Gold and Bitcoin allocation: refuge strategy in 2026

Gold has served as a store of value for 5 000 years. Bitcoin claims to take up that role for the digital age, with two decisive strengths : absolute scarcity (21 million×21 millionMaximum number of bitcoins that will ever exist, hard-coded in the protocol. This programmed scarcity is a founding feature. The last sat will be mined around the year 2140.See in the lexicon →, against a gold supply growing 1.7 % per year) and immaterial portability (a memorised seed phrase×Seed phraseSequence of 12 or 24 words (usually in English) that encodes your master key. Universal wallet backup : with these words, you can restore your funds on any compatible software.See in the lexicon → crosses a border, a kilo of gold does not). By 2026, the joint gold plus Bitcoin allocation in a wealth portfolio has become a serious topic, no longer only an activist one.

The two assets do not substitute for each other. Gold has a 5 000-year track record and a negative correlation to risk assets during geopolitical crises. Bitcoin is 17 years old, has the volatility of a risk asset, but a decreasing correlation to equities since 2023 and a rising institutional respectability (spot ETF, state reserves). A typical recommended allocation : 5 to 10 % gold, 1 to 5 % Bitcoin, in a balanced portfolio.

This article retraces the history of gold as a store of value, compares the monetary properties (scarcity, divisibility, portability, verifiability, durability), exposes the current correlations to equities/bonds/USD, proposes allocation grids by wealth profile, and addresses the practical vehicles (physical gold, gold ETF, BTC self-custody×Self-custodyModel in which you hold your own private keys. Your bitcoins depend on no third party. This is Bitcoin's founding promise.See in the lexicon →, spot Bitcoin ETF).

5,000 years of gold as store of value

To understand gold's position today against Bitcoin, we must situate its history. Gold as store of value is not a modern postulate: it is an empirical fact verified over five millennia of civilisations.

• Antiquity (3000 BC - 500 AD). First gold coins in Lydia (640 BC, King Croesus). Wealth standard in Roman Empire (aureus, ~7.9 grams), then Byzantine (solidus, 4.5 grams). Gold survives Rome's fall.
• Middle Ages and Renaissance (500-1700). Florentine gold coins, Venetian ducats, French écus. Gold finances trade routes, crusades, explorations. Discovery of the Americas brings massive gold to Europe (16th century), but does not dilute its value durably.
• Classical gold standard (1816-1914). United Kingdom establishes gold standard in 1816. All major economies follow in 19th century. One sterling pound = 7.32 g of fine gold. System works for 98 years, until First World War.
• Bretton Woods (1944-1971). After Second World War, Bretton Woods agreement (July 1944) establishes US dollar as world reserve currency, convertible to gold at 35 USD/ounce. Hybrid system: national currencies tied to dollar, dollar tied to gold. Massive storage at Fort Knox (8,100 US tons).
• Nixon shock (15 August 1971). President Richard Nixon unilaterally suspends dollar-gold convertibility, to finance Vietnam War and Great Society programmes. End of Bretton Woods. Birth of the pure fiat×Fiat (fiat currency)State currency with legal tender status (euro, Swiss franc, dollar), issued by a central bank and not backed by a physical asset. By contrast, Bitcoin has an issuance capped at 21 million units, with no central issuer.See in the lexicon → monetary system we live in since. Gold becomes freely traded on markets.
• Free gold since 1971. Gold USD price: 35 USD/ounce in 1971 → 850 USD/ounce in January 1980 (x24 in 9 years, oil shock + stagflation) → 250 USD/ounce in 1999 (-70 % in 19 years, end of stagflation) → 1,900 USD/ounce in 2011 (x7.6 in 12 years, post-2008 QE) → ~3,400 USD/ounce in May 2026 (estimate, x1.8 in 15 years).

Three lessons from gold's history.

