How does copio judge a resilient portfolio?
Copio's analytical standard is maximum expected long-term geometric growth subject to liquidity needs and losses the investor can sustain. Geometric growth measures how gains and losses multiply through time. A 50% fall requires a 100% gain to recover; a simple average of percentage returns misses that effect.
We believe cost-controlled convex protection can help preserve a productive growth allocation through severe losses and provide gains to reinvest when prices fall. The complete protected portfolio must be assessed against feasible alternatives, including one that permanently holds less equity. That comparison includes both the hedge's explicit costs and the opportunity cost of reducing growth exposure. Protection that merely smooths the ride while lowering expected compound growth does not meet our investment standard.
This is a portfolio-construction objective, not a promise of a maximum future return. A costly or badly matched hedge can reduce growth, and money needed for a near-term bill cannot be judged as though it were available for permanent investment.
Does mean-variance optimisation maximise geometric growth?
Mean-variance optimisation does not establish maximum geometric growth without specifying the objective, sizing and return assumptions. It examines expected arithmetic return, variance and covariance; the chosen risk trade-off determines the allocation. Under certain assumptions a quadratic mean-variance objective can approximate log-growth optimisation, as Busseti, Ryu and Boyd explain in section 6 of Risk-Constrained Kelly Gambling. A maximum Sharpe ratio or a chosen volatility target alone is insufficient, especially for strategies whose value depends on rare, nonlinear payoffs.
The investors in this series also use different objectives. Yale's 2015 report describes mean-variance analysis alongside market judgement. AQR publishes compound-return estimates and uses Sharpe ratios to assess scalable long-short premia. Pershing Square explicitly seeks long-term compound growth in intrinsic value per share, while Spitznagel judges a hedge by its effect on the whole portfolio's compound growth. Their case studies provide the source passages.
Diversification can increase geometric growth by reducing shared losses without giving up too much expected return. We therefore assess each allocation on its evidence; we do not infer inferior compounding simply from the use of mean-variance tools.
What makes an asset allocation conservative?
A conservative asset allocation limits exposure to losses the investor cannot afford and preserves access to money needed for spending. The relevant risk is therefore connected to a goal: a house deposit due soon faces different constraints from retirement spending decades away.
Cash and short-term, high-quality debt can provide liquidity without relying on an equity-market recovery. They still face inflation risk, and foreign-currency holdings can change value relative to the currency of a future bill. Longer bonds add sensitivity to interest rates; corporate credit adds the possibility that borrowers fail to pay.
The distinction appears in Warren Buffett's asset allocation. His 2013 shareholder letter instructed the trustee of a bequest for his wife to use 90% in a low-cost S&P 500 index fund and 10% in short-term government bonds. That is a simple allocation with high equity exposure. It is neither Berkshire's balance-sheet mix nor a conservative default for every household.
How is resilient portfolio allocation different?
For copio, resilience supports long-term compounding by addressing losses that can damage wealth or force the investor out of the plan. Cushioning a fall through lower exposure and generating gains from an explicit hedge are different mechanisms. An equity crash, a surge in inflation and an urgent withdrawal can each expose a different weakness.
| Risk | What the allocation needs to address | Where the response can fail |
|---|
| Recession and falling profits | Dependence on equities and corporate borrowers | High-yield bonds can fall alongside stocks |
| Inflation and rising rates | Purchasing power and bond duration | Cash can lose real value; long bonds can lose market value |
| Sudden equity crash | Exposure to abrupt losses | Options may not cover the relevant asset, timing or scale |
| Prolonged bear market | Ability to maintain the plan through repeated losses | Recurring hedge premiums can compound the loss |
| Cash needed unexpectedly | Saleability and withdrawal terms | Private funds can restrict access to capital |
| Currency movement | Match between holdings and future spending | A safe asset in one currency can fluctuate in another |
These risks can overlap. Our comparison of asset classes in past crashes includes the 2022 stock-and-bond sell-off, while the bond-duration guide explains why rising yields reduce bond prices.
How does cushioning a fall differ from active downside protection?
Cushioning reduces the loss carried by the portfolio; explicit downside protection seeks gains during a covered shock that can offset losses and fund reinvestment. Holding cash instead of shares mechanically reduces equity exposure, but the cash balance does not jump because the stock market crashes. Government bonds can also rise when yields fall; that gain depends on the economic shock and does not provide a contractual equity-crash payoff.
A purchased put provides a defined payoff tied to its underlying asset, strike and expiry. Its value before expiry also depends on market conditions, and it can lose the entire premium. The contractual payoff establishes what it covers; cost-effectiveness requires a separate assessment of the whole portfolio's compounding after costs. Trend following and discretionary short positions can generate gains in falling markets too, but depend on positions, timing and execution.
| Mechanism | How it responds to an equity fall | Geometric-growth test |
|---|
| Reduced equity exposure | Less capital participates in the loss | Does the lower loss and return on the reserve justify the growth forgone? |
| Diversification | Other holdings may fall less or rise | Do their combined returns and costs improve compounding across different shocks? |
| Convex crash hedge | Can generate a disproportionate gain in the covered decline | Do payouts and reinvestment outweigh premiums, costs and missed coverage? |
Is a 60/40 or All Weather portfolio conservative?
