Every generation of investors discovers the same argument and thinks they invented it: "Why bother with bonds? Stocks always win over the long run." The logic feels airtight. The historical charts confirm it. And if you're 30 with decades of compounding runway ahead of you, going all-in on equities seems almost obvious. Why would you voluntarily accept lower returns?
I've heard this argument from genuinely smart people. I've also watched those same smart people sit on their hands in March 2020, their all-stock portfolios down 34% in five weeks, no dry powder, no mechanism to buy anything at fire-sale prices. They were right about the long run. They just couldn't survive the short run. The argument about long-run stock superiority is technically correct on average - but averages don't retire. You do. Your specific decades, your specific market environment, your specific income needs. That's what matters.
And here's the part that really bothers me: the research foundation for "stocks always win" is shakier than most people realize. Professor Edward McQuarrie of Santa Clara University spent years reconstructing 19th-century bond market data that previous researchers had botched or ignored entirely. What he found upended a cornerstone assumption - the historical case for stocks' dominance depends almost entirely on data from a narrow, unusually favorable 20th-century window. Before that window? Bonds held their own for decades at a stretch. Then in 2022, the bond-equity relationship that underpins most diversification theory broke down completely. And 2025's tariff shock tested it again in ways that show just how conditional - not absolute - the bond-as-diversifier story really is.
What follows is the honest, complicated version of the answer to "why not 100% stocks?" Not the textbook version.
The Assumption We Never Questioned
The intellectual foundation for "stocks for the long run" comes mostly from Jeremy Siegel's book of the same name - genuinely important work that shaped how an entire generation thinks about financial planning. Siegel pulled historical data showing stocks dramatically outperforming bonds from 1926 to the present. The chart became famous. The conclusion became dogma. And nobody questioned the starting date.
That's the problem. What relatively few investors know is that the data before 1926 - everything before the standard starting point that everyone uses - tells a very different story. Professor Edward McQuarrie spent years reconstructing the actual US bond market from 1793 onward, using real bond prices and coupon payments from market transactions. Not theoretical models. Not estimated yields. Actual prices people paid and coupons they collected. His findings, published in "The US Bond Market Before 1926: Investor Total Return from 1793," upended what most of us assumed was settled science.
Stocks Always Win Long-Term
Data from 1926 onward shows stocks dramatically outperforming bonds across every major long-term horizon. The conclusion: for long-horizon investors, equities dominate. Bonds are just a volatility-reduction tool with an expected return cost attached. Simple. Clean. Persuasive.
The problem: 1926 - 2025 is a single data window. It happens to capture the greatest period of US industrial and economic expansion in human history, fueled by demographics, policies, and a global order that no longer exist in the same form.
Stocks' Dominance Is a 20th-Century Phenomenon
McQuarrie found that in the 19th century, long-term bonds didn't just hold their own against equities - they outperformed over multi-decade periods. After the Panic of 1837, through the post-Civil War deflationary era, and in other long stretches, bonds delivered superior real returns. Not by a little.
Siegel's earlier datasets had relied on synthetic or estimated bond returns, often using New England municipal yields with unique tax privileges - systematically understating what bonds actually returned before 1926. The data was wrong.
This is not some minor academic footnote you can wave away. If stocks' dominance is primarily a product of a specific 20th-century environment - post-WWII industrial expansion, baby boom demographics, favorable monetary policy, the dollar as global reserve currency - then the confidence with which we project that dominance onto the next 40 years of your life should be a lot lower. McQuarrie isn't saying bonds will outperform going forward. He's saying the margin of certainty that most investors assume is dramatically wider than the data actually supports.
Sit with this for a second: you get one investment lifetime. The sample of independent 40-year periods in the modern data is essentially two. Two. Would you bet your retirement on a sample size of two?
The 2022 Wake-Up Call: When the Relationship Broke
Even investors who acknowledged some theoretical uncertainty about stocks' long-run edge still relied on one relationship with near-absolute confidence: when stocks fall, bonds rise. This negative correlation was the bedrock of modern portfolio theory. It was the engine behind the 60/40 portfolio. It was the implicit assumption behind every "safe" allocation recommendation financial advisors made for twenty years straight. Nobody questioned it because it kept working.
