How much bitcoin belongs in a diversified portfolio?
Educational and illustrative only; not investment advice. Covers bitcoin as an asset only, with no UTXO Management fund, service, or performance. All portfolio results are hypothetical and gross of fees. Data as of 9/30/2026. See Disclosures.
Market data and month-end performance statistics as of 9/30/2026. ≈ denotes an approximation. See Data Notes.
Over the ten years ended September 2026, bitcoin returned +63.6% annualized with a Sharpe Ratio of 0.99, versus 0.72 for a 60/40 portfolio. It ranked first among major assets in 7 of 11 calendar years.
Annualized volatility was 36% in 2025 and near 50% in 2026, down from 70% or more in every year from 2017 to 2021. The 2025-26 drawdown troughed near -53% on daily closes, versus -77% to -85% in prior cycles.
Bitcoin's upside deviation is 67% against 34% downside, and its monthly skew is +0.48 versus -0.43 for the S&P 500. Its gains have been larger than its losses.
In a hypothetical backtest, a 2.5% bitcoin position rebalanced quarterly lifted the 60/40 return from 9.7% to 12.0% and its Sharpe Ratio from 0.72 to 0.88, while maximum drawdown moved only from -20.1% to -20.9% (Oct 2016 to Sep 2026; gross of fees, costs and taxes; not actual results).
Bitcoin ("BTC") has completed a full market cycle since the launch of U.S. spot bitcoin ETFs in January 2024. The asset reached an all-time high above $126,000 (October 2025), drew down more than 50% into June 2026, and recovered to approximately $83,700 as of September 30, 2026. ETFs hold roughly $108 billion in assets, and the United States established a Strategic Bitcoin Reserve by executive order in March 2025. These developments make bitcoin worth a structured review by institutional investors.
This paper gives institutions, advisors, and capital allocators a framework for evaluating bitcoin's strategic role within a diversified portfolio.
We exclude bitcoin's low-adoption early years. Over the 10 years ended 9/30/2026, bitcoin returned 63.6% annualized with a 72.2% standard deviation, for a Sharpe Ratio of 0.99. A dollar invested grew to roughly $137, against $4.14 in the S&P 500 and $2.54 in a 60/40 portfolio (Exhibit B).
Despite severe drawdowns in 2018, 2022, and 2025-26, bitcoin outperformed every other major asset class in 7 of the past 11 calendar years, finished last in three of the others, and ranks last of five in 2026 year to date (Exhibit C).
| Asset | Ann. return | Std dev | Sharpe | Sortino | Max drawdown |
|---|---|---|---|---|---|
| Bitcoin | +63.6% | 72.2% | 0.99 | 2.07 | -74.6% |
| S&P 500 | +15.3% | 15.3% | 0.85 | 1.34 | -23.9% |
| Bloomberg U.S. Aggregate | +1.1% | 5.2% | -0.24 | -0.32 | -17.0% |
| 60 S&P 500 / 40 U.S. Agg | +9.7% | 10.2% | 0.72 | 1.11 | -20.1% |
| Gold | +11.8% | 15.5% | 0.64 | 1.11 | -23.5% |
| Nasdaq 100 (QQQ) | +20.9% | 19.0% | 0.97 | 1.64 | -32.6% |
Exhibits A and B. Monthly total returns, Oct 2016 to Sep 2026. Bitcoin: ChartRow BTC-USD month-end closes; SPY, AGG, GLD, QQQ as proxies, distributions reinvested. 60/40 rebalanced quarterly. Risk-free: 3-month T-bill. Standard deviation is the annualized dispersion of monthly returns around their average. Max drawdown is the largest peak-to-trough decline on month-end values. Sharpe Ratio is return in excess of T-bills per unit of standard deviation; Sortino Ratio is the same excess return per unit of downside deviation, so it penalizes only returns below the T-bill rate. Source: ChartRow, UTXO Management.
