Preface

This biweekly quantitative report (May 13–26, 2025) focuses on the market performance of Bitcoin and Ethereum, providing a data-driven analysis of key indicators such as long-short ratios, open interest, and funding rates. The quantitative strategy section highlights the practical application of the “Bollinger Band Reversal Strategy” across the top 10 cryptocurrencies by market capitalization (excluding stablecoins), detailing its core logic, signal criteria, and execution framework. Through systematic parameter optimization and historical backtesting, the strategy demonstrates strong performance in trend identification and risk control, with clear execution discipline. Compared to simply holding BTC or ETH, the strategy shows superior results in both returns and drawdown management, offering a practical framework for quantitative crypto trading.

Summary

• In the past two weeks, BTC broke above 110,000 USDT to reach a new all-time high and continues its upward trend, while ETH has entered a consolidation phase.
• Open interest in BTC and ETH futures has increased, indicating continued capital inflow; however, the Long/Short Taker Size Ratio (LSR) has not strengthened accordingly, suggesting market sentiment remains cautious.
• ETH’s funding rate has shown slightly more volatility than BTC’s, signaling short-term bearish pressure from localized short positions.
• Liquidations are predominantly on the long side, reflecting strong buying interest during BTC’s high-level consolidation.
• A quantitative analysis using a “Bollinger Band Reversal Strategy” shows that with optimized parameters, ETH achieved an annualized return of over 66%, outperforming a simple buy-and-hold strategy.

Market Overview

To systematically present the evolving capital behavior and trading structure of the cryptocurrency market, this report analyzes five key dimensions: the price volatility of Bitcoin and Ethereum, long/short taker size ratio (LSR), open interest in derivatives, funding rates, and liquidation data. These indicators collectively reflect price trends, market sentiment, and risk conditions, offering a comprehensive view of trading intensity and structural characteristics. The following sections will analyze the recent changes in each indicator since May 13th.

1. Price Volatility Analysis of Bitcoin and Ethereum

According to CoinGecko data, since May 13, BTC has maintained an upward trend while ETH has remained range-bound. ETH was initially more active, climbing to 2,600 USDT early in the period, but subsequently entered a consolidation phase. In contrast, BTC steadily strengthened, with capital gradually flowing back in. On May 22, BTC officially broke above the key 110,000 USDT level, reaching a new all-time high of 111,814 USDT after three months, reflecting rising market attention and increased allocation interest.

Driven by multiple favorable factors, BTC’s upward momentum has been strong, with a year-to-date gain exceeding 20%. Its total market capitalization has climbed to $2.19 trillion, surpassing Amazon ($2.13 trillion) and Google ($2.05 trillion), making it the fifth-largest asset globally, behind only Apple. This rally has been fueled by two key drivers: first, the easing of U.S.-China trade tensions, which reduced market risk aversion; and second, Moody’s downgrade of the U.S. sovereign credit rating, prompting investors to seek alternatives to the U.S. dollar, thereby enhancing BTC’s appeal as a hedge.[1][2][3][4]

Figure 1: On May 22, BTC officially broke through the key 110,000 USDT level, reaching a new all-time high of 111,814 USDT after three months.

In terms of volatility, since May 13, ETH has experienced multiple sharp spikes, with overall volatility consistently higher than BTC. This indicates stronger short-term market sentiment swings during ETH price movements, particularly with a significant peak observed in mid-May. In contrast, BTC’s volatility remained relatively stable, showing no major fluctuations and reflecting greater resilience and structural support in its price action.

As shown in the chart, ETH’s volatility surged above 4% in mid-May, significantly higher than BTC during the same period, suggesting more intense short-term speculation during ETH’s breakout phases. Overall, BTC exhibited a more evenly distributed volatility profile, while ETH’s spikes were concentrated at key moments, highlighting its greater sensitivity to external bullish catalysts.

Figure 2: BTC’s volatility remained relatively stable, while ETH saw multiple sharp surges.

