All Seasons (Robbins)
Tony Robbins published this five-sleeve allocation in Money: Master the Game in 2014, after Ray Dalio described a version of All Weather that a reader could hold with index funds. It is a different portfolio from the one this site calls All-Weather, and the difference decides what it does when inflation rises.
Forward-looking estimates over a long horizon, not a measurement of any past period. Built from J.P. Morgan’s 2026 Long-Term Capital Market Assumptions, annualized and compounded, with a 3.10% risk-free rate. What this portfolio actually returned, month by month from 1970, is further down the page. For Ray Dalio’s All-Weather, which is a different allocation, see its own record.
Allocation
What if you add Bitcoin?
These are forward-looking estimates rather than a backtest. Nothing below is what Bitcoin did; it is what the portfolio would return if every asset delivered its expected return from here. Each row funds Bitcoin by reducing the other positions proportionally.
The assumption doing the work is Bitcoin at 15.00% a year with 42.5% volatility. That is far below its history and deliberately so: a forward estimate for an asset this young is a judgment, not an extrapolation. It also explains why the returns below move less than people expect. A 10% position in a 15% asset can only add about a point a year to a portfolio, because it is 10% of the portfolio. What changes more than the return is the Sharpe ratio, which is the column worth reading.
| Portfolio | Return | Volatility | Sharpe |
|---|---|---|---|
| Base (All Seasons (Robbins)) | 5.33% | 8.32% | 0.27 |
| With 5% Bitcoin | 5.81% | 8.68% | 0.31 |
| With 10% Bitcoin | 6.30% | 9.50% | 0.34 |
Returns are geometric (compound) and forward-looking, built from J.P. Morgan’s 2026 Long-Term Capital Market Assumptions with Bitcoin at 15.00% (our own estimate). Sharpe ratio uses a 3.10% risk-free rate. No time period is involved: these are expectations for a long horizon rather than a measurement of any past window. For what this portfolio actually did, month by month, see the record back to 1970 below.
Where the All Seasons portfolio comes from
Tony Robbins interviewed Ray Dalio for Money: Master the Game, published in 2014. Bridgewater's All Weather strategy is not something an individual can buy, so Robbins asked for a version a reader could hold with index funds. What the book printed is a fixed five-sleeve allocation, rebalanced once a year: 30% stocks, 40% long-term US Treasuries, 15% intermediate-term US Treasuries, 7.5% gold, 7.5% commodities.
Robbins names the S&P 500 for the stock sleeve and Treasuries at two maturities for the bonds. The reasoning he attributes to Dalio is that a portfolio should hold something built for each of four conditions: growth above expectations, growth below expectations, inflation above expectations, and inflation below expectations. Stocks carry the first, long Treasuries the second and fourth, gold and commodities the third.
Robbins called it All Seasons and Bridgewater's strategy is called All Weather. The two names get used for each other constantly, including by pages that publish one portfolio's weights under the other's name. They are different allocations, and holding one while reading about the other is how an investor ends up surprised by a year like 2022.
How All Seasons differs from All-Weather
This site's All-Weather holds seven sleeves and All Seasons holds five. Two of the differences change how the portfolio behaves when inflation rises.
| Sleeve | All Seasons | All-Weather (this site) |
|---|---|---|
| Equities | 30% US | 30% global |
| Long US Treasuries | 40% | 15% |
| Intermediate US Treasuries | 15% | 15% |
| Inflation-linked bonds | none | 15% |
| Gold | 7.5% | 7.5% |
| Commodities | 7.5% | 7.5% |
| Cash | none | 10% |
Bridgewater's own strategy is a third thing again. It balances risk contributions rather than fixed dollar weights and levers the result to a volatility target, which neither column above does.
The bet is 55% in nominal Treasuries
The missing inflation-linked sleeve is the first difference. Those bonds pay a coupon that rises with the consumer price index, which is the one thing a nominal bond cannot do. The buyable version of Bridgewater's strategy, State Street's ALLW fund launched in March 2025, holds inflation-linked bonds at roughly the size of its equity position. All Seasons holds none, and the job falls to the 15% split between gold and commodities.