| Rank | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | '26 YTD |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1st | BTC +125 | BTC +1,330 | AGG +0.4 | BTC +88.1 | BTC +302 | BTC +59.5 | GLD -0.9 | BTC +155 | BTC +122 | GLD +63.7 | QQQ +20.8 |
| 2nd | SPX +12.0 | QQQ +32.7 | QQQ -0.1 | QQQ +39.0 | QQQ +48.6 | SPX +28.6 | AGG -13.0 | QQQ +54.9 | GLD +26.9 | QQQ +20.8 | SPX +12.8 |
| 3rd | GLD +8.6 | SPX +21.7 | GLD -2.1 | SPX +31.0 | GLD +24.9 | QQQ +27.4 | SPX -18.2 | SPX +26.4 | QQQ +25.6 | SPX +17.7 | AGG -2.9 |
| 4th | QQQ +7.1 | GLD +12.7 | SPX -4.5 | GLD +18.0 | SPX +18.6 | AGG -1.7 | QQQ -32.6 | GLD +12.8 | SPX +24.9 | AGG +7.1 | GLD -3.9 |
| 5th | AGG +2.6 | AGG +3.4 | BTC -72.3 | AGG +8.6 | AGG +7.5 | GLD -4.0 | BTC -64.2 | AGG +5.6 | AGG +1.2 | BTC -6.5 | BTC -4.6 |
Exhibit C. Calendar-year total returns (%), ranked. BTC bitcoin, SPX S&P 500, QQQ Nasdaq 100, GLD gold, AGG Bloomberg U.S. Aggregate. 2017 onward from monthly data; 2016 from public annual data; 2026 through 9/30. Source: ChartRow, UTXO Management.
Bitcoin's standard deviation, above 60% in most calendar years of its first decade, has long given investors pause. That gap has narrowed materially. Annualized volatility of monthly returns fell to 36% in 2025 and has run near 50% through the 2026 drawdown, against 70% or more in every year from 2017 to 2021. Several of the largest S&P 500 constituents were more volatile: Tesla, with a market value above $1 trillion, had annualized volatility of roughly 55% on monthly returns in 2025, against bitcoin's 36%.
Two structural changes are consistent with the compression. First, the holder base has shifted toward regulated vehicles and long-duration balance sheets. In particular, spot ETFs and corporate treasuries trade less and lever less than the offshore participants who set prices in prior cycles. Second, the market's aggregate cost basis now sits far closer to the cycle peak. Realized price, the average price at which every coin in circulation last moved on-chain and therefore a proxy for the market's aggregate cost basis, was roughly 44% of the October 2025 high, versus about 25% at prior tops, leaving less unrealized profit to be liquidated in a downturn.
The result is visible in the drawdown. The 2025-26 decline bottomed near -53% on daily closes in June 2026, compared with -77% to -85% in each of the three prior cycles. Bitcoin remains a high-volatility asset, though recent cycles suggest the range of outcomes has narrowed.
| Dec 2013 · Jan 2015 | -85% |
| Dec 2017 · Dec 2018 | -84% |
| Nov 2021 · Nov 2022 | -77% |
| Oct 2025 · Jun 2026 | -53% |
Cycle drawdowns on daily closes. Sources: ChartRow, Glassnode, 21Shares, UTXO Management calculations.
Volatility and drawdowns have been lower in the most recent cycle.
Standard deviation does not care which direction a return goes. A +65% month and a -37% month count as the same amount of risk. Skew fills in that missing information: it measures whether a return series leans toward large gains (positive skew) or large losses (negative skew). Over the 120 months since October 2016, bitcoin's monthly returns have a skew of +0.48, compared with -0.43 for the S&P 500 and -0.39 for the 60/40 portfolio.
Splitting volatility into its two halves makes the point concrete. Upside deviation measures the size of the positive months only, and downside deviation the size of the negative months only. Bitcoin's upside deviation is 67% and its downside deviation is 34%, a ratio of about 2 to 1; for the S&P 500 the ratio is 1.4 to 1. In other words, bitcoin's gains have been larger than its losses, a difference a volatility screen does not register. Bitcoin has had 28 months above +20% and 7 below -20%. The 60/40 portfolio has never had a month outside ±10%.
