FREE NEWSLETTER
Go to main Forum page »
“The riskiness of an investment is not measured by beta but rather by the probability—the reasoned probability—of that investment causing its owner a loss of purchasing-power over his contemplated holding period. Assets can fluctuate greatly in price and not be risky as long as they are reasonably certain to deliver increased purchasing power over their holding period. And a non-fluctuating asset can be laden with risk.” — Warren Buffett, in his 2011 Berkshire Hathaway shareholder letter.
Diversified stock portfolios, e.g., index funds, have proven to be the best long-term investment available for reducing the real risk Buffett is talking about. Since 1928, U.S. stocks averaged about 9.9% nominal annual returns, translating to roughly 6–7% real returns after inflation. Bonds averaged 4.6%, cash 3.3%, gold 5%, and real estate 4.3%.
The following plots in these posts include a “MAX” function that allowed me to draw traces that locked in the peaks of the S&P 500 curves. Thus, you see thick cyan-colored table tops when the market dips for a while. Very cool. I haven’t seen this done before, but I don’t get around much.
I included P/E and CAPE ratios, which aren’t a core topic in these posts, but they don’t clutter the presentation and some of you will appreciate them. P/E is current market price divided by trailing 12-month earnings, and while not inflation-adjusted, the division removes its effect somewhat. CAPE is Shiller’s P/E that uses the same “P” but adjusts earnings for inflation and uses a trailing 10-year average. It’s a much improved look at the long-term, as shown in these plots.
Nominal S&P 500 1960–2025 Graph
The post-dot-com bubble of the 2000s is sometimes called the “lost decade,” but in real buying power it was actually about 17 years. Worse still, by far, were the three lost decades from 1965 to 1995. I haven’t seen this mentioned anywhere, but again, I don’t get around much.
I included a CPI curve that is useful in interpreting the differences between the nominal S&P 500 curve (lower) and the inflation-adjusted curve (upper). I also lightly shaded the real lost decades.
Investing is about preserving and growing usable wealth, which means what you can purchase, not how many units of the local currency you have. In mid-1922 Germany, the exchange rate for one US dollar was about 320 marks. By November of the next year, it was one US dollar to 4.2 trillion marks.
The goal isn’t numbers—it’s purchasing power—yet gains are usually measured with a nominal ruler. Becoming a millionaire isn’t the flex it used to be, and realizing this is foundational to financial wisdom.
(Graphs are high-resolution and should be viewed full screen. You can download them.)
I experienced all of these “lost decades”. 1965-1995 was highly volatile, with annual returns (all dividends reinvested) of -14.66% to + 37.58%. $100 invested in 1965 and all dividends reinvested each year would have grown to about $1745. However, that is misleading because $100 in 1965 had the equivalent purchasing power of approximately $559 in 1995 due to the cumulative effect of inflation over that 30 year period.
I’ve always taken the published returns with a grain of salt. My financial future was not determined by the stock market. It was determined by my career and investing in my business, and my willingness to save at a rate substantially greater than inflation. A portion of those savings were invested in a variety of ways, including growing and maintaining my business. I don’t have much personal reliable data prior to 1995 (1965 to 1995), but in 1996 my CAGR was 0.00% with a net worth of $10,727! Stock market, business and personal financial disruptions caused this.
For the 30 year period, from 1995 to 2025 my GAGR was 19.16%. There was a lot of volatility. In poor years I put cash into the business and was unable to add to my retirement accounts. My worst year of percent change was -92.1% and the best was 526%. There were eight negative years and 19 positive years. Three years were 0.00% change.
Note: I used AI for stock market returns and inflation numbers.
We use a TIPs bond ladder for low-risk preservation of purchasing power in a liability matching portfolio against future expenses. Stocks are held for long term growth.
Rob, could you expound on this a bit more, please? I don’t know what you mean by “a liability matching portfolio.”
Liability matching is the strategy I use to run my portfolio. The concept is straightforward: you map out your anticipated spending needs across different time horizons, then align those needs with assets that match each timeframe.
