InvestLab ResearchFoundationsNo. 03 · Oct 2026

What's the Right Long-Term Portfolio for You?

Part 1: your five options, tested since 1990. What each one earns, how far it can fall, and how often it lets you down.

Stephane RenevierInvestLab · former multi-asset portfolio manager, London

Published: 15 Oct 2026
Reading time: 24 min
Figures: 16

In Building Your Investment Policy Statement, I set out the five decisions every portfolio should start from: your investment horizon, the return you need, the losses you can live with, what you're allowed to own, and how long you can stand looking wrong. Those objectives are the yardstick for everything that follows. Without them there's no way to say whether one portfolio is better than another, so if you haven't written yours down yet, start there.

Now comes one of the most important portfolio decisions you'll ever make: choosing your strategic asset allocation.

Your strategic asset allocation (SAA) is the long-term mix of stocks, bonds, and other assets you'll hold. It's the foundation of your portfolio: the blend you'll live with through both the good years and the ugly ones.

It sounds almost too simple to matter. In reality, few decisions have a bigger impact on your investing experience. Most investors spend their time worrying about stock picks and market forecasts, when the far more important question is how much they own of each asset class in the first place. Research shows that more than 90% of how a portfolio swings around over time traces back to its asset mix, not stock picking, not timing (Brinson, Hood & Beebower, 1986).

The framework we use at InvestLab reflects this idea. Strategic asset allocation sits at the base of the pyramid because it carries most of the portfolio's weight. Tactical strategies and satellites then build on that core: they adapt as conditions change and add other sources of return, and they can make a real difference to the result. But they work best on top of a well-designed core, not as a fix for a weak one.

A pyramid with three tiers: a wide strategic core at the base (essential), an active tactical layer in the middle that adapts as conditions change, and a small satellites tier at the top. — FigureStrategic asset allocation is the widest tier, because it carries most of your risk and return. Get it right first.

That's why this decision comes first. Before deciding what you think about today's market, it's worth deciding what kind of portfolio you're trying to build. So how do you decide what your core should look like?

That's where your IPS comes in. Your return objective, investment horizon and tolerance for losses act as constraints. They don't tell you exactly what portfolio to own, but they help narrow the field.

The job of an SAA isn't to maximise returns. It's to maximise the probability that you'll achieve your objectives. Put differently, to minimise the risk of failure.

That's why there is no universally "best" portfolio. The right one for a 25-year-old saving for retirement may be entirely wrong for someone buying a house in three years. Success depends on the investor, their goals, and the constraints they face.

Here's the menu we'll walk through, at a glance. The rest of this piece explains what each option really signs you up for.

Table of the five options, July 1990 to September 2026, net of costs. 100% stocks: 8.1% a year, 15.3% volatility, worst fall −54%, 26 of 35 positive years, simulated 30-year chance of a 50%+ fall 46% and of under 2% a year 12%. Aggressive: 8.1%, 11.6%, −40%, 26 of 35, 26%, 5%. 60/40: 7.8%, 10.2%, −30%, 27 of 35, 15%, 4%. Balanced: 8.2%, 9.4%, −27%, 29 of 35, 9%, 2%. All-weather lean: 7.5%, 8.7%, −23%, 28 of 35, 6%, 2%. From January 2012 to September 2026 the all-weather lean made 6.0% a year and global stocks 11.6%. — FigureThe five options, July 1990 – September 2026, net of trading costs. The 30-year odds are simulated, so they're indicative only: they depend on the assumptions behind them.

Option 1: How about 100% equities?

Bottom line: the highest expected return, but too high a chance of a 50%+ fall or a disappointing 30 years for my goals.

For many investors, this is the obvious starting point. Stocks have historically delivered higher returns than almost every other liquid asset class. Over long periods they've rewarded investors for taking risk, and for many people a globally diversified equity portfolio is a perfectly sensible solution.