Then 2022 happened. The Fed launched the fastest rate-hiking cycle in four decades to fight inflation that had been building since 2021. Bonds - which mechanically lose value when rates rise - fell hard. At the same time, rising rates and recession fears crushed equities. Both asset classes lost money. Simultaneously. For the entire year. The 60/40 portfolio posted its worst year since the 1930s. The negative correlation that everyone's retirement plan depended on just... vanished. Two assets that were supposed to zig and zag in opposite directions zigged together, month after month.
The chart above tells the story better than I can describe it. For roughly twenty years - 2000 through 2021 - the rolling correlation between S&P 500 returns and bond returns was predominantly negative. Stocks fell, bonds gained. This wasn't a coincidence or a lucky streak. It reflected a specific macroeconomic setup: low inflation, interest rates trending down over decades, and a Federal Reserve that reliably cut rates whenever equity markets got shaky. In that environment, the bond-as-diversifier thesis wasn't just theory. It was mechanically reliable. You could set your watch by it.
What 2022 showed us - brutally and unmistakably - is that this diversification relationship is conditional on the inflation environment. When inflation runs hot, the Fed tightens, rates rise, bond prices fall. And stocks fall simultaneously because higher rates reduce the present value of future earnings. Same driver hitting both asset classes at once. The negative correlation flips positive. The diversification you were counting on disappears at precisely the moment you need it most desperately. It's like discovering your fire insurance doesn't cover the kind of fire that actually burns houses down.
Negative Correlation Era
Low inflation, falling rates, Fed cutting on equity stress. Bonds reliably rose when stocks fell. The 60/40 portfolio's golden era.
Both Fall Together
Fastest rate-hike cycle in 40 years. Inflation drove both stocks and bonds lower simultaneously. The 60/40's worst year since the 1930s.
Oscillating, Uncertain
Correlation unstable, regime-dependent. Some restoration of negative relationship as inflation slowed, but not the reliable pattern of pre-2022.
Tariff Shock Test
Treasury yields rose while stocks fell - the opposite of the expected safe-haven behavior. A new driver: foreign selling of US Treasuries by China and Japan.
The 2025 Tariff Shock: A New Wrinkle
The April 2025 tariff announcements produced something that genuinely alarmed people who watch the bond market closely: Treasury yields rose during the initial equity selloff. That's backwards. When stocks crash, money is supposed to flow into Treasuries, pushing bond prices up and yields down. Instead, 10-year yields went from roughly 4.0% before the April 2 announcements to 4.5% by April 11. Bond prices fell while stocks also sold off. For a brief, deeply unsettling period, neither asset class was doing what it was supposed to do.
Morningstar's analysis of this episode is worth reading carefully. Several explanations emerged: foreign governments - China and Japan specifically - appear to have been selling US Treasuries, draining demand for bonds at exactly the moment domestic investors needed them as a safe haven. Tariff-driven inflation expectations made things worse. Investors were simultaneously worried about recession risk (which normally pulls yields down) and inflation risk (which pushes yields up). The two forces partially canceled each other out, leaving bonds behaving erratically rather than providing the reliable ballast everyone's allocation models assumed they would.
The Crucial Context That Got Buried
Despite the yield spike during peak tariff uncertainty, the fuller picture is less alarming for bond holders. Since the current market turbulence began in earnest on February 19, 2025, the Morningstar US Market Index has lost nearly 13%. Over that same period, long- and intermediate-term Treasuries have posted small gains. The recent episode was a short-term disruption in a longer-term pattern that largely held. Importantly, today's higher starting yields - the 10-year at ~4.3% versus 1.76% at the start of 2022 - provide meaningful cushion against bond price declines. This is materially different from the 2022 starting conditions, and it matters for any investor assessing bonds' current protective capacity.
So are bonds broken? No. But here's what you need to internalize: bond diversification is probabilistic and environment-dependent. It is not guaranteed. It is not mechanical. Understanding the specific conditions where bonds fail as diversifiers - high inflation combined with foreign selling pressure on US Treasuries - lets you make smarter decisions about how much to hold and when cash or other alternatives might serve you better. The key word in all of this is "conditional." If someone tells you bonds always protect against stock losses, they're either oversimplifying or they haven't been paying attention since 2021.
The Battery Concept: Why Rebalancing Still Works
Even granting everything I just laid out - the shakier historical foundation, the 2022 correlation breakdown, the 2025 tariff weirdness - there's still a powerful argument for holding bonds that has nothing to do with diversification in the textbook sense. It's the rebalancing argument. And it works through a mechanism that market historians have called the "battery concept." This is, frankly, the most compelling reason to hold bonds that nobody talks about enough.