The gains also arrive in bursts. Remove the best 12 months and bitcoin's ten-year annualized return falls from 64% to about 4%. Those months were not identifiable in advance, which favors a fixed allocation rule over attempts to time entries.
| Metric | Bitcoin | S&P 500 | 60/40 |
|---|---|---|---|
| Skew | +0.48 | -0.43 | -0.39 |
| Upside deviation | 67% | 13% | 8% |
| Downside deviation | 34% | 9% | 6% |
| Average up month | +20.7% | +3.5% | +2.3% |
| Average down month | -12.0% | -4.1% | -2.7% |
| Months beyond ±20% | 28 / 7 | 0 / 0 | 0 / 0 |
| Up/Down capture vs S&P | 2.69× / 0.63× | 1.00× / 1.00× | 0.65× / 0.67× |
Bitcoin's upside deviation has been twice its downside deviation.
Bitcoin's correlation to stocks depends on the macro environment. When the shock is specific to bitcoin, as in the leverage unwind of late 2025, the two move independently or in opposite directions. When the shock is macro, as in March 2026, when a hawkish Federal Reserve repriced every risk asset at once, bitcoin trades as a high-beta risk asset and falls with equities. The 2026 drawdown contained both regimes within a few months.
Over the full decade the correlation of monthly returns is 0.34 with the S&P 500 and 0.14 with U.S. bonds. On a rolling two-year basis both have moved between slightly negative and moderately positive (Exhibit G).
Three regimes recur. In liquidity-driven rallies and tightening cycles (2020, 2022, early 2026), bitcoin behaves like a leveraged expression of the same macro bet as equities, and correlation rises. In bitcoin-specific episodes, such as the exchange failures of 2022 and the late-2025 unwind, correlation falls toward zero. In calm markets it drifts between the two. Bonds are simpler: bitcoin has shown no reliable relationship to fixed income in either direction.
The practical conclusion is that bitcoin's low long-run correlation diversifies return drivers; it does not hedge equity drawdowns. On the 25 days the S&P 500 fell more than 1% over the past year, bitcoin fell on 22 of them. Its diversification shows up across quarters and years.
Bitcoin is most often compared with gold and with technology stocks, reflecting its dual narrative as a store of value and an emerging technology. It shares gold's scarcity, fungibility, and durability, and differs in divisibility, verification, transfer, and storage.
The nearly three years since the ETF launch have tested the "digital gold" framing directly. Gold returned approximately +64% in 2025, its strongest year in over four decades, while bitcoin fell 6.5%. 2025 was the third calendar year in the study period, after 2018 and 2022, in which gold outperformed bitcoin. Through September 2026 the two are nearly level year to date (bitcoin -4.6%, gold -3.9%).
The correlation of their monthly returns over the past decade is 0.10, which argues for treating them as complements rather than substitutes.
Bitcoin is also perceived as an emerging technology. Does that make it a leveraged technology bet? Against the Nasdaq 100, using the QQQ ETF as a proxy, bitcoin has delivered roughly 3 times the annualized return, 3.8 times the standard deviation, and 2.3 times the maximum drawdown over ten years, with a similar Sharpe Ratio (0.99 against 0.97).
The resemblance ends there. Bitcoin's correlation to the Nasdaq 100 is approximately 0.32, so about 10% of its variance is explained by technology equities and 90% by other factors. Bitcoin fell in 2025, when the Nasdaq 100 gained 20.8%, and in 2018, when QQQ was flat.
Adjusting for volatility makes the point. A hypothetical, backtested portfolio of 20% bitcoin and 80% 3-month T-bills, rebalanced quarterly, carried volatility below the Nasdaq 100 (16.4% against 19.0%) while returning 21.0% a year, with a Sharpe Ratio of 1.10, a maximum drawdown of -18.7%, and a correlation to equities near 0.3 (Exhibit I). Historically, bitcoin has behaved as a distinct return stream, not as leveraged technology exposure.
| 10 years | BTC | QQQ | 20/80 |
|---|---|---|---|
| Ann. return | +63.6% | +20.9% | +21.0% |
| Std dev | 72.2% | 19.0% | 16.4% |
| Sharpe | 0.99 | 0.97 | 1.10 |
| Sortino | 2.07 | 1.64 | 2.62 |
| Max drawdown | -74.6% | -32.6% | -18.7% |
| Worst 12 months | -72.3% | -32.6% | -16.9% |
| Correlation to QQQ | 0.32 | 1.00 | 0.30 |
Exhibit I. October 2016 to September 2026, monthly returns. The 20/80 portfolio is hypothetical, constructed with hindsight, and does not reflect actual trading or any account managed by UTXO Management. Results are gross of management fees, transaction costs, custody fees, and taxes, which would reduce returns. 20/80 is 20% bitcoin, 80% 3-month U.S. T-bills, rebalanced quarterly. QQQ total returns with distributions reinvested; correlation is of monthly returns over the full period. Source: ChartRow, Invesco, UTXO Management calculations.
