Author: Michel Athayde , Bitget Wallet BD Ambassador
summary
In traditional financial markets, long-term buying of deep out-of-the-money put options is often seen as a typical "black swan insurance" strategy: continuously losing premiums during normal times, only to recoup losses or even make a large profit in a single instance when extreme risks occur. The question is, does this logic still hold true in the more volatile and frequently crashing crypto market?
The answer provided by this backtesting is not simple. Based on historical data from January 1, 2018 to March 14, 2026, we conducted a systematic test on BTC and ETH: purchasing one deep OTM Put asset monthly and holding it until maturity without market timing. The results show that this is not a "stable profit" strategy, but rather an extremely right-skewed, extremely low-win-rate, and highly path-dependent tail insurance tool. Under the benchmark portfolio, BTC's final total return was 97.62%, while ETH's was -73.07%; in other words, the same logic can yield completely opposite long-term results on different assets.
Research Background
"Buying insurance long-term" has always been one of the most controversial strategies. Supporters believe that what truly determines the fate of long-term funds is not daily fluctuations, but a few systemic crashes; opponents, on the other hand, believe that paying premiums year after year is essentially gambling on uncertain disasters with certain losses.
The crypto market makes this debate even more interesting. On one hand, the volatility of BTC and ETH is far higher than that of traditional assets, and crashes are not uncommon; on the other hand, extreme market conditions occur more frequently, theoretically making them more suitable for tail hedging for survival. Thus, a key question arises: if one doesn't time the market or judge cycles, but simply mechanically buys one deep OTM Put contract every month since 2018—this "dumbest" tail insurance method—can it actually make money in the long run?
Strategy Design
The rules for this test are very simple, yet very strict. Positions are opened on the first calendar day of each month; if no initial observation data is available on that day, they are postponed to the next tradable day. One European Put contract is purchased and held until expiration, with no early profit-taking or stop-loss orders. The target Delta of the benchmark portfolio is 0.01, with an expiration date of 30 days. The initial capital is $2000, and 2% of the account's net value is invested each time. If the budget is insufficient to purchase a full contract, the position size is scaled down according to the budget.
To conduct sensitivity analysis, the backtesting was extended to Delta 0.02 and 0.05, investment ratios of 1%, 2%, and 5%, and maturity days of 14, 30, and 45, for a total of 54 parameter combinations.
Backtesting method
Pricing uses Black-Scholes, assuming a risk-free interest rate of 0, and volatility is proxied by the annualized volatility of the daily returns over the past 30 days. The strike price is not set arbitrarily but is derived by back-calculating the target Delta. In terms of data, the daily samples for both BTC and ETH cover up to March 14, 2026. After this cleansing, BTC has 3055 records and ETH has 3048 records, with no missing dates or closing prices requiring correction. Throughout the parameter grid, all skipped transactions occurred when the expiration date exceeded the backtesting end date; no silent failures were observed.
In other words, this conclusion is not based on a small sample or incomplete data, but rather on a complete historical review of BTC and ETH under the same set of rules.
Backtesting results
First, let's look at the benchmark portfolio, which is the version closest to the proposition of "buying a fixed amount of deep OTM Put every month": Delta 0.01, investment of 2%, and holding for 30 days.
The results were highly divergent. BTC executed 98 transactions, achieving a total return of 97.62%, with an annualized compound return of 8.66%. ETH also executed 98 transactions, but the total return was only -73.07%, with an annualized compound return of -14.78%. In conclusion, this strategy is profitable in the long run on BTC, but not on ETH.
However, what's truly noteworthy isn't the returns themselves, but the cost of those returns. The maximum drawdown for the BTC benchmark portfolio reached a staggering -97.24%, while ETH also hit -93.82%. BTC's win rate was only 2.04%, and ETH's was even lower at 1.02%. The median return per trade for both was -100%, and their worst trades were also -100%. This means that most of the time, this strategy isn't just about "small losses," but about losing all the premiums by maturity.
Figure 1: The benchmark portfolio equity curve shows a clear divergence in long-term results between BTC and ETH. BTC ultimately achieved positive returns, while ETH experienced a long-term decline in net asset value.