Neither label establishes a portfolio's risk level without the holdings and implementation details. A 60/40 portfolio still puts 60% of capital in stocks, while the risk in its 40% bond allocation depends on duration, credit and currency.
Ray Dalio's All Weather portfolio balances exposure to growth and inflation surprises. The ALLW ETF reported 70.7% nominal government bonds, 43.6% equities, 36.5% inflation-linked bonds and 32.7% commodities in its 30 September 2026 exposure document. Those figures total 183.5% of fund value because derivatives can add exposure beyond the capital invested. They measure notional exposure, not each asset's share of risk.
The familiar 30% stocks, 40% long-term Treasuries, 15% intermediate Treasuries, 7.5% gold and 7.5% commodities mix comes from Tony Robbins's 2014 presentation of an unlevered version. It still has substantial interest-rate exposure through its 55% Treasury allocation. The Dalio case study separates that mix from Bridgewater's private strategy and the public ETF.
What do ten leading investors teach about defensive allocation?
Their approaches address different risks, and several retain high equity exposure or use leverage. The table distinguishes their methods; our assessment is whether each supports expected geometric growth after costs.
| Investor | Approach examined | Main limitation for a household |
|---|
| Warren Buffett | Cash reserves, productive businesses and a separate 90/10 trust instruction | Berkshire's insurance obligations and controlled businesses differ from a personal portfolio |
| Seth Klarman | Cash when bargains are scarce; buying below estimated value | Historical cash levels do not supply a fixed target |
| Howard Marks | Choosing the balance between owning and lending | Credit income is not an equity-crash payout; assess defaults, liquidity and compounding |
| Ray Dalio | Balancing growth and inflation risks | A balanced risk budget does not establish maximum geometric growth or a crash payout |
| Cliff Asness | Combining value, momentum, carry and defensive styles | Sharpe ratios alone do not establish the best sizing for net geometric growth |
| David Swensen | Diversification, with mean-variance analysis and market judgement | A smoother valuation series does not supply liquid gains during a crash |
| Mark Spitznagel | Tail hedging judged by its effect on portfolio compounding | Matches our objective; reported hypothetical results still need scrutiny |
| Bill Ackman | Compound-growth objective with occasional macro hedges | One successful hedge does not establish a recurring protection policy's net value |
| Stanley Druckenmiller | Concentrated positions and rapid changes when a view fails | Depends on active decisions; public filings omit much of the macro portfolio |
| Paul Tudor Jones | Inflation trades, position sizing and trend-based exits | A historical ranking or interview does not specify a complete allocation |
Yale makes the liquidity distinction particularly clear. Its 2025-2026 Financial Report, covering the year to 30 June 2026, describes roughly 97% of the endowment pool as invested in assets expected to produce equity-like returns, up from roughly 91% in the 2024-2025 report. Diversification across private managers does not remove that economic exposure or make the holdings immediately saleable.
What do eleven more investors and institutions add?
Six further case studies cover stock pickers, a value formula and a barbell, and five cover pension funds, a sovereign fund, an endowment and a foundation. Each limitation below is documented in the linked study.
| Investor | Approach examined | Main limitation for a household |
|---|
| Benjamin Graham | Between 25% and 75% in stocks, 50/50 by default | In 2022 the 50/50 mix in our backtest still lost 17.9%, because bonds fell too |
| Charlie Munger | A few holdings kept for decades | Two banks were 88.8% of the Daily Journal's reported 13F value at 30 September 2026 |
| Nassim Taleb | 85-90% in very safe assets, the rest in speculative bets | If the bets expire worthless, their cost compounds against the whole portfolio |
| Peter Lynch | Stock picking with full equity exposure | Magellan fell 33.1% from August to November 1987, and investors' timing cost them part of its return |
| Joel Greenblatt | A rules-based value formula, now long and short | Long stretches of lagging; we estimate the fund's investors earned far less than the fund |
| Michael Burry | Concentrated positions, including large put options | A 13F values options at the underlying shares, not at what the puts cost |
| GPIF | 25% each in domestic and foreign bonds and equities, rebalanced | Its worst quarter since 2015 was −10.71%, and it has no household spending needs |
| CPP Investments | A total portfolio with private assets and leverage | Leverage was 33.0% of net assets, and private valuations lag public markets |
| ADIA | A published range for each asset class | It publishes no total assets or single-year returns |
| Harvard endowment | Mostly private equity and hedge funds | Illiquid, access-limited holdings; it lost 27.3% in fiscal 2009 |
| Gates Foundation Trust | A diversified balance sheet funding a dated spend-down | Its 13F covers only part of the portfolio; losses interact with grants and future gifts |
How much cash and credit belongs in a conservative portfolio?
