MarsBit published an analysis arguing that if the standard is defined narrowly as buying at any point, holding for a full four years, and ending with a positive nominal return, Bitcoin is the only freely traded major high-risk asset that has achieved a 100% success rate across all rolling four-year windows.
The article adds that some low-risk fixed-income instruments can also deliver positive nominal returns over every four-year period, including four-year government bonds, rolling short-term debt, and time deposits. But it says that kind of 100% record is fundamentally different from Bitcoin’s.
The S&P 500 has not delivered gains in every four-year window
Citing long-term data maintained by New York University professor Aswath Damodaran, the article says S&P 500 total return data, including dividends, runs from 1928 to 2025 and includes several four-year periods that ended in losses.
One example is 1929 through 1932, when annual returns were about -8.3%, -25.1%, -43.8%, and -8.6%. Compounded over four years, the total return was about -64.8%. In the article’s example, a $1 million investment would have fallen to roughly $350,000 after four years.
Another example is 1999 through 2002. Even including the final surge of the dot-com bubble in 1999, annual returns of +20.9%, -9.0%, -11.9%, and -22.0% still produced a four-year total return of about -24.4%.
The 2007 to 2010 period also stayed negative despite a sharp rebound after the financial crisis. Annual returns were +5.5%, -36.6%, +25.9%, and +14.8%, yet the four-year total still came to about -3.2%.
Based on those cases, the article says the S&P 500 has a high long-term win rate, but not a perfect one over four-year holding periods.
The Nasdaq 100 also posted a full negative four-year cycle
Using official Nasdaq data, the article points to the dot-com unwind. Nasdaq 100 returns for 2000 through 2003 were -36.4%, -30.8%, -38.9%, and +48.5%. Even with a near-50% rally in the fourth year, the total return across the period was still about -60%.
The article also notes that Nasdaq has said the index fell about 83% from the 2000 bubble peak. It argues that this is where Bitcoin and technology stocks diverge historically: tech equities can spend a full four-year cycle without recovering prior losses, while Bitcoin has not done that so far.
Gold and real estate also failed the four-year test
The article says gold, often treated as a classic long-term store of value, can still produce steep losses over four years. Using Damodaran’s data, it gives 1981 through 1984 as an example. Annual returns were -32.6%, +15.6%, -16.8%, and -19.4%, leaving a cumulative four-year return of about -47.7%.
It also cites research from the World Gold Council showing that gold went through a nearly 12-year bear market from November 1987 to August 1999, with a cumulative price decline of about 48%.
On real estate, the article says the proper comparison is broad U.S. housing rather than a single property bought at a favorable time. Damodaran’s data shows annual real estate returns of about -5.4%, -12.0%, -3.85%, and -4.12% from 2007 through 2010, for a cumulative loss of about -23.3%.
It argues that housing often looks smoother than BTC or equities only because homes do not trade on a public market every day. If daily quotes existed for each property, the swings would look much larger than people usually perceive.
Long-duration Treasuries and corporate bonds did not qualify either
The article says U.S. Treasuries are often described as risk-free, but that claim needs a distinction between holding a bond to maturity and trading a long-duration bond instrument.
It gives 10-year U.S. Treasury total returns from 2021 through 2024 as an example: -4.42%, -17.83%, +3.88%, and -1.64%, for a cumulative four-year return of about -19.8%. The piece attributes that result to the rate-hiking cycle that began in 2022, which pushed down the market price of older bonds with lower coupons.
On that basis, the article says TLT, 10-year Treasury indexes, and long-bond funds are not assets that guarantee gains over four years.
It makes the same point about corporate credit. Baa corporate bond returns over 2021 to 2024 were listed as +1.02%, -15.23%, +8.74%, and +1.74%, leaving a cumulative return of about -5.3% over the four-year period.
Which assets can actually produce a 100% nominal four-year win rate?
The article places a separate group of assets in that category, all tied to low-risk contractual repayment.