Figure 2: The drawdown curves illustrate that even if this type of strategy eventually profits, it may still result in prolonged periods of significant losses. The BTC benchmark portfolio experienced a maximum drawdown of -97.24%, while ETH saw a drawdown of -93.82%.
This is very intuitive to see from the charts. The equity curve does not rise smoothly, but stays at a low level for a long time, only occasionally jumping out when there are extreme profits; the drawdown curve is deep underwater for a long time; the histogram of single-trade returns clearly shows that the left side is densely packed with -100%, while the right side only has a very few extremely large profits. It is not a "win rate-based strategy" in the traditional sense, but a typical "positive skewness strategy".
Figure 3: The distribution of single-transaction profits is highly right-skewed. Premiums drop to zero in the vast majority of months, with a few extremely profitable transactions determining the long-term outcome.
Revenue Source Analysis
The core feature of this strategy is that the profits do not come from "making money in most months", but from "making a lot of money in a very few months".
The percentage of months with positive returns for the BTC benchmark portfolio was only 4.08%, and for ETH, it was only 2.04%. BTC had its longest streak of consecutive losses at 57 months, while ETH had an even longer streak of 87 months. This means that even if investors eventually make a profit, they may first have to endure years of continuous losses and the feeling of strategy failure.
On the other hand, the best single trade in the BTC benchmark portfolio yielded returns exceeding 620 times, with the average return per trade boosted to 5.36% by extreme profits; the return distribution skewness reached 9.90, and the kurtosis approached 98. The top 5 profitable trades contributed 10.54 times to the total net profit of BTC. These figures almost spell out the essence of this type of strategy: it doesn't rely on consistent carry to make money, but rather on a few extreme moments to "recoup" years' worth of premiums in one go.
ETH's structure exhibits similar characteristics, but the problem lies in the fact that ETH's extreme profits are insufficient to cover long-term losses. Under the benchmark portfolio, ETH paid a total premium of $2984.47 and only recovered $1523.06 at maturity, resulting in a Premium Burn Ratio of 48.97%. In contrast, BTC paid a total premium of $18903.47 and recovered $20855.85, with a Premium Burn Ratio of -10.33%. This indicates that with BTC, extreme market conditions ultimately recouped the premium loss, even leaving a surplus.
Figure 4: The contribution of extremely profitable trades shows that BTC's long-term profits are almost entirely driven by a few top trades, which is a typical outcome of the tail insurance strategy.
The contribution of extreme market conditions to strategy returns
One interesting finding is that a major market crash does not necessarily mean that your puts will make money. Under the benchmark portfolio, we focused on examining the performance of positions around the time of the March 2020 pandemic crash, the May 2021 crypto crash, the May 2022 LUNA event, and the November 2022 FTX event.
The results were far from optimistic. BTC only saw a profitable event window in May 2021, with a single profit of 127.09%, a recovery multiple of 2.27 times at maturity, and a contribution of 15.87% to the total net profit. However, in March 2020, May 2022, and November 2022, the corresponding holdings all lost 100% of their premiums. ETH fared even more extremely, incurring a 100% loss in all four event windows.
This illustrates that tail insurance doesn't automatically generate profits whenever a black swan event occurs; rather, it's highly dependent on the timing of the entry, the remaining maturity, and the strike price. While fixed monthly rolling purchases can guarantee your "long-term presence," they cannot guarantee that you'll "just happen to hold the most profitable contract."
Performance differences between BTC and ETH
Based on the average results of the full-parameter grid, BTC significantly outperforms ETH. The average total return for BTC across all parameters is 1326.06%, with an average CAGR of 23.87% and an average Tail Insurance Score of 0.544; for ETH, these figures are 519.48%, 13.39%, and 0.362, respectively. In the benchmark portfolio, BTC is positive while ETH is negative, a difference that is also very evident in the net asset value comparison charts for different Deltas and durations.
More importantly, BTC's extreme profitability is better able to cover long-term premium losses, while ETH is more likely to remain in a state of "paying premiums for a long time, but not necessarily getting compensation at critical moments." Simply put, both are tail insurance, but BTC is more like "insurance that can be cashed out," while ETH is more like "insurance with higher costs and more difficult to cash out."