If my only objective were to maximise expected returns, and I had a very (very) long horizon, the discussion might end here. But those aren't my objectives. My goal is to maximise the probability of achieving them: roughly 7–10% a year over 30 years, while avoiding losses larger than 50% and keeping the odds of very poor long-term outcomes acceptably low. Against that yardstick, a 100% equity portfolio struggles, for two reasons.

The first issue: drawdowns

Equities are wonderful wealth-building assets, but they can be brutal companions. Large drawdowns aren't rare accidents: they're part of the package. Even counting dividends and inflation, US stocks lost almost 80% after 1929 and took about 15 years to get back to their peak, and the Nasdaq-100 took 17 years after the dot-com bust. Others waited far longer: about three decades in Japan after 1989, 37 years in Italy after 1961, and Portuguese investors who bought in 1973 are still waiting. In Russia in 1914 and China in 1949, shareholders lost everything. These weren't obscure or speculative markets: several were among the world's largest right before they fell.

Years to get back to the previous peak, after inflation and with dividends reinvested: Portugal 1973, 47+ years and still 91% below in 2020 (worst fall −98%); Japan 1989 in US dollars, 37½ years, back only in September 2026 (−72%); Italy 1961, 37 years (−86%); Sweden 1912, 33 years (−77%); Japan 1989 in yen, 28 years (−68%); Belgium 1928, 27 years (−73%); Spain 1973, 23 years (−85%); Nasdaq-100 2000, 17 years (−82%); US 1929, 15½ years (−79%); Russia 1914 and China 1949, never (−100%). — FigureThe longest waits to get back to a previous peak, after inflation and with dividends reinvested.

For me, that's a problem. To reach my long-term goals I need a portfolio I can realistically hold through difficult periods. Losing 80% of my wealth, or waiting decades to break even, falls outside that range, which is why my IPS sets a maximum acceptable drawdown of 50%.

Could global diversification solve the problem? It certainly helps: owning thousands of companies across dozens of countries is far safer than concentrating in a single market. But it doesn't eliminate equity risk. Global stocks still lost roughly 55% during the Global Financial Crisis, and there's no reason future drawdowns couldn't be larger or longer.

To get a sense of the range of outcomes, I used the IPS Risk Simulator, one of the free tools on InvestLab. Feeding in assumptions similar to a global equity portfolio (7% return before inflation, 15% volatility), it generated thousands of possible future paths. Over a 30-year horizon, close to half (46%) experienced a drawdown larger than 50%, and around one in eight (13%) suffered losses greater than 70%. Those are estimates that depend heavily on the inputs, but the conclusion is clear.

Simulated chance that global stocks fall at least this far from a peak over 30 years: more than 10%: 100%; 20%: 99.2%; 30%: 89.3%; 40%: 73.6%; 50%: 46.3%; 60%: 30.1%; 70%: 13.2%. — FigureA 50%+ loss along the way is close to a coin flip: too high for my risk tolerance.

The second issue: long-term disappointment

We're often told stocks always win in the long run. Broadly, that's true. The problem is that the "long run" can be much longer than most investors realise.

Consider Japan. The chart below shows the range of real returns Japanese equity investors experienced over holding periods from 10 to 121 years. Even after 30 years, an exceptionally long horizon, investors still faced roughly a 25% chance of losing money after inflation. To be highly confident of a positive real return, you'd have needed more than 50 years. That's an extreme example, but it makes the point: long horizons reduce risk, they don't eliminate it.

Fan chart of annualised real returns for Japanese equities by holding period from 10 to 121 years. The spread only converges to a small positive figure after roughly 50 years. — Figure*How to read it:* pick a holding period on the bottom axis, say 30 years. The chart looks at every 30-year stretch in Japan's stock market since 1900 and shows the return a year after inflation. The dark line is the typical result. The shading shows the spread: the darker the shading, the rarer and more extreme the outcome. Wherever the shading dips below the zero line, investors lost money after inflation. At 30 years, about a quarter of all stretches still lost money after inflation. Only after about 50 years is even the worst stretch above zero. Source: Credit Suisse Global Investment Returns Yearbook.