Rolling 3-month U.S. T-Bills
A 3-month Treasury bill is one example. The article says returns do not depend on market price appreciation. Instead, investors buy at a discount and receive face value at maturity. In its example, an investor pays $99.5 for a bill with a $100 face value and receives $100 from the U.S. Treasury at maturity, with the difference representing interest.
It cites TreasuryDirect’s description of the mechanism, which says the purchase price is usually below face value and the investor receives the full face value at maturity. It also says Damodaran’s annual T-Bill data since 1928 has been nominally positive in most cases, so continuously rolling 3-month T-Bills naturally produced positive returns over any full four-year historical window.
The article adds one caveat: positive nominal dollars do not guarantee preserved purchasing power. It gives an example in which inflation is 8% while the T-Bill yield is 2%, leaving account balances higher but real purchasing power lower.
Four-year deposits or CDs
The article says that if an investor can lock in a four-year certificate of deposit paying a fixed 4% rate, does not redeem early, stays within insurance limits, and the rate contract remains valid, the nominal dollar profit after four years is largely known on the purchase date.
It notes that the Federal Deposit Insurance Corporation insures eligible bank deposits and CDs up to the standard limit of $250,000 per depositor, per bank, per ownership category. In the article’s framing, that is why the 100% four-year win rate comes from contract terms and credit protection rather than price gains.
U.S. Treasuries maturing within four years
The article makes the same argument for Treasuries that mature within four years. If an investor buys a U.S. Treasury that matures in four years with a 4% yield to maturity and then holds it to maturity, interim price swings do not change the contractual repayment of principal and interest, assuming the U.S. government pays as scheduled.
It contrasts that with buying a 10-year Treasury and being forced to sell it in year four. The article stresses that a bond maturing in four years is not the same thing as holding a long-duration bond for four years. The first can lock in nominal returns; the second carries duration risk.
Strategy data shows a 100% win rate for rolling four-year BTC holdings
The article says a Strategy report used BTC/USD daily price data from July 2010 to August 2026 and measured rolling holding periods starting from each day in the sample.
The results were presented as follows:
| Holding period | Share ending in profit | Worst total return |
|---|---|---|
| 1 year | 73.1% | -83.6% |
| 2 years | 84.0% | -68.3% |
| 3 years | 99.3% | -34.7% |
| 4 years | 100.0% | +32.6% |
According to the article, none of the 4,419 rolling four-year BTC windows in that dataset ended with a loss.
It says the weakest four-year stretch in the sample ran from April 16, 2021 to April 16, 2025 and still returned 32.6%, which works out to roughly 7.3% CAGR.
Bitcoin’s 100% is not the same as a Treasury’s 100%
The article argues that the real comparison is not simply who has a perfect four-year nominal record, but how that record is achieved. Assets that can do it tend to sit at two extremes: very low-risk instruments backed by contractual repayment on one side, and Bitcoin, described in the piece as a high-growth asset still in the early phase of monetization, on the other. Between those extremes, traditional risk assets such as stocks, gold, real estate, and long-duration bonds do not qualify.
It then draws a sharper distinction. A T-Bill’s 100% record comes from knowing, at the time of purchase, roughly how many dollars will be paid back in the future, with the main exposure tied to U.S. sovereign credit.
Bitcoin works differently. The article says Bitcoin has no issuer, no promise of principal repayment, no coupon, and no cash flow. Its four-year 100% record comes entirely from historical market prices ending higher over every completed four-year holding window in the dataset.
The article’s investment takeaway
The piece concludes that Bitcoin may be one of the most prominent modern risk assets with a historical record of positive nominal returns across every four-year holding period.
In its view, what makes the record unusual is not only the 100% figure. Bitcoin has gone through repeated drawdowns of 70% to 90% and still has not produced a single completed four-year negative return cycle.
The article ends by arguing that the ability to hold matters more than frequent trading, and that strategies such as selling calls, using leverage, swing trading, or yield enhancement all change this unusually rare long-term return distribution in one way or another.