The second difference is duration. All Seasons puts 40% in long-term Treasuries where this site's All-Weather puts 15%, funded from the inflation-linked sleeve and the cash. A bond's sensitivity to interest rates comes from its duration, so moving a quarter of the portfolio from a five-year maturity to a twenty-year one changes the risk by much more than the weight suggests.
Across both Treasury sleeves, 55% of All Seasons sits in bonds whose only defence against inflation is the yield they were bought at. The comparable figure for this site's All-Weather is 30%, with a further 15% in inflation-linked bonds and 10% in cash. That 55% is the position every other number on this page runs through, and the record since 1970 below shows what happened to it when rates rose sharply.
What All Seasons is expected to return now
The three figures at the top of this page are forward-looking. They take J.P. Morgan's 2026 Long-Term Capital Market Assumptions for each of the five sleeves and combine them through the full correlation matrix, giving 5.33% expected return at 8.32% volatility and a Sharpe ratio of 0.27. No past period is being measured.
One assumption carries most of that. J.P. Morgan puts US long Treasuries at 4.90% a year with 13.02% volatility, and All Seasons holds 40% of them. That sleeve supplies 1.96 of the 5.33 points of expected return and 44% of the portfolio's variance. The 30% equity sleeve supplies 2.01 points and 40%. If the long-Treasury assumption is wrong, this page is wrong.
On the same assumptions this site's All-Weather is expected to return 5.15% at 7.30% volatility, and a 60/40 5.95% at 10.63%. All Seasons sits between the two on both counts, which is what a portfolio holding more duration and less cash than one and less equity than the other should do.
These forward figures are well below the measured record further down this page, for a reason that has nothing to do with the allocation. A bond's return over a long horizon is anchored to the yield it was bought at, and the 1970 to 2025 window contains forty years of falling yields, over which long Treasuries compounded at 9.21% a year. Starting from today's yield, J.P. Morgan's estimate for the same asset is 4.90%.
How All Seasons performed since 1970
Every sleeve of All Seasons has a published series reaching January 1970, so the record below substitutes nothing, which is unusual for a fifty-year backtest. The All-Weather reconstruction beside it has to stand something in for inflation-linked bonds, which did not exist in the United States until 1997, and for non-US equities before 1975. All Seasons holds neither, and is the cleaner of the two to measure.
The window runs to May 2025, 665 months in all, because that is where the commodity series ends. Every row is rebalanced monthly and priced on the same months, so the only thing being compared is the allocation.
| Portfolio | Return | Volatility | Sharpe | Worst fall | 2022 |
|---|---|---|---|---|---|
| All Seasons (Robbins) | 8.76% | 7.76% | 0.56 | -19.44% | -16.62% |
| All-Weather (this site's version) | 8.29% | 6.39% | 0.60 | -19.26% | -9.19% |
| 60/40 | 8.44% | 8.82% | 0.45 | -29.16% | -12.34% |
| US equities | 10.86% | 15.87% | 0.40 | -50.31% | -19.92% |
January 1970 to May 2025, 665 months, rebalanced monthly. Return is the compound annual rate. Sharpe uses the realized Treasury bill return over the same months rather than a fixed constant. The worst fall is measured month end to month end, so an intra-month low deeper than the figures above will not appear. $10,000 invested at the start became $1,049,611 in All Seasons, $891,628 in a 60/40 and $3,023,032 in US equities.
All Seasons compounded at 8.76% a year, which is 0.47 percentage points above this site’s All-Weather over the same months and 1.37 percentage points above it in volatility. Per unit of risk it finished behind, at 0.56 against 0.60. Against a 60/40 it earned more over the full window at lower volatility and a shallower worst fall, and lost more in 2022.
One input carries that 8.76%. The window contains the four decades from 1982 to 2021 in which long-term Treasury yields fell almost without interruption, and over those years long Treasuries alone compounded at 9.21% a year. Over the inflationary stretch before them, 1970 to 1981, the same bonds returned 3.35%. A portfolio holding 40% of them inherits whichever of those two eras it is measured over.