And it's not just Japan. Across more than 200 years of data and 39 developed markets, Anarkulova, Cederburg & O'Doherty (2022) found investors still faced roughly a 12% chance of ending a 30-year holding period with a real loss. My own analysis points the same way: with global-equity-like assumptions, the IPS Risk Simulator estimates roughly a 12% chance of earning less than 2% a year before inflation over the next 30 years, a 4% chance of losing money outright, and more than a one-in-four chance (28%) of earning less than 4% a year.

Simulated chance that global stocks earn less than each return a year over 30 years: below −5%: 0.1%; −2%: 1.1%; 0%: 4.2%; 2%: 12.1%; 4%: 27.9%; 6%: 51.5%; 8%: 76.1%. — FigureMore than a one-in-four chance of under 4% a year, and a real chance of missing my 2% floor.

So 100% equities is a perfectly good fit for plenty of investors. Just not for mine: the odds of falling short are simply too high.

Option 2: What about adding bonds?

Bottom line: a smoother ride for nearly the same return, but the risk is still mostly stocks, and it struggles when inflation hits.

At first glance bonds seem a strange addition: if stocks return more over time, why hold a lower-returning asset? Because maximising returns isn't the goal. Maximising the probability of hitting your objectives is. And bonds, especially "high-quality" government bonds (in quote marks, given today's fears about even the safest government debt), help with that in two ways.

First, they reduce your risk. In recessions, crises and flights to safety, high-quality government bonds have historically provided a valuable cushion, cutting the risk of severe losses and making the portfolio less dependent on any single outcome. The reason is simple: in a downturn, central banks cut rates and investors rush to safety, and both lift bond prices. That should keep working (unless inflation is the problem, more below), and the cushion is bigger when yields start high, as they do today. It's like insurance (imperfect, but real), except that, unlike home or car cover, you're actually paid a positive expected return to hold it.

Second, they may cost less return than you think, and can even add to it. A portfolio's compounded return depends not just on what it earns but on how volatile the ride is. Big losses are especially damaging, because recovering takes disproportionately large gains: a 50% loss needs a 100% gain just to break even; a 50% gain then a 33% loss leaves you flat (not the +8.5% a simple average implies). By shrinking those losses, bonds offset part of the return they appear to give up, and sometimes a diversified portfolio compounds faster than a more volatile one with a higher expected return. That's one reason diversification is called the closest thing to a free lunch in investing.

Little wonder the default strategic asset allocation for decades has been the classic 60/40: 60% stocks and 40% bonds.

A note on scope: throughout this piece I stick to portfolios any individual investor can run easily, so no leverage and as few ETFs as possible.

For my own version I make two deliberate departures. I narrow the bond sleeve to high-quality government bonds, long-duration (TLT):

Government, not credit. I want bonds for diversification: something that rallies when stocks fall. Governments do; corporate and lower-quality bonds largely don't, because their credit risk bites exactly when equities drop. Too much credit and your "diversifier" is just stocks-lite.
Long-duration (TLT). It's pacier: bigger moves, more diversification punch per dollar. With no leverage (most of us), that's more bang for your buck: real ballast from a smaller slice, leaving room for the return engine. The trade-off: more rate risk (TLT fell about 31% in 2022) and US-centric, so a defensive or local government bond is a fine swap.

Here's a simple 60% VT / 40% TLT portfolio, extended back to 1990:

The 60/40: 60% global stocks (VT), 40% long Treasuries (TLT), rebalanced quarterly. — FigureThe 60/40.

Growth of $100 and falls from peak for the 60/40 against 100% stocks, July 1990 to September 2026. Net of costs: return 7.8% vs 8.1% a year, volatility 10.2% vs 15.3%, Sharpe 0.53 vs 0.42, worst fall −30% vs −54%, 27 vs 26 positive years out of 35, $100 grew to $1,516 vs $1,693. — FigureThe 60/40 vs 100% stocks, net of trading costs.