Discharging During Crashes
When stocks fall sharply, your bond allocation holds its value while equities go on sale. Rebalancing means selling bonds and buying depressed equities - which is the thing every investor says they want to do but almost nobody can pull off emotionally without a mechanical rule forcing the trade. The bonds fund the purchase. Without them, you'd need to use earned income to buy the dip. And earned income tends to be scarce during recessions - the exact moments when stocks are cheapest.
Recharging During Bull Markets
When stocks rally and your equity allocation drifts above target, rebalancing means selling some equities and buying bonds. You're systematically taking profits at elevated prices and rebuilding your reserve for the next crash. This is selling high - literally the thing every investor claims they want to do. But without a mechanical rebalancing rule, the emotional pull of a rising market makes it nearly impossible. "Why would I sell when everything is going up?" Because the battery needs recharging. That's why.
People understate the timing dimension of this argument. Market crashes don't show up when you're flush with cash and feeling confident about everything. They show up during economic downturns - when bonuses evaporate, job security gets shaky, emergency expenses spike, and new investment dollars become scarce. The 2020 Covid crash compressed 34% of losses into five weeks. Who bought the March 2020 lows and captured that extraordinary recovery? Not the people who coincidentally received a windfall in late March. It was the people who had built reserves - bonds, cash, or both - during the bull market years that preceded it. They had a battery. And they used it.
McQuarrie's research adds another dimension that often gets overlooked: bond yields tend to rise during economic expansions and fall during recessions. That creates a natural counterbalancing dynamic. Bonds appreciate in value precisely when the rebalancing opportunity is most attractive - when stocks are cheap and you want to be buying. This isn't coincidence. It's the mechanism through which bonds earn their place in a portfolio even when their absolute returns look underwhelming in isolation. A 4% bond return sounds boring until you realize it funded your ability to buy stocks at 2009 prices.
Not All Bonds Are Created Equal
Here's a mistake I see people make constantly after episodes like 2022 or the April 2025 tariff spike: they conclude that "bonds failed." But the bonds that behaved worst - long-duration Treasuries and high-yield corporate bonds - aren't the same instruments that should be serving the battery function in your portfolio. This distinction isn't a technical detail. It's the difference between diversification that works and diversification that's an illusion. You have to know which bonds you're holding and why.
Short-to-Intermediate Government & Investment-Grade
These are the bonds that actually do what people think all bonds do. Short duration means limited sensitivity to rate changes; high credit quality means they don't suddenly behave like stocks during credit stress. Morningstar specifically names core and core-plus intermediate bond funds as the best equity ballast. This is where your battery should be built.
Long-Duration Government Bonds
Fantastic in falling-rate environments - the most valuable bonds you can own during a 2008-style deflationary recession. But they carry enormous interest rate risk when rates are rising. In 2022, long Treasuries fell 25%+. That's stock-like volatility in what's supposed to be the safe part of your portfolio. Better as a tactical position when you have a view on rates than as a permanent core holding.
High-Yield / Junk Bonds
These are bonds in name only when it comes to diversification. During market stress, high-yield bonds behave like stocks - credit spreads widen, prices drop, and they fall right alongside equities. The exact opposite of what you need from the bond allocation in your portfolio. They pay more income, sure. But when the crash comes, they provide zero protection. If you hold HYG or JNK and think you're diversified against equity risk, you're fooling yourself.
Morningstar also makes a case that I find increasingly persuasive: cash - specifically Treasury bills, the shortest-duration instruments - has been the best equity diversifier over the past three years, outperforming longer bonds precisely because T-bills are immune to the interest rate risk that wrecked bond funds in 2022. For investors who need to spend from their portfolio in the next couple of years, individual bonds held to maturity or defined-maturity "bond ladder" funds eliminate price volatility entirely. You know exactly what you're getting at maturity, and the interim price fluctuations are irrelevant because you're not selling. This is the kind of practical, unsexy strategy that actually works.
The TIPS Case - Inflation Protection That Matters
TIPS (Treasury Inflation-Protected Securities) deserve their own mention because they address the single most common objection to holding bonds: "inflation will eat my returns." TIPS adjust their principal value with CPI, so the real value of your investment is protected against inflation erosion. In a portfolio context, they behave like traditional Treasuries during deflationary stress while also protecting against the exact scenario - high, persistent inflation - where both stocks and regular bonds get hammered simultaneously. Morningstar specifically recommends TIPS as a complement to nominal bonds for anyone drawing income from their portfolio. Think of them as the one fixed-income instrument designed for the environment that breaks everything else.