Bitcoin's qualities and performance record distinguish it from traditional assets. It is neither a real asset nor simply a leveraged technology play, but something in between. Since it does not fit neatly into conventional categories, where should it reside in a portfolio?
Bitcoin is not a single stock or security. It is a digital protocol whose native currency is interdependent with the network that settles it. One framework some investors use positions bitcoin as an investable base layer protocol, analogous to owning a stake in TCP/IP during the adoption of the internet.
It is also a liquid alternative in the literal sense. Venture capital, private equity, and most hedge funds offer diversification with lockups, gates, capital calls, and lagged valuations. Spot bitcoin trades continuously, U.S. spot ETFs trade and redeem daily, and the position is marked to market every day, so it can be sized, rebalanced, or exited on any business day.
There is no simple classification, which leaves the choice of sleeve to each allocator. Under traditional bucketing, bitcoin has been placed in venture capital, technology, commodity, absolute return, or opportunistic sleeves. It has:
Wherever it resides, the bucket matters less than the size and the rebalancing rule.
How do portfolio return and risk change when small bitcoin allocations are added to a traditional 60/40 portfolio? We backtest 1%, 2.5%, and 5% allocations over the 10 years ended September 2026, each funded pro rata from equities and bonds (a 2.5% allocation reduces equities by 1.5% and bonds by 1.0%). Each is rebalanced to target quarterly; the 2.5% and 5% allocations are also shown rebalanced annually and, for contrast, never (Exhibit J).
| Portfolio | Ann. return | Std dev | Sharpe | Sortino | Worst 12M | Max DD | $1 grew to |
|---|---|---|---|---|---|---|---|
| 60/40, quarterly | +9.7% | 10.2% | 0.72 | 1.11 | -16.0% | -20.1% | $2.54 |
| +1% BTC, quarterly | +10.7% | 10.4% | 0.79 | 1.24 | -16.5% | -20.4% | $2.76 |
| +2.5% BTC, quarterly | +12.0% | 10.8% | 0.88 | 1.42 | -17.3% | -20.9% | $3.12 |
| +5% BTC, quarterly | +14.3% | 11.7% | 0.99 | 1.68 | -18.6% | -21.8% | $3.80 |
| +2.5% BTC, annually | +13.8% | 11.5% | 0.97 | 1.65 | -17.3% | -21.0% | $3.63 |
| +5% BTC, annually | +17.3% | 13.5% | 1.07 | 2.06 | -18.8% | -22.0% | $4.93 |
| +2.5% BTC, never rebalanced | +20.2% | 28.2% | 0.71 | 1.22 | -44.9% | -48.5% | $6.28 |
| +5% BTC, never rebalanced | +25.4% | 37.0% | 0.73 | 1.30 | -53.4% | -57.8% | $9.63 |
Exhibit J. Hypothetical portfolios constructed with hindsight; they do not reflect actual trading or any account managed by UTXO Management. Results are gross of management fees, transaction costs, custody fees, and taxes, which would reduce returns. Results depend heavily on the period chosen: a different start or end date would produce materially different outcomes. Monthly total returns, October 2016 through September 2026. 60/40 is 60% S&P 500 (SPY), 40% Bloomberg U.S. Aggregate (AGG). Standard deviation and Sortino annualized from monthly observations; risk-free rate is the 3-month T-bill. Source: ChartRow, UTXO Management calculations.
In the backtest, every rebalanced allocation improved risk-adjusted return, and annual rebalancing produced the highest Sharpe Ratios: 0.97 at 2.5% and 1.07 at 5%, against 0.72 for 60/40 and 0.88 and 0.99 rebalanced quarterly. At 2.5%, annual rebalancing returned 13.8% a year against 9.7% for 60/40, with volatility of 11.5% against 10.2%.
Annual resets let gains run through bitcoin's trending years, an edge specific to this sample. Never rebalancing earned the most but stopped being a balanced portfolio: a 2.5% position bought in October 2016 grew to 66% of the portfolio at its peak, and the portfolio lost 45% in its worst twelve months (Exhibits K and L).