The headline is almost too good. Over 36 years the 60/40 came close to all-stocks' return (7.8% vs 8.1% a year) with two-thirds of the volatility (10% vs 15%), a worst drawdown of −30% vs −54%, and a higher Sharpe ratio (0.53 vs 0.42). Nearly the same destination, a far smoother road: variance drag working for you. But two problems remain.

→ Follow the 60/40 on InvestLab (Lean 60/40, free with an account)

Problem one: it's far more concentrated in risk than it looks

You put 60% of your money in stocks, but they supply roughly 80% of the portfolio's risk, and the blend still moves with global equities almost one-for-one (correlation 0.87). The reason is simple: stocks are far more volatile, so they dominate whatever they're mixed with.

Two bars. Your money: stocks 60%, bonds 40%. Your risk: stocks 81%, bonds 19%. — FigureStocks are 60% of the money but about 80% of the risk.

Problem two: the insurance lapses just when inflation strikes

Bonds did their job in the dot-com bust and 2008, but not in 2022, when the 60/40 fell about as hard as all-stocks from peak to trough (−27% vs −26%) because long Treasuries fell 31%, more than equities. The diversifier became an amplifier. Diversification depends on the kind of trouble: in a recession bonds rally as rates fall; in an inflation shock central banks raise rates, hammering bonds (long-duration most) and stocks together, exactly when you needed them apart.

Option 3: What about adding more asset classes?

Bottom line: the smoothest ride and the smallest falls, at the cost of long stretches behind stocks.

By now the natural question is: why stop at stocks and bonds? The instinct is right. The trick is not simply owning more assets, but owning assets that behave differently.

Many popular additions don't help as much as investors think. Private equity and real estate, for example, often look wonderfully diversified on paper. In reality, much of that diversification comes from infrequent pricing. Under the hood they're still heavily exposed to economic growth and equity markets (in AQR's chart below, both sit right next to stocks), and in a serious crisis they tend to suffer alongside stocks rather than protect you from them.

A better starting point is to ask: what actually drives asset returns? Most market environments can be thought of as combinations of two forces: growth and inflation. Each can surprise to the upside or downside, creating four broad economic regimes. Stocks and bonds already cover two of them reasonably well:

Stocks thrive when growth is strong and inflation stays contained.
Government bonds do best when growth disappoints and inflation falls.

The problem is that both can struggle when inflation becomes the dominant risk. That's exactly what happened in 2022. To make the portfolio more resilient, we need assets that help in the remaining environments:

Reflation (growth and inflation both rising) → commodities. Real assets (oil, metals, agriculture) often benefit directly from rising prices and strong demand. When inflation picks up, commodities are often part of the reason.
Stagflation (weak growth, high inflation) → gold. Gold doesn't depend on growth, profits or cash flows. It has historically done best when investors lose confidence in paper assets, real rates fall, or inflation erodes the value of money.

A two-by-two map with inflation from low to high left to right and growth from low to high bottom to top. Goldilocks (growth up, inflation down): stocks (VT). Reflation (growth up, inflation up): commodities (DBC). Recession (growth down, inflation down): government bonds (TLT). Stagflation (growth down, inflation up): gold (GLD). — FigureFour growth and inflation regimes, with one building block built to lead in each.

The data backs this up. AQR mapped how each asset has responded to growth and inflation surprises over more than 50 years. Stocks sit in the Goldilocks corner, Treasuries in recession, gold in stagflation and commodities where growth and inflation both run hot (AQR calls it overheating). Notice where the US 60/40 sits: right next to US equities, which is Problem one in a single dot. Private equity and real estate sit in the same corner, close to stocks.

AQR scatter of growth sensitivity against inflation sensitivity for assets and hypothetical strategies, January 1972 to June 2026. US equities and the US 60/40 sit top-left, US Treasuries bottom-left, gold and breakevens bottom-right, commodities right. — FigureSource: AQR, Exhibit 4, 'The Long View: Growth and Inflation Sensitivities of Assets and Hypothetical Strategies', January 1972 – June 2026.