The Practical Allocation Framework
So after all of this - the shakier historical case, the conditional nature of bond diversification, the real but imperfect value of the rebalancing battery - how should you actually think about allocation? The honest answer is that there's no universal answer. Anyone who gives you one is selling something. But there is a framework for thinking through the variables that matter for your specific situation.
| Investor Profile | Suggested Range | Reasoning | Bond Type |
|---|---|---|---|
| 25–35, stable income, high risk tolerance | 80–85% stocks | Long horizon, recession recovery ability, steady contributions provide some natural rebalancing | Intermediate-term Treasuries / core bond fund for the 15–20% |
| 35–50, moderate income stability | 70–75% stocks | Rebalancing benefits are highest over this horizon; portfolio size makes drawdown recovery more consequential | Mix of intermediate bonds and TIPS; cash buffer for 1–2 years expenses |
| 50–60, approaching retirement | 55–65% stocks | Sequence-of-returns risk becomes material; a bad decade early in retirement has permanently impaired outcomes | Individual bonds / bond ladders for spending needs; TIPS for inflation protection |
| Retired, drawing from portfolio | 40–55% stocks | Preservation of spending power matters; bonds fund near-term withdrawals without forced equity sales | Short-to-intermediate Treasuries for 2–3 year spending needs; growth equities for the long tail |
| Any profile, guaranteed pension income | Can increase by 5–10% | Pension acts as a bond - reduces need for portfolio bond allocation proportionally | The guaranteed income stream functions as a very long duration bond in practice |
"The 100% stocks debate is almost always framed as a question about expected returns. That's the wrong framing. It should be about optionality and timing. Bonds aren't in a portfolio primarily to reduce volatility - they're there so you have purchasing power available at the exact moments when buying equities is most valuable and your earned income is most scarce. Yes, 2022 was a genuine warning that bond diversification is conditional, not guaranteed. But the rebalancing mechanism - the battery - works regardless of what the bond-equity correlation is doing in any given year, as long as you're holding high-quality instruments that maintain value during equity stress. The lesson from 2022 is not 'avoid bonds.' It's 'hold the right bonds and size them for your specific situation.' There's a world of difference between those two conclusions."
Conclusion: One Lifetime, One Portfolio
The "stocks always win over the long run" argument has a fatal flaw baked into its framing: the long run is not a statistical average across millions of investors' lifetimes. It's your specific decades. Your specific market environment. Your specific income trajectory and spending needs. Japan's Nikkei took 34 years - thirty-four years - to recover from its 1989 peak. An investor who retired in 1999 with 100% S&P 500 had negative real returns over their first decade of retirement. The 1966 - 1981 period produced negative real returns for US equity investors. These aren't fringe scenarios that academics invented to scare people into buying bonds. They're recorded history. And they happened to real people who had real retirements to fund.
McQuarrie's work extends the lens further back and the picture gets worse for the "stocks always win" crowd: the data window from which we draw that conclusion is genuinely narrow, and the historical performance of bonds before 1926 was substantially better than the reconstructed datasets Siegel relied on suggested. We're more uncertain about the long-run equity premium than any textbook will tell you. That uncertainty alone justifies holding some bonds.
Key Takeaways
The historical case for stocks' superiority rests on less than 100 years of data and two fully independent 40-year periods. McQuarrie's work on pre-1926 bond markets shows bonds outperforming equities across multi-decade stretches in the 19th century. The confidence with which people state "stocks always win long-term" dramatically exceeds what the data actually supports. Be honest with yourself about that gap.
Bond diversification is conditional on the inflation environment, not guaranteed. 2022 proved this conclusively; the April 2025 tariff shock tested it again. If you want the battery function to work reliably across different environments, hold high-quality, short-to-intermediate duration bonds. Not long-duration. Not high-yield. Those are different instruments doing different things, and confusing them is how people end up saying "bonds failed me."
The most powerful argument for bonds isn't about return enhancement or even volatility reduction. It's about optionality at the moment of maximum opportunity - having purchasing power available when equities are dirt cheap and your earned income is hardest to come by. Cash and individual bonds held to maturity complement bond funds for anyone with near-term spending needs. You get one investment lifetime. Size the insurance accordingly.
Research Desk, PolyMarkets Investment Strategies, May 18, 2025