The start date decided what a Bitcoin slice did
Added to a balanced Australian portfolio, a 5% Bitcoin slice lifted the return from 7.8% to 15.3% a year from 2016, but only from 7.5% to 10.1% from 2018. It has diversified less over time, and gold did that job better.
The answer depends on the start date. A 5% slice took the portfolio from 7.8% to 15.3% a year from 2016 and from 7.5% to 10.1% from 2018. The earlier window includes Bitcoin's 2017 rise of more than twelve-fold.
It diversified less over time. Its correlation with global shares went from 0.19 in 2016 to 2019 to 0.37 in 2020 to 2026.
It has not been a crash hedge. It fell in 7 of the 13 worst months for global shares. Gold rose in 11 of them.
01 · The start date decides
The same slice, two very different results
Prices are from EODHD, in Australian dollars. The balanced portfolio is 35% VAS (Australian shares), 30% VGS (global shares) and 35% VAF (bonds), rebalanced yearly. Gold is GOLD, an ASX-listed gold fund, so its figures include the fund's fees and tracking.
Add a 5% Bitcoin slice from 2016 and the balanced portfolio's return lifts from 7.8% to 15.3% a year. Start in 2018 and the same slice lifts it from 7.5% to 10.1%. Bitcoin's calendar-year returns in Australian dollars were very uneven: 2017 alone was up more than twelve-fold, and 3 of the ten calendar years from 2016 to 2025 were losses. The 2016 window contains that run and the 2018 window does not.
| Bitcoin slice | From 2016: return / yr | Sharpe | From 2018: return / yr | Volatility | Sharpe | Worst fall |
|---|---|---|---|---|---|---|
| None | 7.8% | 1.01 | 7.5% | 8.5% | 0.90 | −20% |
| 2% | 11.1% | 1.24 | 8.6% | 9.0% | 0.99 | −20% |
| 5% | 15.3% | 1.31 | 10.1% | 10.1% | 1.04 | −20% |
| 10% | 21.3% | 1.28 | 12.5% | 12.8% | 1.04 | −22% |
| 20% | 31.0% | 1.23 | 16.6% | 19.0% | 0.97 | −29% |
Risk rose with the slice. Up to 5% the portfolio's worst fall did not change (−20.1% with none, −20.4% with 5%). At 20% it deepened by about nine points to −29.0%. In the 2018 window the Sharpe ratio, return per unit of volatility, was highest with a 5% or 10% slice (1.04) and lower at 20% (0.97). That describes one period, and it is not a size to aim for.
On its own, Bitcoin made 64.1% a year from 2016 and 23.5% from 2018, with a worst fall from a peak of −82% and −80%.
02 · Does it protect a portfolio?
It has gone along with shares more over time
Bitcoin's monthly returns moved with global shares by 0.19 in 2016 to 2019 and by 0.37 in 2020 to 2026, so it was a weaker diversifier in the second half. Against bonds the link was small (0.15 and 0.07).
The 13 worst months for global shares (the bottom tenth) averaged −5.3%. Bitcoin's average in them was positive (+2.3%), but it fell in 7 of the 13: a coin flip, not a hedge. Gold rose in 11 of the 13.
03 · Bitcoin and gold did different jobs
One was the engine, the other the shock absorber
Take the same balanced portfolio and add a 5% slice of Bitcoin, a 5% slice of gold (an ASX gold fund), or 2.5% of each.
| Portfolio | From 2016: return / yr | Sharpe | From 2018: return / yr | Sharpe | Worst fall |
|---|---|---|---|---|---|
| Balanced portfolio | 7.8% | 1.02 | 7.6% | 0.91 | −20.1% |
| Balanced +5% Bitcoin | 15.3% | 1.29 | 10.2% | 1.03 | −20.4% |
| Balanced +5% gold | 8.1% | 1.11 | 8.0% | 1.02 | −18.7% |
| Balanced +2.5% each | 12.0% | 1.30 | 9.2% | 1.05 | −19.5% |
Bitcoin added far more return (+2.6 points a year from 2018, against +0.4 for gold). Gold steadied the ride: the worst fall was −18.7% with gold and −20.4% with Bitcoin. Half of each gave the highest Sharpe ratio of the four in both windows (1.05 from 2018, 1.30 from 2016).
So as a source of growth Bitcoin did more. As a diversifier, gold did more: its correlation with global shares stayed near −0.18 across both periods, while Bitcoin's drifted up.
What this does not show
- One balanced mix, two windows. The portfolio is 35% Australian shares, 30% global shares and 35% bonds, rebalanced yearly. Other mixes and other start dates give different numbers.
- Three market cycles are not a forecast. Bitcoin fell 60% or more from a peak in both the 2018 and 2022 cycles, and had a handful of very large years. What comes next is not in the data.
- The stress sample is small. Thirteen months is enough to say Bitcoin did not reliably rise when shares fell, not enough to size the relationship. Gold's run since 2016 also includes a long climb in its price, and the gold figures are for an ASX-listed fund, so they include its fees and tracking.
- Tax, custody and exchange risk are outside it. Returns are before tax. Nothing here covers how the coins are held, or what happens if an exchange fails.
General information only. Past results, even on honest data, are not a forecast. Nothing here is financial, tax or investment advice.