Several indicators are commonly used to assess potential entry points into bitcoin. One long-standing reference level is the 200-week moving average, roughly four years of prices and a proxy for participants' average cost across a cycle. Bitcoin has closed a month at or below it only 14 times in 150 months (2015, 2022-23, June-July 2026), each near a cycle low. At $83,700 on September 30, 2026, spot stood about 28% above it.
Across the 12 month-ends at or below the average with a full year of data, the median 12-, 24-, and 36-month returns were +123%, +320%, and +431%, respectively (Exhibit N).
On-chain realized price, the average cost basis of every coin at the time it last moved, measures the same idea directly. It stood near $53,600 in September 2026 (Glassnode); spot has spent less than a fifth of its history below it.
One approach is to fund a bitcoin allocation on a fixed schedule and, if policy permits, weight purchases toward periods when spot trades within 10-20% of the 200-week average or below realized price. These signals have a short record and may not repeat. These are backtested results identified with hindsight; they are not a trading record and do not predict future returns.
| Forward return | At/below 200 WMA · median | Positive | Worst | Any month · median | Positive | Worst |
|---|---|---|---|---|---|---|
| 12 months | +123% | 100% | +29% | +65% | 69% | -74% |
| 24 months | +320% | 100% | +194% | +215% | 81% | -52% |
| 36 months | +431% | 100% | +189% | +325% | 100% | +5% |
Exhibits M and N. Month-end closes, Apr 2014 to Sep 2026, log scale. 200-week average approximated by a 46-month average of month-end closes, so intra-month touches (Dec 2018, Mar 2020) do not register. Forward returns from month-ends at or below the average (12 observations; June and July 2026 lack forward data). Exhibit N is a backtest of spot bitcoin price returns identified with hindsight, based on 12 overlapping observations from two market cycles (2015 and 2022-23); it is not a trading record, reflects no strategy or account, excludes fees, transaction costs, custody fees, and taxes, and should not be read as predictive of future returns. Source: returnsview.com, UTXO Management.
We suggest sizing the position by the loss you are prepared to explain, not the return you hope for. Exhibit O applies three assumptions to each target weight: bitcoin volatility of 45%, a correlation of 0.3 to the rest of the portfolio, and a -65% drawdown on daily closes. These are planning inputs, not forecasts.
A 2.5% position adds about 11 bps of volatility and risks about 1.6% of the portfolio.
| Target weight | Added portfolio vol | Tracking error vs 60/40 | Portfolio loss at -65% | Weight after +150% year |
|---|---|---|---|---|
| 1.0% | ≈ +3 bps | ≈ 43 bps | -65 bps | 2.2% |
| 2.5% | ≈ +11 bps | ≈ 108 bps | -163 bps | 5.5% |
| 5.0% | ≈ +33 bps | ≈ 215 bps | -325 bps | 10.7% |
Exhibit O. Forward-looking assumptions: bitcoin volatility σb 45%, 60/40 volatility σp 11%, correlation ρ 0.3. Added volatility is the change in portfolio standard deviation when the position replaces 60/40 pro rata: σ² = (1−w)²σp² + w²σb² + 2w(1−w)ρσbσp. Tracking error = w × √(σb² + σp² − 2ρσbσp). Loss is w × 65% with the rest of the portfolio flat; if the 60/40 also has its worst 12 months (-16%), the loss beyond the 60/40's own is 49, 123, and 245 bps. At 60% bitcoin volatility, a 2.5% position adds 27 bps of volatility and 144 bps of tracking error; at 72%, 40 and 174 bps. Drift assumes the remainder of the portfolio returns +10%. Illustrative only.
"Target 2.5% of total assets in spot bitcoin, funded pro rata from public equity and fixed income. Rebalance when the weight exceeds 3.75% or falls below 1.25%, and review quarterly."
We examined the case for bitcoin in a diversified portfolio across a full market cycle, including a drawdown of more than 50%. The past two years added evidence on both sides: volatility has compressed, the drawdown was shallower than in any prior cycle, and in a hypothetical backtest a small, rebalanced allocation improved risk-adjusted returns over the full period even though bitcoin finished last among major assets in 2025 and remains negative for 2026 to date.