Overall, BTC has shown strong performance in this cycle, not only setting a new all-time high but also maintaining a steady upward trend with relatively balanced volatility. This reflects its role as a hedge and store of value in the current market environment. In contrast, although ETH briefly rose to 2,600 USDT in the early phase, its overall gains were limited, and its subsequent performance weakened. Its volatility spikes were concentrated, indicating a reactive but unsustained response to positive external news, with relatively weaker capital inflows.

The divergence in their trends highlights a shift in market asset allocation preferences. In the short term, BTC’s price momentum and volatility pattern serve as key indicators of market risk appetite and liquidity direction. Meanwhile, ETH’s outlook still depends on whether fundamental catalysts can further unlock its potential.

2. Analysis of Long/Short Taker Size Ratio (LSR) for Bitcoin and Ethereum

The Long/Short Taker Size Ratio (LSR) is a key indicator that measures the volume of aggressive buying versus aggressive selling, often used to gauge market sentiment and trend strength. An LSR greater than 1 indicates that the volume of market buys (aggressive longs) exceeds that of market sells (aggressive shorts), suggesting a bullish market bias.

According to Coinglass data, since May 13, BTC has steadily climbed and even set a new all-time high during intraday trading on May 22. However, the LSR did not rise in tandem—instead, it fluctuated around 1 and even dropped below it several times. This indicates that despite BTC’s price rally, significant short positioning remained in the market. Notably, after May 18, while BTC continued to climb, the LSR dipped below 0.95, reflecting increased hedging or short activity at higher price levels and a cautious market sentiment.

ETH, on the other hand, showed more volatile LSR movements. Although its price once rose to 2,600 USDT, the LSR never consistently stayed above 1. Instead, it experienced sharp swings, indicating a lack of sustained bullish dominance and a market still driven by short-term speculation and strong bearish pressure.

Overall, despite the upward trend in BTC over the past two weeks, the LSR failed to strengthen, suggesting skepticism about the rally’s sustainability. Investors remained cautious, adopting hedging strategies during the uptrend. This implies that the current rally is largely driven by short-term sentiment and news catalysts rather than structural bullish support. In the near term, a sustained recovery in the LSR will be crucial to confirming whether the market can maintain its upward momentum.[5]

Figure 3: BTC price has been steadily rising, but the LSR has not followed suit, fluctuating around 1 and even falling below it multiple times.

Figure 4: ETH intraday price once surged to 2,600 USDT, but the LSR showed significant volatility, indicating clear long-short divergence and a lack of sustained bullish dominance.

3. Open Interest Analysis

According to Coinglass data, since May 13, both BTC and ETH have seen a continuous increase in open interest, indicating rising market activity and growing leverage participation. BTC’s open interest steadily climbed from around $70 billion and surpassed $80 billion around May 20. Although it slightly pulled back afterward, it remained at relatively high levels, reflecting sustained capital inflow during the price uptrend. In contrast, ETH’s open interest rose more moderately, from approximately $27 billion to around $32 billion, with smaller fluctuations, suggesting a more measured pace of positioning in the ETH derivatives market.

Overall, despite the simultaneous rise in BTC and ETH open interest and heightened trading activity, the LSR did not strengthen accordingly. This indicates that while capital is entering the market, sentiment remains cautious. The ongoing price increase has been accompanied by significant divergence and hedging, suggesting that a structurally bullish trend has yet to be firmly established.[6]

Figure 5: BTC open interest shows stronger upward momentum, while ETH open interest remains relatively stable.

4. Funding Rate

According to Coinglass data, BTC and ETH funding rates have remained slightly volatile around the 0% mark, frequently flipping between positive and negative. This indicates that long and short forces in the market remain relatively balanced, with no clear tilt toward a single direction. Although both price and open interest increased during the period, funding rates stayed within the narrow range of 0% to 0.01%, without significant upward spikes.

This trend suggests that while there is some bullish expectation in the market, leveraged long positions remain cautious, with no signs of aggressive buying pressure. ETH’s funding rate showed slightly greater volatility than BTC, with occasional signs of localized short pressure, reflecting cautious sentiment near key price levels. Overall, the funding rate movement confirms that the current market remains in a moderately optimistic phase, not yet entering an overheated state.[7][8]

Figure 6: BTC and ETH funding rates fluctuate slightly around 0%, reflecting a balanced long-short dynamic in the market.