• (1) Gold preserves purchasing power but does not multiply it enormously. Over 55 years since 1971, gold made x97 in USD (35 → 3,400) while cumulative USD inflation is about x7. Gold thus beat inflation by a factor x14. Good preservation, modest performance.
• (2) Gold resists monetary crises. Each major monetary crisis (1971, 1973-1980 stagflation, 2008 GFC, 2020 COVID) was accompanied by marked gold appreciation. The negative correlation gold-vs-fiat-confidence is empirically documented.
• (3) Gold can be seized by the State. 5 April 1933: President Roosevelt's Executive Order 6102 forbids private gold holding in the United States (except collector coins and < 100 USD). Effective confiscation at the 20.67 USD/ounce official price, then devaluation to 35 USD/ounce. Private gold holding was only re-legalised in the US in 1974. Major historical precedent that Bitcoin replicates but with different resistance architecture (cryptographic self-custody×Self-custodyModel in which you hold your own private keys. Your bitcoins depend on no third party. This is Bitcoin's founding promise.See in the lexicon →).

Claire's gold (8 bars bought between 1995 and 2005) was probably acquired around 300-400 USD/ounce, and is worth today ~3,400 USD/ounce. Over 25-30 years, that's about x8-x10 performance in USD, or x4-x5 in CHF, and x3-x4 in real purchasing power after inflation. Honest but modest compared to equity markets over the same period (about x15 for S&P 500 dividends reinvested).

Bitcoin as "digital gold": the Ammous thesis

The thesis positioning Bitcoin as "digital gold" goes back to writings by Hal Finney (first Bitcoin transaction recipient in January 2009, who speaks from 2009 of monetary potential), Nick Szabo (who theorised the "bit gold" concept from 1998, direct predecessor of Bitcoin), then popularised massively by Saifedean Ammous in The Bitcoin Standard (2018).

Saifedean Ammous' argumentation.

• Sound money is rare and difficult to create. Gold became world money because it combines durability (chemically inert), divisibility, relative portability, universal recognition and above all scarcity. Any substance that can be easily produced (copper, iron, paper-fiat×Fiat (fiat currency)State currency with legal tender status (euro, Swiss franc, dollar), issued by a central bank and not backed by a physical asset. By contrast, Bitcoin has an issuance capped at 21 million units, with no central issuer.See in the lexicon →) loses its monetary role due to dilution.
• Bitcoin shares these properties but in a superior digital version. Absolute cap at 21 million×21 millionMaximum number of bitcoins that will ever exist, hard-coded in the protocol. This programmed scarcity is a founding feature. The last sat will be mined around the year 2140.See in the lexicon → BTC (vs gold that can still be mined, ~3,000 tons/year, S2F×S2F (stock-to-flow)Valuation model comparing existing stock with annual production, popularised by PlanB. Appealing to illustrate Bitcoin's growing scarcity, but empirically contested.See in the lexicon → ratio ~62), instant portability (1 seed phrase×Seed phraseSequence of 12 or 24 words (usually in English) that encodes your master key. Universal wallet backup : with these words, you can restore your funds on any compatible software.See in the lexicon → = global transfer, vs gold requiring physical transport), satoshi×Satoshi (sat)The smallest unit of bitcoin. 1 BTC = 100 million satoshis. Named after the creator. In 2026, talking in sats becomes common as the price of one BTC rises.See in the lexicon →-level divisibility (10^-8 BTC), cryptographic verification (vs gold requiring physical assay to verify×Don't trust, verifyBitcoiner mantra. Trust no one (bank, government, exchange, influencer), verify on your own through your own node.See in the lexicon → purity).
• Bitcoin solves gold's historical defects. (1) Seizability: Bitcoin self-custody×Self-custodyModel in which you hold your own private keys. Your bitcoins depend on no third party. This is Bitcoin's founding promise.See in the lexicon → is cryptographically resistant to 1933-type Executive Order 6102. (2) Storage cost: physical gold 0.1-0.5 %/year vault×VaultCustody setup for long-term storage, often multisig, kept offline and touched rarely.See in the lexicon → storage (Claire pays ~600 CHF/year to UBS for her 8 bars). Bitcoin self-custody: 70 EUR hardware wallet×Hardware walletSmall dedicated device (Ledger, Trezor, Coldcard, BitBox, etc.) that keeps the private key away from a potentially compromised computer. Signs transactions inside the device itself.See in the lexicon → amortised lifetime. (3) Transaction cost: sending 1 million USD of gold = international secure transport, several days, ~2-5 % fees. Sending 1 million USD of BTC = 1 Lightning transaction of 10-100 sats, instant.
• Bitcoin has a higher S2F than gold since the 2020 halving×HalvingScheduled event every 210,000 blocks (roughly every 4 years) that cuts the miner reward in half. This mechanism makes Bitcoin issuance decline towards the total cap of 21 million.See in the lexicon →. Bitcoin Stock-to-Flow ratio post-2024 halving = ~121 (see Bitcoin halving cycles). Gold S2F ratio = ~62. Bitcoin is now the rarest existing good. This "scarcity of scarcity" is the central Ammous thesis argument.