The amount depends on spending needs, the time available to recover from losses and the purpose of the reserve. Cash for a known bill has a date and an amount; cash held for investment opportunities depends on a decision about prices.
Klarman's historical cash levels illustrate the second use. Press reports put Baupost at an average of 43% cash in 2015 and 26% in early April 2020 after the firm invested during the sell-off. Those observations do not establish its current cash level or a target for individuals; the Baupost study provides each dated source.
Credit needs a separate assessment. In Ruminating on Asset Allocation, dated October 2024, Marks cited prospective returns starting at roughly 7% for public credit and 10% for private credit. The contractual payments appealed to him, but a quoted yield does not deduct every possible default loss or guarantee access to the money when needed.
Can cost-effective crash protection preserve more growth exposure?
Copio's research thesis is that cost-controlled convex protection can improve the geometric growth of a productive portfolio. The mechanism is gains during covered crashes and capital for reinvestment, while retaining more growth exposure than a permanently reduced-exposure portfolio. The benefit must exceed premiums, trading costs and relevant taxes across both quiet and stressed periods.
Universa's client letters used a hypothetical mix of 3.33% in its tail strategy and 96.67% in the S&P 500 in 2018 and 2020. The reported 3,612% March 2020 return concerned capital required for the hedge; the hypothetical blended portfolio was reported at +0.4% for the month. The Spitznagel case study explains the shared compounding objective and the limits of those figures.
AQR's research finds a high cost to recurring put purchases and compares them with trend following. That is evidence the growth test must address: a hedge that consumes more expected growth than its portfolio effect adds fails our standard.
Our put-option guide checks a published 1986-2018 benchmark in which an S&P 500 portfolio protected with monthly puts about 5% out of the money cut the maximum month-end drawdown from 50.95% to 38.92% but compounded at 6.64% a year, against 9.80% for the index. One benchmark does not show that every put strategy reduces growth, so the complete construction still has to be tested rather than assuming any option purchase is cost-effective protection.
Our study of assets during seven stock-market crashes compares the periods when cash, government bonds and gold helped, and the inflation shock in which bonds fell with shares. The euro-investor diversification study checks correlations alongside losses and compounded wealth, rather than treating a low correlation as sufficient evidence.
What can an individual investor check in their own allocation?
The review starts with expected whole-portfolio geometric growth after costs, within the investor's spending needs and capacity for loss.
- Does the defensive allocation improve expected compound growth, or mainly make losses smaller?
- What spending must remain possible after a market fall?
- How much equity exposure sits inside funds as well as direct shares?
- How much interest-rate, credit and currency risk sits in the defensive allocation?
- Which assets could be sold promptly, and which have redemption limits?
- Does the hedge create gains during the losses it covers, and how would those gains be reinvested?
- Do its premiums and implementation costs leave the complete portfolio with higher expected geometric growth than the alternatives?
- Is there a rebalancing rule that can be maintained during a decline?
The asset-class correlation study helps test whether apparently different holdings have moved together. Historical behaviour can reveal concentrations; it cannot guarantee the next crisis will follow the same pattern.
FAQ
What is a conservative asset allocation?
A conservative asset allocation commonly reduces equity exposure and reserves more capital for cash or bonds. That can cushion equity losses, but it does not itself provide a contractual crash payout or establish better compound growth. Copio compares it with a growth portfolio supported by cost-controlled convex protection, using expected whole-portfolio geometric growth after costs and within the investor's constraints.
What is the difference between conservative and resilient portfolio allocation?
Conservative allocation commonly prioritises limiting losses. Copio's standard for resilient portfolio allocation is to maximise expected long-term geometric growth within spending needs and capacity for loss. Explicit protection must earn its cost through the complete portfolio's compounding, including reinvestment of gains; a smoother return path alone is insufficient.
Is a 60/40 portfolio conservative?
A 60% stock and 40% bond portfolio cushions equity losses through lower equity exposure, and bonds may also gain when yields fall. That conditional response is different from a contractual equity-crash payoff. Copio assesses whether the mix improves expected compound growth after costs compared with other feasible allocations.
Is Ray Dalio's All Weather portfolio low risk?
All Weather balances growth and inflation risks, but risk balance does not establish maximum geometric growth or cost-effective crash protection. The ALLW ETF reported 183.5% notional exposure relative to fund value on 30 September 2026. Copio evaluates its leverage, costs and loss distribution at whole-portfolio level, alongside the alternative of a growth allocation with explicit protection.
How much cash belongs in a defensive portfolio?
Cash needed for near-term spending serves a different purpose from long-term investment capital. Cash cushions equity losses by reducing exposure; it does not generate an insurance payout when stocks crash. For investable capital, copio assesses its yield, liquidity, opportunity cost and contribution to expected whole-portfolio geometric growth.