Together, stocks, bonds, commodities and gold give exposure to a far broader range of outcomes than stocks and bonds alone. So should we just split the portfolio 25% each? Not quite.

As we saw with the 60/40, equal dollars doesn't mean equal risk. More volatile assets contribute more to the portfolio's behaviour, regardless of how much capital they receive. That observation sits at the heart of risk parity and Ray Dalio's All Weather approach: don't balance dollars, balance risks. There are sophisticated ways to calculate those weights, but let's keep it simple with a practical version, the all-weather lean:

The all-weather lean: 30% global stocks (VT), 30% long Treasuries (TLT), 20% gold (GLD), 20% broad commodities (DBC), rebalanced quarterly. — FigureThe all-weather lean.

Notice we're not aiming for perfect risk balance. Stocks and bonds are productive assets with positive expected returns; gold and commodities are primarily diversifiers and inflation hedges. They're there to improve outcomes, not to drive long-term wealth creation. Here's how that portfolio would have performed:

Growth of $100 and falls from peak for the all-weather lean against 100% stocks, July 1990 to September 2026. Net of costs: return 7.5% vs 8.1% a year, volatility 8.7% vs 15.3%, Sharpe 0.58 vs 0.42, worst fall −23% vs −54%, 28 vs 26 positive years out of 35, $100 grew to $1,397 vs $1,693. — FigureThe all-weather lean vs 100% stocks, net of trading costs.

The pattern should look familiar. Over 36 years this all-weather portfolio came within about half a point a year of 100% stocks (7.5% vs 8.1%), with just over half the volatility (8.7% vs 15.3%), less than half the worst drawdown (−23% vs −54%), and the highest Sharpe ratio of the portfolios we've examined so far (0.58). That's diversification earning its keep. Even more striking, gold and commodities earned less than stocks on their own (about 6.8% and 5.0% a year, against 8.1%), yet they made the whole portfolio better: their value came from improving the portfolio, not from being great standalone investments.

→ Follow the all-weather lean on InvestLab (Lean All-Weather Unlevered, free with an account)

Option 4: Want more stocks anyway? Aggressive SAA

Bottom line: stock-like returns with smaller falls, but still mostly a bet on stocks.

Fair enough: not everyone needs a portfolio that's fully balanced across every regime, and many people would find it too hard to deviate far from stocks. If that's you, there are two sensible ways to do it.

Equities with diversifiers as supporting players

The simplest approach keeps equities as the dominant part of the portfolio and uses bonds, gold and commodities as supporting players. One example:

Aggressive SAA: 70% global stocks (VT), 10% long Treasuries (TLT), 10% gold (GLD), 10% broad commodities (DBC), rebalanced yearly. — FigureAggressive SAA.

At those weights, the portfolio is still overwhelmingly driven by equities: in a major regime shift, you'll feel it. But the diversifiers can reduce drawdowns, improve resilience, and smooth the compounding. Over 1990–2026 this simple mix matched 100% stocks (8.1% a year for both) while cutting the maximum drawdown from −54% to −40%.

→ Follow it on InvestLab (Lean Aggressive, free with an account)

Option 5: Or meet in the middle: Balanced SAA

Bottom line: growth with ballast, and my starting point for many investors.

All-weather with a growth tilt

The next step is closer to an all-weather portfolio, but with a larger equity allocation than a traditional risk-parity approach. Think of it as a growth-oriented all-weather: enough stocks to drive long-term wealth creation, enough diversification to avoid being entirely dependent on one outcome. It's the shape behind our Lean Balanced strategy.

Balanced SAA: 20% US stocks (SPY), 20% developed ex-US stocks (VEA), 10% emerging-market stocks (EEM), 30% 20+ year US Treasuries (TLT), 10% gold (GLD), 10% broad commodities (DBC), rebalanced yearly. — FigureBalanced SAA.