For near-term expenses—say anything you’ll need in the next five years—you hold short-term bonds or TIPS. These are stable, liquid, and won’t leave you scrambling if the market tanks right when you need the cash. As you move further out on your spending timeline, you can afford to take on more risk since you’ve got time to ride out volatility. So for expenses 10, 15, or 20+ years out, you match those with progressively more aggressive assets like equities.
What you end up with is an asset allocation that’s actually pretty similar to the three-bucket strategy people talk about—safe money for immediate needs, moderate investments for mid-range goals, and growth-oriented holdings for long-term spending. The difference is you’re building it from your actual spending forecast rather than following some arbitrary bucket framework. It’s a more intentional approach that ties directly to what you’ll actually need and when you’ll need it.
Thank you for that detailed explanation, Mark. That makes sense to me, though I hadn’t assigned specific assets to the “buckets” you and David touch on. I have created an income plan for us for the next 10 years and assumed I would stop reinvesting the interest and dividends to have cash available to pay my husbands RMDs (which begin this year) and to supplement our pension(s) and social security if/when needed. I’ve been thinking of changing our asset allocations this year to move to a lower risk portfolio. We are currently at 68/32 but are not needing to withdraw anything beyond his RMDs. That would all change if I were on my own, though. Lots to think about.
Actually Morningstar’s Christine Benz’s three bucket portfolio is not arbitrary at all. In brief you determine your annual budget. Then you subtract other sources of income such as Social Security, pension, annuities etc. The resultant amount is then put into buckets of various investments. Say someone has a million dollar portfolio and needs $10k per year of income from it. You could have 2 years of cash or 20k in bucket one. Eight years or 80k in short term assets such as bonds or TIPS, and the balance of the portfolio in equities. The thinking is you have your withdrawls set for 10 years in relatively safe assets while the balance of funds are available for portfolio growth. Since it is rare that a bear market lasts 10 or more years you have a nice relatively safe assets to ride out the bear and thus can avoid selling your equities in a down market.
Fair point, David. I should probably have said it’s a more tailored version of the three-bucket strategy.
Nicely written! It also explains why proper asset allocation differs for different people. It’s also a moving target as life events or the portfolio changes. Asset allocation is the single most important aspect of investment. Get that right and most downstream decisions become obvious.
Yes, my returns are an illusion. The 10% average S&P 500 returns are comprised of about 32% dividends. The return is about 6.8% if we don’t accumulate those dividends. (Different sources provide varying numbers). Then consider inflation.
“From 2000 to 2025, the average annual inflation rate in the United States was approximately 2.56%, resulting in a cumulative price increase of about 88.14%.”
In other words, if I want to feel like a millionaire, I’ll need at least $2 million.
If bonds return average 4.6% and inflation is 2.56%, then the average annual gain is 2.04%. And there is risk. In 2022 Vanguard’s BND bond index fell from $84.75 and today is $74.27, a decline of 12.4%. Inflation has also chewed into that.
If it’s all an illusion, then what matters to me is my purchasing power, today and until the day of my and the spouse’s death. The averages are of no value to me, whatsoever.
When I have discussed saving and investing I emphasize the consequences of inflation. Begin retirement saving at age 25, and continue for 40 years. Sounds good? “$1 million today will have the purchasing power equivalent to approximately $364,900 in 40 years.“
Hmmm, how far can we go with a $14,596 annual withdrawal (4% of $364,900). So what can we do? Save more in our peak earning years. Not easy if one is committed to spending on “experiences” or spending $hundreds per month on “Treat Culture”. (Fortune December 26, 2025).
“If it’s all an illusion, then what matters to me is my purchasing power, today and until the day of my and the spouse’s death. The averages are of no value to me, whatsoever.”
This is my biggest complaint about financial media and tools. Most sources use 7% as an estimate of real gains, merging gains and inflation into a single value, when they represent different things. Money grows in nominal terms, not “real terms”.
This does two things.
One million today is worth less in the future because the buying power is reduced by inflation, but it is still a million dollars.
For these reasons I ended up building my own models that show nominal growth estimates, and the reduced purchasing power due to compounding inflation.