Counter-arguments by Peter Schiff and gold defenders.

• Bitcoin lacks 5,000 years of history. 17 years in 2026 vs 5,000 years for gold. The test of time is not yet crossed for Bitcoin. Gold has traversed all civilisations and crises; Bitcoin has traversed only one major crisis (COVID 2020, and the 2022 Ukraine war where it behaved more like a risk asset than a refuge).
• Bitcoin is intangible. Physical gold is psychologically reassuring (one can touch, see, bequeath in hand). Bitcoin is abstract, dependent on global IT infrastructure, vulnerable to extended power failures or national cyberattacks (unlikely but theoretically possible).
• Bitcoin remains correlated to risk assets. During the FTX×FTXCentralised exchange that collapsed spectacularly in November 2022. Sam Bankman-Fried was convicted. Dragged Blockfolio and many funds down with it.See in the lexicon → crisis (November 2022) and the SVB crisis (March 2023), Bitcoin fell at the same time as tech stocks, not like gold which rose. Bitcoin has not yet proven its refuge status in all contexts.
• Limited universal acceptance. Indian wedding: gold is given, not Bitcoin. Chinese wedding: same. Latin American cultures: family gold passed over 4-5 generations. Bitcoin does not yet have this anthropological cultural penetration. It's ongoing, but not yet acquired.

Synthesis position in 2026. The "Bitcoin is better than gold" thesis is plausible but not empirically demonstrated over sufficient duration. The "Bitcoin and gold are complementary" thesis is more defensible: gold brings 5,000-year stability and tangibility; Bitcoin brings portability and upward asymmetry. The absolute debate (one OR the other) is mainly ideological; pragmatic 2024-2026 net worths mix both.

Gold-Bitcoin correlation and risk profile asymmetry

Before comparing assets on 8 criteria (section 5) and performance (section 6), we must understand how gold and Bitcoin behave relative to each other. Correlation and volatility asymmetry condition allocation strategy.

Weekly gold-Bitcoin correlation (2015-2026).

• Over the whole 2015-2026 period: weekly Pearson correlation coefficient ~0.15-0.25. Weakly positively correlated. Far from strong correlation (> 0.7) that would make Bitcoin a substitute for gold.
• By sub-periods: 2015-2019 (before spot ETFs): ~0.05 (near-zero correlation). 2020-2022 (COVID + post-COVID inflation): ~0.3 (notable positive correlation, both perceived as refuges against monetary dilution). 2023-2024 (pre-2024-halving×HalvingScheduled event every 210,000 blocks (roughly every 4 years) that cuts the miner reward in half. This mechanism makes Bitcoin issuance decline towards the total cap of 21 million.See in the lexicon → cycle): ~0.1. 2025-2026 (post-halving, ETF): ~0.2 (slight increase due to institutional portfolios often holding both).
• Interpretation: gold and Bitcoin are not substitutes, they are mutual diversifiers. They can rise together in inflation crisis, fall together in liquidity crisis, or move in opposite directions depending on context. This moderate correlation is exactly what justifies holding both.

Annualised volatility asymmetry.