Personally, I think this is one of the strongest starting points for many investors. It has a meaningful growth engine, but enough ballast that you're more likely to stay invested when markets get difficult. Over our sample it earned the highest return of the five, slightly ahead of 100% stocks (8.2% vs 8.1% a year), while keeping volatility low (9.4%) and drawdowns relatively modest (−27%). Here's how the two stock-heavy options compare:

Growth of $100 and falls from peak for Balanced, Aggressive and 100% stocks, July 1990 to September 2026. Net of costs: return 8.2%, 8.1% and 8.1% a year; volatility 9.4%, 11.6% and 15.3%; worst fall −27%, −40% and −54%; positive years 29, 26 and 26 of 35. — FigureBalanced and Aggressive vs 100% stocks, net of trading costs.

→ Follow Lean Balanced on InvestLab (free with an account)

How much should we trust any of this?

The numbers above describe one past, not the future, so trust the shape of the trade-offs, not the exact magnitudes. A few reasons to stay humble.

So why trust any of it? Not because of the backtest, but because of economic logic. Each asset is tied to a different engine: stocks to growth, government bonds to falling growth and deflation, commodities and gold to inflation. Hold all four and, whatever the economy throws at you, something is built to do well. Since nobody can know which regime comes next, owning assets that each thrive in a different one is simply safer: a conclusion that would hold even if markets had never produced a single data point.

Put it to work on InvestLab

Follow any of these portfolios, free: Lean Balanced, Lean All-Weather Unlevered, Lean Aggressive and Lean 60/40, with live weights and performance.
Check your own odds in the IPS Risk Simulator.
Test your own mix or create your own strategy back to 1990 in the backtesting tool. Every number in this piece can be re-run there.
Build the whole portfolio (core, tactical and satellites) in Portfolio Studio.

→ Create a free InvestLab account

For advisers, wealth managers and family offices: we run firm pilots of InvestLab, with model portfolios, client-ready reports and the full backtesting tool for your team.

Figures from InvestLab backtests, July 1990 – September 2026, monthly, in US dollars and net of trading costs. Before each ETF launched, history uses index proxies. "Chance of" figures come from the InvestLab IPS Risk Simulator (rescaled bootstrap model, seed 42) and are estimates, not forecasts. Past performance does not guarantee future returns. Educational, not personalised investment advice.

Important information

Not investment advice. This article is for education only. It is not investment advice, a personal recommendation, or an offer or solicitation to buy or sell any security or strategy, and it does not consider your circumstances, objectives or risk tolerance. Before making any investment decision, consider speaking to an adviser licensed in your jurisdiction who knows your situation.

Hypothetical performance. Every backtested figure here is hypothetical. No account achieved these results. Backtests are built with hindsight, assume trading costs by formula, ignore taxes, and use index proxies before each ETF existed. Real results would have differed, often for the worse.

Simulations are not forecasts. The 30-year odds come from a statistical model. They depend entirely on the assumptions used, and real outcomes can fall outside the ranges shown.

Risk. All investing involves risk, including the loss of the money you invest. Past performance, real or simulated, is not a reliable guide to future results. Investing in assets priced in another currency adds exchange-rate risk.

Data. Figures come from third-party data providers and academic datasets (including Kenneth French's data library, the Jordà-Schularick-Taylor Macrohistory Database, AQR and the Credit Suisse Global Investment Returns Yearbook). They may contain errors and can be revised. Third-party charts remain the property of their owners.

Interests. I founded InvestLab, and links to the platform point to products it offers, some of them paid. I may hold positions in some of the assets or funds mentioned.

Jurisdiction. This content is not directed at anyone in a jurisdiction where its publication would be unlawful or would require a licence InvestLab does not hold. Full disclosures: app.investlab.app/disclosures.

How to cite

Renevier, S. (2026). What's the Right Long-Term Portfolio for You? InvestLab Research, Foundations No. 03, 15 October 2026. www.investlab.app/research/choosing-your-strategic-asset-allocation

For information and education only. Nothing here is investment advice. Backtested results are hypothetical and shown with their assumptions; past performance does not guarantee future returns.