Inflation is far more insidious than taxes at depleting the purchasing power of a portfolio over medium to long time frames.
Robert,
I like and use that approach – calculating in both nominal and real terms – myself.
“what matters to me is my purchasing power, today and until the day of my and the spouse’s death”
Perfect.
Excellent article. I also haven’t heard the 1966-1995 period explicitly mentioned before, but I did hear Bill Bengen (of 4% rule fame) mention in a recent interview that the 1966 period (I think 1968 exactly as a retirement date) as a worst period than the great depression in his study. He said something to the effect that during the great depression, while stocks collapsed, the country also went through a period of deflation, while the period following 1966 included both a stock market downturn coupled with a period of significant inflation. He said that combination was the worst period on a portfolio in his study. Gene
“He said something to the effect that during the great depression, while stocks collapsed, the country also went through a period of deflation, while the period following 1966 included both a stock market downturn coupled with a period of significant inflation.”
That is where I’m headed in Part II. : )
Excellent. I’ll be looking forward to reading it.
I understand and “get it” that focusing on long-term purchasing power is mathematically sound. My thoughts are it ignores the SORR risk retirees actually face. When spending from your portfolio in the real world, volatility is a clear and present risk: selling shares during a crash permanently destroys wealth, period.
I was thinking, your “lost decades” period from ’65 to ’95—does this include dividends? If it doesn’t, I believe it might change the whole 30-year period narrative.
A time-segmented approach with cash and bonds isn’t necessarily a mistake; it’s possibly the bridge that lets you survive the short-term to reach the long-term. Most people don’t have Buffett’s billions to smooth the process.
The key phrase is “contemplated holding period”. Anything less than 15 years with the broad stock market there is a risk of not keeping up with inflation – the shorter the period the higher the risk.
With respect to SORR this article (What Returns Are Safe Withdrawal Rates REALLY Based Upon?) points out that the SORR risk is not correlated with succeeding 30 year returns, but with the succeeding 15 year returns. It also provides data on the worst years to retire in from a SORR perspective (1907, 1929, 1937 and 1966).
One final thing. What is SORR for someone just starting retirement is an opportunity for someone just starting their career – if they can save diligently and have the stomach to invest into a bad market and wait for the good times to return. For instance, buying into the S&P 500 starting in 1965 and holding to 2000-2005 would have led to excellent results as they would have bought low for 15 years after which they would have seen an epic run. This is also true for the last 25 years or so – a bad run from 2000-2008, followed by excellent returns since.
This Sequence of Return Risk (SORR) discussed by Mark and Adam is a really good point. The link in Adam’s post is excellent and makes a (real) case about the success achieved with 30-year holding periods. Lots of folks are familiar with similar studies showing how rolling 20-year periods avoid (real) losses – but these don’t include retirement type series withdrawals. This is my favorite among the latter.
Series withdrawals, such as 4% annually, add another reduction to the portfolio’s return once they commence, whereas the inflation hit starts on day one.
Regardless, trade in your nominal ruler for a real one. : )
You are absolutely right Mark. The Buffett quote triggered this post (and Parts II and III). This series is theory and fun and maybe useful. The latter only if you have enough money and heirs that you’re able to keep swinging for the fence, a rare and wonderful problem. Then again, if you’re young and wise, this would also be useful – another rare problem. : )
Observation: Isn’t it interesting that people warn us about single country allocations through Japan’s experience while ignoring ours? We really should take Buffett’s advice and focus on real returns in the long-term.
I’ve noticed a heavy focus on the S&P 500 rather than the broader market. I understand the appeal—it has outperformed more often than not. If I were a US citizen, I’d probably think the same way.
If U.S. residents invest 70/30 in favor of U.S. equities (or even implement a higher U.S. allocation) I’m curious how you allocate your non-U.S. and U.S. investments, Mark?
… (and other non-U.S. readers feel free to weigh in).
Andy,I normally invest using market-cap weighting, which for me in the UK means 95% international and 5% domestic. About 65% of my international equity goes to the US, with the remaining 35% spread across the rest of the world.