• Gold: historical annualised volatility 15-20 %. Maximum drawdowns ~-40 % (1980-1982 and 2011-2015). Typical recovery in 5-10 years.
• Bitcoin: annualised volatility 60-80 % (dropping from 100 % to 60 % with maturation). Maximum drawdowns -80 % to -86 % (4 out of 4 cycles). Recovery in 2-4 years.
• Bitcoin/gold volatility ratio: about x4. For equivalent risk exposure, you must hold 4 times less Bitcoin than gold in value. If Claire wants refuge exposure equivalent to 200,000 CHF of gold, the risk-equivalent in Bitcoin is ~50,000 CHF.

Annualised return asymmetry.

• Gold 2014-2026: ~8 %/year annualised in USD (from 1,200 to 3,400 USD/ounce in 12 years = x2.8, so (2.8)^(1/12) - 1 ≈ 8.9 %).
• Bitcoin 2014-2026: ~50 %/year annualised in USD (from 600 to 92,000 USD/BTC in 12 years = x153, so (153)^(1/12) - 1 ≈ 51 %).
• Bitcoin/gold return ratio: about x6 over this period. Bitcoin generated 6 times more annualised return than gold over 12 years, at the price of 4 times more volatility. The return/risk ratio is therefore favourable to Bitcoin over 2014-2026, but the future may differ (decreasing multipliers per Bitcoin halving cycles).

Behaviour in major crises (refuge test).

• COVID March 2020: gold ~-12 % (liquidity panic) then +25 % in 6 months. Bitcoin ~-50 % in 3 days, then +500 % in 18 months. Gold classic refuge. Bitcoin not refuge but asymmetric opportunity.
• Post-COVID inflation 2021-2022: gold ~+8 % in 18 months. Bitcoin -77 % in 18 months. Gold refuge against moderate inflation. Bitcoin not refuge in this specific context.
• SVB crisis March 2023: gold +10 % in 1 month. Bitcoin +30 % in 1 month. First episode where Bitcoin clearly played the refuge role against banking risk.
• Post-2026-halving bear: gold stable around 3,400 USD/ounce. Bitcoin -16 % from ATH×ATH (All-Time High)Highest historical price of an asset. Breaking the previous ATH is a strong psychological signal in Bitcoin cycles.See in the lexicon →. Gold passive refuge, Bitcoin in cycle correction.

Partial conclusion. Gold and Bitcoin are not functionally equivalent. Gold is a proven refuge against moderate inflation and slow monetary crises. Bitcoin is an upward asymmetry asset that behaves sometimes as risk asset, sometimes as refuge depending on crisis type. For Claire in 2026, keeping the gold provides the "safe refuge" leg of the portfolio. Adding Bitcoin provides the "patrimonial asymmetry" leg with increased risk but superior return potential.

Central table: 8 comparison criteria gold vs Bitcoin

Structured comparison on the 8 fundamental monetary properties. For each criterion, quantified or qualitative advantage.

Reading the table. Score: Bitcoin advantage on 4 criteria (scarcity, portability, divisibility, seizability), gold advantage on 4 criteria (durability, recognition, volatility, history). Raw tie on 8 criteria. The result depends on investor priorities.

Weighting by profile.

• Conservative elderly investor (Claire case): heavily weights recognition, durability, volatility, history. Gold-Bitcoin score = 4-2 or 5-1 depending on weightings. Gold largely wins.
• Tech-progressive investor: heavily weights portability, absolute scarcity, seizability, divisibility. Gold-Bitcoin score = 1-3 or 0-4. Bitcoin wins.
• Balanced investor (most common): neutral weighting, gives 4-4 or 3-5. Concludes complementarity, 50/50 or 60/40 mix.

Quantified 2014-2026 compared performance

Beyond theoretical principles, let's compare actual observed returns over 12 years, in USD to neutralise currency effects.