Bitcoin's correlation to equities on bad days is real, and the allocation should be sized to the loss it can incur rather than the return it has produced. Its future adoption and performance remain uncertain, but its value will continue to rest on the protocol's ability to enforce a fixed supply without a central authority. We view that property as distinctive among investable assets, and as the basis for a position held in the right size with the right allocation process.
This paper is provided for educational and illustrative purposes only. It addresses bitcoin as an asset and does not describe, offer, or recommend any fund, account, strategy, or service of UTXO Management or its affiliates. It contains no UTXO Management performance and is not a solicitation. All portfolio results, rebalancing outcomes, and allocation examples are hypothetical, constructed with the benefit of hindsight, and do not reflect actual trading. The illustrative policy and risk-budget figures are examples under stated assumptions, not recommendations or forecasts. Readers should consult their own advisors before making any investment decision.
Market figures on page 1 are as of 9/30/2026. Unless noted, performance statistics use month-end total returns through 9/30/2026 from public market data: ChartRow BTC-USD month-end closes for bitcoin; SPY, AGG, and GLD as proxies for the S&P 500, Bloomberg U.S. Aggregate, and gold, distributions reinvested. The 60/40 portfolio is rebalanced at calendar quarter-ends.
Standard deviation is the dispersion of monthly returns, annualized by √12; it treats a large gain and a large loss as equal risk. Maximum drawdown is the largest peak-to-trough decline in month-end values. The Sharpe Ratio divides a portfolio's return in excess of the risk-free rate by its standard deviation, treating upside and downside volatility alike. The Sortino Ratio divides the same excess return by downside deviation only, the volatility of returns below the risk-free rate, so an asset whose large moves are mostly gains scores better on Sortino than on Sharpe. Both use arithmetic mean monthly excess returns over the 3-month T-bill, annualized; on compound returns, which penalize volatility more, the Sharpe Ratios are 0.85 for bitcoin, 0.83 for the S&P 500, and 0.71 for 60/40. Bitcoin trades continuously, so month-end closes differ by vendor; on CoinMarketCap closes, 2017 returns +1,369% (not +1,330%), the best month +69.6% (not +65.2%), and skew +0.62 (not +0.48). Conclusions are unchanged. In Exhibit J, annual rebalancing occurs at December year-end; never-rebalanced portfolios are buy-and-hold from October 2016. Exhibit O is an illustration under stated assumptions, not a forecast.
Statistics marked ≈ are approximations intended to be directionally accurate. Cycle drawdowns and realized price are from the sources cited in Exhibits D and E. ETF assets and the daily S&P 500 comparison on page 5 are UTXO Management estimates from public market data. Tesla volatility: 2025 monthly total returns (PortfoliosLab). Strategic Bitcoin Reserve: Executive Order 14233, March 6, 2025. Market cap: price × circulating supply.
The comments provided herein are a general market overview and do not constitute investment advice, are not predictive of any future market performance, and do not represent an offer to sell or a solicitation of an offer to buy any security. This information is not intended to provide specific advice, recommendations, or projected returns. The views presented represent good faith views of UTXO Management as of the date of this communication and are subject to change as economic and market conditions dictate. Performance statistics are derived from public market data and are approximations; they have not been audited and may differ from figures produced by commercial databases. Though these views may be informed by information from sources we believe to be accurate, we make no representation as to the accuracy of such sources or the adequacy and completeness of such information. Past performance is not indicative of future results. Digital assets are highly volatile and may result in the loss of the entire investment. Exhibits I through L and N, including the 20/80 portfolio, the 60/40 portfolios with bitcoin, the rebalancing rules, and forward returns from the 200-week average, are hypothetical: they were constructed with hindsight, do not reflect actual trading or any account managed by UTXO Management, and are shown gross of management fees, transaction costs, custody fees, and taxes, which would reduce returns. Additional risks include loss or theft of private keys and custodian failure; changes in law, regulation, or tax treatment; exchange-traded products that trade at a premium or discount to net asset value or fail to track the spot price; and forks or other changes to the Bitcoin protocol. UTXO Management is a Nakamoto Company; UTXO Management, its affiliates, and their personnel may hold bitcoin or bitcoin-related investments and may benefit from its wider adoption, which presents a potential conflict of interest.