The counterfactual makes the size of the bet visible. Holding the same five weights with the 40% long sleeve moved to a five-year maturity gives 8.37% a year at 6.38% volatility, with a worst fall of -16.03%. Duration bought 0.39 percentage points of extra return for 1.38 percentage points of extra volatility and 3.42 percentage points of extra drawdown.
Why 2022 was the worst year in the record
The deepest fall All Seasons took in 55 years ran from December 2021 to September 2022, a loss of 19.4%. It took 23 months from the trough to make it back, in August 2024. The second deepest was the financial crisis, at 14.5% into February 2009, which is 4.9 points shallower.
The calendar year cost 16.6%, against 12.3% for a 60/40 and 9.2% for this site’s All-Weather. A portfolio built to hold up in every economic condition lost more in 2022 than the conventional allocation it is meant to improve on.
| Sleeve | 2022 return |
|---|---|
| US equities (30%) | -19.92% |
| Long Treasuries (40%) | -25.95% |
| Intermediate Treasuries (15%) | -9.31% |
| Gold (7.5%) | 0.44% |
| Commodities (7.5%) | 18.72% |
Calendar year 2022, each sleeve held on its own. Weights in brackets are the All Seasons weights.
Long-term Treasuries fell 26.0% over the year, further than US equities’ 19.9%. The 40% sleeve intended to cushion an equity fall was the largest single loss in the portfolio, because the shock was an inflation shock and a nominal bond has no defence against one. Gold returned 0.44% and commodities 18.72%, but at 7.5% each they could not carry a portfolio that had 55% in the asset doing the damage.
This site’s All-Weather lost 9.2% in the same year, holding 15% in long Treasuries instead of 40%, plus inflation-linked bonds and cash. Over the full window it finished 0.47 percentage points behind All Seasons on return and 1.37 percentage points below it on volatility. Read that row with one caveat: inflation-linked bonds cannot be sourced before 1997, so the All-Weather line holds that sleeve as half long Treasuries and half commodities throughout, and its equity sleeve is global rather than US. The All Seasons line has no substitution of either kind.
What this backtest is not
The equity sleeve is the CRSP value-weighted whole US market, which includes small and mid caps the S&P 500 Robbins names does not. Treasury returns are derived by repricing a par bond as the Federal Reserve’s published constant-maturity yield moves, because the Fed publishes yields rather than returns. Gold is the London PM fixing, price only, with no lending income. Commodities are AQR’s equal-weighted futures index, which carries far less energy than the production-weighted indices most commodity funds track.
Every series is sampled at month end. Until August 2026 the Treasury and gold series here averaged the business days in each month instead, which understated their volatility and manufactured autocorrelation, so any figure quoted from an older copy of this dataset has the returns about right and the risk wrong.
Robbins specifies rebalancing once a year and this engine rebalances monthly. Rebalancing every December over the same window gives 8.97% a year rather than 8.76%, a difference of 0.21 percentage points. The monthly figure is the one in the table because every other row is rebalanced the same way, which keeps the comparison honest even where it costs All Seasons a fifth of a point.
None of this is a fund’s track record. There are no fees, no bid-offer spreads and no taxes in these numbers, and the five ETFs that would hold this allocation did not all exist until November 2006. The same engine and the same dataset produce the fifty-year All-Weather record in the All-Weather article, including the check that reproduces Bridgewater’s own published figures to within a few tenths of a point.
How to hold All Seasons with ETFs
The allocation can be held as written with five US-listed funds. These are the ones this site prices each asset class from, so the weights and the record above describe the same thing an investor would actually buy.
| Sleeve | Weight | US-listed fund | Priced here from |
|---|---|---|---|
| US stocks | 30% | SPY (S&P 500) | February 1993 |
| Long-term Treasuries | 40% | TLT (20 year) | August 2002 |
| Intermediate Treasuries | 15% | IEF (7 to 10 year) | August 2002 |
| Gold | 7.5% | GLD (bullion) | December 2004 |
| Commodities | 7.5% | DJP (broad futures) | November 2006 |
Dates are the first month of price history this site holds for each fund. The five only exist together from November 2006, which is why the 1970 record above is built from index data instead. VTI is the usual alternative for the stock sleeve, holding the whole US market rather than the S&P 500.