Gold performance 2014-2026:

• 1 January 2014: ~1,200 USD/ounce. May 2026: ~3,400 USD/ounce. Raw performance: x2.83 over 12 years.
• Annualised: (2.83)^(1/12) - 1 ≈ 8.9 %/year.
• Maximum intermediate drawdown×DrawdownDecline from a previous peak. Bitcoin has gone through several drawdowns of more than 75 percent in its history. To factor into your psychological planning.See in the lexicon →: 1,700 USD January 2015 then 1,050 USD December 2015 (-13 % in 1 year), then slow recovery. No -50 % drawdown over period.
• Honest performance comparable to S&P 500 over 2014-2026 (~10 %/year dividends reinvested), with less volatility (gold vol ~17 % vs S&P vol ~16 %, surprisingly close).

Bitcoin performance 2014-2026:

• 1 January 2014: ~600 USD/BTC. May 2026: ~92,000 USD/BTC. Raw performance: x153 over 12 years.
• Annualised: (153)^(1/12) - 1 ≈ 51 %/year.
• Intermediate drawdowns: -85 % in 2015 (152 USD bottom), -84 % in 2018 (3,200 USD bottom), -77 % in 2022 (15,700 USD bottom), ongoing in 2026 (-16 % from 110k USD).
• Performance crushes all assets over this period, but with volatility 4-5x higher. Someone panicked and sold at 2018 or 2022 bottom would have realised massive losses.

Performance weighted by gold-Bitcoin mix (over 12 years 2014-2026).

• 100 % gold: +183 % cumulative. 10k USD → 28k USD.
• 70 % gold, 30 % Bitcoin: +4,700 % weighted cumulative (gold x2.83 × 70 % + Bitcoin x153 × 30 % = 1.98 + 45.9 = x47.9). 10k USD → 479k USD.
• 50 % gold, 50 % Bitcoin: +7,800 % cumulative. 10k USD → 780k USD.
• 30 % gold, 70 % Bitcoin: +10,900 % cumulative. 10k USD → 1.09 M USD.
• 100 % Bitcoin: +15,200 % cumulative. 10k USD → 1.53 M USD.

Reading. Over 2014-2026, the higher the Bitcoin share, the higher the final return. But with a volatility traversed that would have panicked most investors at 2018 or 2022 bear.

Risk-adjusted performance (approximate Sharpe ratio).

• 100 % gold: excess return 8 % (above risk-free rate ~1 % average over period), vol 17 %, Sharpe ~0.4.
• 100 % Bitcoin: excess return 50 %, vol 70 %, Sharpe ~0.7.
• 50/50 mix: weighted return ~30 %, weighted vol (with low correlation) ~40 %, Sharpe ~0.75.
• 70 % gold, 30 % Bitcoin mix: weighted return ~22 %, weighted vol ~25 %, Sharpe ~0.84. Best return/risk ratio over period.

Crucial limit. These figures are retrospective. The 2014-2026 period included Bitcoin's transition from microcap (~10 Bn USD cap) to megacap (~1,800 Bn USD cap). This exceptional rise will not repeat over 2026-2038 (decreasing multipliers, see Bitcoin halving cycles). More modest projections for 2026-2038: gold 7-9 %/year annualised, Bitcoin 15-25 %/year annualised (optimistic consensus), still favourable to a mix with a dose of Bitcoin, but with a less crushing advantage than historically.

5 gold-Bitcoin allocation strategies by profile

Here are 5 mix strategies observed in pragmatic 2024-2026 portfolios, with their logic, target profile and traps.

Strategy 1: 100 % gold (classic conservative).

• Logic: keep proven refuge position, refuse Bitcoin risk perceived as speculative. Profile: retirees without Bitcoin-friendly heirs, risk-averse profiles, or "Bitcoin will stay niche" conviction.
• Advantage: maximum emotional stability, no FOMO×FOMO (Fear Of Missing Out)Fear of missing the rally, which pushes people to buy at the worst moment, near the tops. DCA is the classic antidote.See in the lexicon →, simple taxation, classic transmission.
• Drawback: moderate expected return (~7-9 %/year), zero upward asymmetry, vulnerability to a future where Bitcoin would replace gold as dominant refuge.
• For Claire: keep 200,000 CHF of inherited gold intact. Strategy consistent with her 62-year-old retiree profile, but she misses Bitcoin asymmetry if her conviction is solid.