Figure 5: A comparison of benchmark strategies for BTC and ETH shows that while both assets exhibit extreme volatility, BTC offers better tail insurance efficiency.
Parameter sensitivity: Are deeper imaginary values always better?
The parameter comparison results provide a more complete picture. Sorted by total return, the highest-yielding combination appeared on BTC: Delta 0.01, 2% investment, 45-day maturity, with a total return of 9747.06%, corresponding to a CAGR of 74.99%. If only "profitability" is considered, combinations with deeper out-of-the-money values, longer durations, and moderate leverage performed extremely well on BTC.
However, if we look at risk constraints, the answer isn't so simple. The combination with the lowest maximum drawdown is the ETH Delta 0.05, 1% investment, 14-day version, with a maximum drawdown of -37.89%. Looking at the overall average metrics, Delta 0.02 provides a better balance between Total Return, Calmar Ratio, and Premium Burn Ratio, making it the most "reasonable" Delta in this full-parameter grid analysis.
Figure 6: The net asset value curves for different Deltas show that the deeper the Delta, the more the strategy resembles a typical black swan insurance policy; after the Delta is slightly raised, a more balanced combination of returns and capital experience will emerge.
Figure 7: A comparison of different investment ratios shows that position ratios can significantly amplify return elasticity, but also simultaneously amplify long-term drawdowns and capital path risks.
Advantages and disadvantages of the strategy
The advantages of this type of strategy are very clear. First, it has a simple structure, fixed rules, and does not rely on subjective judgment. Second, it has extremely strong convexity; once it hits an extreme market condition, the profit potential is astonishing. Third, it is completely different from mainstream long-only strategies, and is closer to a risk coverage tool in asset allocation than directional trading.
However, the drawbacks are equally severe. First, the long-term trading experience is extremely poor, with low win rates, prolonged drawdowns, and potential years of consecutive losses. Second, it is highly sensitive to "holding path"; even if the market crashes, you may not be holding the most profitable put option. Third, while this backtesting is rigorous enough, it is still based on Black-Scholes and historical volatility proxying. Implied volatility skew, liquidity, fees, and slippage in the real market will further affect the live trading results.
What type of investors is this strategy suitable for?
It is more suitable for three types of people. The first type is institutional funds that use it as a portfolio insurance rather than a primary return strategy; the second type is highly disciplined investors who can withstand years of seemingly continuous losses; and the third type is people who clearly hope to obtain convex returns in extreme market conditions rather than pursuing a high win rate on a daily basis.
This strategy is unsuitable for investors looking for a "stable money-making model," or for funds with short-term performance targets and an intolerance for long-term underperformance. Backtesting results show that even the ultimately profitable BTC benchmark portfolio experienced a maximum drawdown of -97.24% and 57 consecutive months of losses. Without extremely high capital tolerance, this strategy is almost impossible to hold.
Final conclusion
If you consistently buy one deep OTM Put every month starting in 2018, could you make money in the long run? This backtesting results in the answer: yes, but not in a universally true sense. On BTC, the benchmark portfolio ultimately achieved a total return of 97.62%; on ETH, the same logic resulted in a loss of 73.07%. This means it's not a consistently effective money-making machine across assets and environments.
Is it a strategy that relies purely on extreme market conditions? Based on backtesting results, it basically is. The extremely low win rate, the extremely low percentage of months with positive returns, the extremely high skewness and kurtosis, and the overwhelming contribution of the top 5 profitable trades to the final net profit all indicate that it is not essentially a strategy of "working hard for a long time and receiving a monthly salary," but rather a strategy of "paying premiums for a long time and waiting for a few lucky moments."
If we must draw a more practical conclusion from these results, it is this: BTC is more suitable for this kind of tail insurance than ETH; Delta 0.02 is a more balanced choice because it offers better overall cost-effectiveness in a full parameter comparison; and if investors are looking for more extreme convexity, Delta 0.01 remains the most aggressive version and the most "black swan insurance" version.
Ultimately, buying deep OTM Put options long-term is neither the most expensive nor the best strategy. It's more like an expensive but irreplaceable asset allocation tool at certain times. The question is never "whether it's worth buying," but rather whether you actually need this insurance and whether you can withstand the long period before the actual claim arrives.