Holding it from outside the United States
The commodity sleeve is where the choice of fund changes the answer. Broad commodity indices differ mostly in how much energy they carry: the S&P GSCI runs about 60 percent energy where the Bloomberg index DJP tracks caps it near a third. Two funds both described as broad commodities can behave differently in an oil shock, which for a 7.5% sleeve meant to cover inflation is the whole point of holding it.
European investors cannot buy most US-listed ETFs, and UCITS versions of all five indices exist from several issuers. A UCITS tracker of the same index holds the same securities with the same unhedged currency exposure, and differs by fee and tracking difference of a few basis points a year. No European ticker is named here because no free data source carries reliable European exchange history, so nothing could be verified before publishing it.
One decision a European holder faces that a US holder does not is whether to take the currency risk. For a euro or sterling investor, 55% of this portfolio in US Treasuries is a dollar position as much as a duration position, and in a year when the dollar falls that sleeve can lose money while the underlying bonds gain. Hedged share classes exist and cost approximately the short-term interest rate difference between the two currencies. You can price the alternatives in the optimizer.
How these numbers are calculated
Expected returns and volatilities come from J.P. Morgan's 2026 Long-Term Capital Market Assumptions (30th edition). Portfolio risk is computed using the full 27x27 correlation matrix, not simple weighted averages. The Sharpe ratio uses 3.10% (US Cash) as the risk-free rate.
For full methodology details, see the methodology page.
Customize this portfolio
Adjust weights, add constraints, try different optimization methods.
Frequently asked questions
Is the All Seasons portfolio the same as the All-Weather portfolio?
No. All Seasons is Tony Robbins' fixed five-sleeve version: 30% stocks, 40% long-term US Treasuries, 15% intermediate-term US Treasuries, 7.5% gold and 7.5% commodities, with no inflation-linked bonds and no cash. This site's All-Weather holds seven sleeves, including 15% inflation-linked bonds and 10% cash, and puts 15% in long Treasuries rather than 40%. Bridgewater's institutional All Weather is different again: it balances risk contributions rather than fixed dollar weights and levers the result to a volatility target.
What is the All Seasons portfolio allocation?
30% stocks, 40% long-term US Treasuries, 15% intermediate-term US Treasuries, 7.5% gold and 7.5% commodities, rebalanced once a year. Robbins names the S&P 500 for the stock sleeve. The five weights sum to 100% with no cash position.
How has the All Seasons portfolio performed?
Over the 665 months from January 1970 to May 2025 it compounded at 8.76% a year with 7.76% volatility, turning $10,000 into $1,049,611. Its deepest fall was 19.4%, from December 2021 to September 2022, and it took 23 months from there to make the loss back. These are month-end figures from index data rather than from funds, which did not all exist until 2006.
How does the All Seasons portfolio compare to the S&P 500?
Over the same 665 months US equities compounded at 10.86% a year against All Seasons' 8.76%, and fell 50.3% at worst against 19.4%. Per unit of risk All Seasons was ahead, with a Sharpe ratio of 0.56 against 0.40. The equity line here is the whole US market rather than the S&P 500 itself, because that is the series that reaches 1970.
What ETFs make up the All Seasons portfolio?
A US investor can hold it with SPY or VTI for the stock sleeve, TLT for long-term Treasuries, IEF for intermediate Treasuries, GLD for gold and a broad commodity fund such as DJP. European investors can buy UCITS trackers of the same indices, which hold the same securities with the same unhedged currency exposure and differ by fee and tracking difference.
What happened to the All Seasons portfolio in 2022?
It lost 16.6% over the calendar year, more than a 60/40's 12.3%. Long-term Treasuries fell 26.0%, further than US equities' 19.9%, and the 40% position in them meant the sleeve intended to cushion equities was the largest single loss. This was the deepest fall in the whole 1970 to 2025 record.