Strategy 2: 70/30 gold-Bitcoin mix (cautious transition).

• Logic: keep proven refuge majority, add a Bitcoin dose to capture asymmetry without disproportionate exposure. Profile: conservative investors open to innovation.
• Advantage: compound benefit of best historical Sharpe ratio (~0.84 over 2014-2026), majority refuge exposure, sufficient Bitcoin dose to participate in bullish cycles.
• Drawback: Bitcoin fraction remains modest, will not significantly move net worth in case of major bullish cycle. Partial renunciation of asymmetry.
• For Claire: keep 140,000 CHF of gold (5.6 bars), sell 60,000 CHF of gold to buy Bitcoin via spot ETF (see Bitcoin spot ETF comparison) or directly via self-custody×Self-custodyModel in which you hold your own private keys. Your bitcoins depend on no third party. This is Bitcoin's founding promise.See in the lexicon →. Total net worth allocation: 8.3 % gold, 2.5 % Bitcoin (on 2.4 M CHF). Conservative but open.

Strategy 3: 50/50 gold-Bitcoin mix (modern balance).

• Logic: recognition that gold and Bitcoin are functionally complementary without obvious hierarchy. Profile: balanced investors, age 40-55, built but cautious Bitcoin conviction.
• Advantage: maximum refuge risk diversification, balanced exposure to upward asymmetry.
• Drawback: intermediate weighted performance, may seem undecided. Refuge portfolio volatility much higher than pure 100 % gold.
• For Claire: keep 100,000 CHF of gold, shift 100,000 CHF to Bitcoin. Total net worth allocation: 4.2 % gold, 4.2 % Bitcoin (8.3 % total refuge). Pragmatic balanced strategy.

Strategy 4: 30/70 Bitcoin-gold mix (advanced Bitcoin conviction).

• Logic: solid Bitcoin conviction, considers the "superior digital gold" thesis materialises, but keeps gold as insurance against total Bitcoin failure (unlikely but non-zero). Profile: convinced investors, age 30-50, 10+ year horizon.
• Advantage: majority exposure to Bitcoin upward asymmetry, retains gold safety net.
• Drawback: high refuge portfolio volatility, potential -50 % drawdowns in Bitcoin bear hard to absorb emotionally.
• For Claire: keep 60,000 CHF of gold, shift 140,000 CHF to Bitcoin. Probably too much for her 62-year-old profile, unless very strong conviction and volatility tolerance.

Strategy 5: 100 % Bitcoin (maximum conviction).

• Logic: radical conviction that Bitcoin supplants gold long term, refuses to keep the old monetary world. Profile: long-time convicted, MicroStrategy×MicroStrategy (Strategy)US company led by Michael Saylor, which has made bitcoin its main treasury asset since 2020. More than 400,000 BTC accumulated by 2025.See in the lexicon →-style, young investors (20-35).
• Advantage: maximum exposure to Bitcoin asymmetry, total ideological alignment.
• Drawback: concentrated risk, -80 % drawdowns to traverse, loss of gold diversifier, loss of family culture (Claire case: selling her father's gold is emotionally costly).
• For Claire: not recommended given her profile and the inherited gold's emotional value.

Recommendation for Claire: strategy 2 (70/30 gold-Bitcoin) with progressive rise toward 50/50 over 24-36 months. Phase 1 (June-December 2026): keep 200,000 CHF of gold, add 50,000 CHF Bitcoin via IBIT spot ETF (total allocation 8.3 % gold, 2.1 % Bitcoin). Phase 2 (2027-2028): if conviction confirms, sell 50,000 CHF of gold to bring Bitcoin to 100,000 CHF (60/40 gold-Bitcoin in refuge pocket). This progressive rise allows testing volatility tolerance before engaging larger amounts. Compatible with her cautious banker director temperament.