’m sure Pfau would agree) that people should actually attempt to estimate their own retirement spending needs rather than just guessing based on a percentage from an article. But when doing a study, you need something to plug in when running the calculations.

    As to your calculations, my biggest concern is that, from what I understand, you’re assuming a constant rate of return. The big thrust of Bengen’s research and the Trinity study in the 90s was to show that even with a given average rate of return, the amount you can actually withdraw safely per year is far less (due primarily to sequence of returns risk — a bad handful of years early in retirement causes major problems).

    For the most part, they found that going over a 4% starting withdrawal rate (which adjusts upward for inflation each year) causes significant risk for retirements of 30 years. Others (Pfau for instance, in a different study of his) found that when looked at in other (non-U.S.) markets, even 4% isn’t very safe.

  10. I thought the question was simple: What percentage of my salary should I save? The answers have a lot of percentages, but not to that question. It was my experience that you start with the percentage to get the max from your employer and the add 1% each year to coincide with your annual raise. After 20 years you will be saving enough!

  11. I think 15-20% is a good rule of thumb for anyone just starting out on their working career. If you start saving later, then more is obviously needed. Also, since our careers aren’t linear and stuff can happen to interrupt our earning potential, I personally am saving 30% with the hope that I can ease off as I approach retirement if it appears I am likely to meet my goal before my desired retirement date. Save early and often!

  12. Mike: you are providing some good suggestions about some modifications for the Morningstar study. Their data assumptions are the classic “calibrated to the US historical data since 1926,” but could be modified to be more conservative. Also, different assumptions about Social Security could be used.

    About using a fixed return assumption. I think that is okay when you are still 30 or 40 years from retiring to get some idea about what to do, but it becomes increasingly problematic as you get closer and closer to retirement. If you assume a 7% fixed return, then you assume your money doubles every 10 years. If your money ends up not doubling in those last 10 years before retirement, then suddenly you are falling well behind your retirement wealth goals.

    I already mentioned the article of mine above which also discusses this issue more:

    Mike, I think we discussed this at a Bogleheads thread, about whether sequence of returns risk only applies to the post retirement period. I think you and I were both agreeing that it can apply to pre-retirement too. You are most vulnerable to sequence of returns risk when your wealth is the largest, because then a given % drop in your portfolio means losing a lot more dollars. That is the danger of assuming a fixed return assumption as you get closer to your retirement date.

  13. Yep, I definitely agree that sequence of returns risk matters prior to retirement as well. It’s just the nature of math that when you have addition going on as well as multiplication, the multiplication functions that matter most will be those that occur after most of the additions have occurred.

    Those last 10 working years and those first 10 years of retirement are super important.

  14. As “commenter Larry” referred to by Wade above, I don’t think Mike’s answer is a cop-out either. I am about 2-3 years from retirement (a minimum of 1.5 years if I retire at 65 so I’m eligible for Medicare), and according to the retirement calculators I am not doing too badly especially if I factor in Social Security – which at my age is likely to be grandfathered even if Congress makes decisive changes to entitlements. So this issue is not just academic or speculative for me, but critical.

    But with so many possible variables as Mike notes — and let’s add on inflation, nreimbursed health care costs, the possibility of global financial upheavals such as another Great Depression — I don’t think anybody can make a reliable statement as to how much one needs to save, other than to say the more you comfortably can, the better.

    Another problem with some of the above comments, however, is that they seem to assume one must reach a certain goal prior to retirement, and don’t consider what kinds of productive financial adjustments one can make after retirement. What I’d really like to hear more from in discussions of this kind are from people who have actually retired, and studies of what people have done to make their retirements successful. I don’t think retirement is necessarily a static condition where one simply depletes the assets they have accumulated prior to retirement.

  15. I do not give a specific value how much to save each year. I think it depends on several factors, including income, the percentage of salary increase, interest rate, as well as inflation rate.

Disclaimer: By using this site, you explicitly agree to its Terms of Use and agree not to hold Simple Subjects, LLC or any of its members liable in any way for damages arising from decisions you make based on the information made available on this site. The information on this site is for informational and entertainment purposes only and does not constitute financial advice.

Copyright 2024 Simple Subjects, LLC - All rights reserved. To be clear: This means that, aside from small quotations, the material on this site may not be republished elsewhere without my express permission. Terms of Use and Privacy Policy

My Social Security calculator: Open Social Security