However, at the start of the year I deliberately rotated away from US markets, boosting my exposure to developed Europe and developed Asia. This brought my US allocation below 50%. My goal was to reduce my concentration risk in the US Mag 7 from 24% down to 15%. I’m retired and won the game—I don’t need the concentration risk or the possible higher return.
My overall asset allocation is aggressive for retirement at 85:15. I maintain only a 15% bond allocation because I have a ten-year term annuity covering my essential spending and a ten-year collapsing bond ladder for discretionary spending. This structure allows me to keep my equity exposure on the aggressive side while protecting against sequence-of-returns risk in my early retirement years.
Mark,
You’ve got well-thought-out strategy. Thanks for providing the detail.
Also, per your bond allocation you had mentioned in a different article/thread that you had a short term Vanguard investment. I had asked if it was VBIRX (or BSV ETF) … inquiring again as I invest in that fund and interested if you do as well or have other recommendation(s). Thanks!
Thanks for the question. As a UK citizen, I actually can’t invest in US mutual funds like VBIRX, and US ETFs like BSV aren’t tax-efficient for me due to estate tax issues and dividend treatment. I use the Vanguard Global Short-Term Bond Index Fund instead – it’s a UK-domiciled alternative that serves a similar purpose with short-term, investment-grade bonds. Vanguard offers various options like this for non-US investors that work better from a tax and regulatory perspective.
Got it. Thanks, Mark.
A good observation and I would suggest that in the US (as alluded to in other comments in this thread) we should be cautious relying solely on S&P 500 performance, even though there has been significant outperformance the past decade. Over a longer time horizon, the S&P500 has experienced long periods underperforming other equity classes.
The link included at the bottom is for a color-coded chart from the Paul Merriman Foundation which illustrates graphically the annual performance of the following major asset classes over the past 100 years:
Small Cap Blend (green)
Small Cap Value (cyan)
Large Cap Blend (S&P 500) (red)
Large Cap Value (pink)
One can see that being diversified across these asset classes can help protect against long periods of S&P 500 underperformance e.g. 2000-2013. The ‘four fund’ colored yellow represents the performance of the 4 asset classes listed above when combined in equal measure. The chart is for US Index Funds and International can provide additional diversification.
Over the past 50 years the S&P 500 has a CAGR of 10.9%, the 4-fund by comparison has a CAGR 12.2% (numbers through 2024).
https://irp.cdn-website.com/6b78c197/files/uploaded/(K)_Quilt_Charts_(1928-2024)_-_2024_Returns_(1).pdf
Thanks for providing these links. I’ve seen these results and class breakouts too. Why is it in results like these or published portfolio allocation models, more often than not, “blend” and “value” are used rather than “growth,” “blend” and “value?”
It doesn’t matter in the long term: https://www.reddit.com/r/Bogleheads/comments/1q1fy9m/vti_or_voo_is_a_choice_that_truly_doesnt_matter/?utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_button
Exactly. People think that investing in the SP500 means they’re missing out on huge swaths of the market and that this actually matters. A quick look at long-term returns, volatility, etc. of SP500 vs Total Market and you’ll see there is no meaningful difference.
You beat me to it! I always notice that. I have never held just the S&P500 (never mind I hold bonds and fixed income as well as stocks). I hold Vanguard’s Extended Market and International funds as well.
When I got out of school in ’86 (19!) and started working with my Dad, he was stuck on the Dow 30, but it wasn’t a hindrance to his decision making. That annoyed me after I figured it out – the Dow 30 was so yesterday – but it was and continues to be a useful metric.
The S&P 500 happened in 1957 when the American market first had 500 investable stocks. For a short while it was the market. Now I’m the old codger using a subset index of the market. The S&P 500 also has the huge advantage for me as the default research metric.
A quick look at the 15 year lifespan of two ETF’s that accurately track the S&P 500 and total US market show good correlation.
What annoys me is that my newspaper, the South Florida Sun-Sentinel, shows graphs of the Dow Jones Industrials and the NASDAQ composite but not the S&P 500.
Yep, many local news broadcasts only mention the Dow when reporting daily results.