5. Cryptocurrency Liquidation Chart

According to Coinglass data, the total liquidation of long positions in the crypto market over the past two weeks reached $2.697 billion, significantly higher than the $1.76 billion in short liquidations. This indicates that during the recent period of high-level consolidation, bullish sentiment has been strong, with many traders chasing the rally. However, this also made them more vulnerable to forced liquidations during pullbacks.

Compared to earlier phases dominated by short liquidations, the recent shift toward increased long liquidations suggests that the market has entered a consolidation phase at higher price levels, with growing risks for those chasing upward momentum. Capital has become more sensitive to short-term volatility, and the derivatives market remains characterized by high leverage and elevated risk.

Overall, although BTC and ETH open interest has continued to rise since May 13 — indicating increased trading activity and leverage — the LSR has not strengthened accordingly. This implies that while capital is entering the market, overall sentiment remains cautious. Liquidation data further shows that even during price rallies, long positions frequently face liquidations near local highs. Notably, following BTC’s new high on May 23, increased volatility led to significant pressure on long positions the next day. This underscores that despite capital inflows, the market remains in a state of high volatility, high leverage, and hedging activity — and a structurally sustained bull trend has yet to be confirmed.[9]

Figure 7: Total long liquidations reached $2.697 billion, significantly higher than the $1.76 billion in short liquidations, indicating strong bullish chasing behavior amid high-level consolidation.

Quantitative Analysis – Bollinger Band Reversal Strategy

(Disclaimer: All forecasts in this article are based on historical data and market trends and are for informational purposes only. They should not be considered investment advice or a guarantee of future market performance. Investors should carefully assess risks and make prudent decisions when engaging in related investments.)

1. Strategy Overview

The “Bollinger Band Reversal Strategy” is a mean-reversion approach that combines volatility assessment with trend structure monitoring. Centered on Bollinger Bands, the strategy triggers a long entry when the price falls below the lower band, signaling a potential short-term oversold condition and rebound opportunity. Conversely, an exit is executed when the price crosses above the upper band or reaches predefined take-profit or stop-loss thresholds, indicating a possible end of the current price swing.

To enhance stability and signal reliability, the system also incorporates multiple Simple and Exponential Moving Averages (SMA/EMA) to capture broader trend direction. The use of fixed-percentage take-profit and stop-loss rules further supports effective risk management and profit locking. Overall, the strategy is well-suited for range-bound or high-volatility market conditions, aiming to capture short-term reversals with built-in tolerance and trading discipline.

2. Core Parameter Settings

3. Strategy Logic and Operational Mechanism

Entry Condition

• When there is no existing position, a long entry is triggered when the price breaks below the lower Bollinger Band, indicating an oversold condition.

Exit Conditions:

• Stop-Loss Exit: If the price falls to the entry price × (1 - stop_loss_percent), a forced stop-loss is triggered.
• Take-Profit Exit: If the price rises to the entry price × (1 + take_profit_percent), a take-profit order is executed.
• Upper Band Exit: If the price breaks above the upper Bollinger Band, it is interpreted as a short-term return to normalcy, prompting an exit signal.

Live Trade Example Chart

• Trade Signal Trigger

The chart below shows the ETH/USDT 1-hour candlestick at the most recent trade entry on May 25, 2025. As shown, the price rebounded sharply after breaking below the lower Bollinger Band, accompanied by a spike in trading volume and a MACD bullish crossover. The strategy entered a long position at this point, successfully capturing the start of a short-term rebound—fully aligned with the entry logic of the Bollinger Band strategy.

Figure 8: Example of actual entry signal for ETH/USDT strategy on May 25, 2025

• Trading Action and Outcome
  On the morning of May 26, ETH reached the upper Bollinger Band, triggering an exit based on the strategy’s preset conditions and successfully locking in profits from the short-term rebound. Although the price saw a slight upward movement afterward, this exit adhered to the disciplined exit logic of the Bollinger Band strategy, demonstrating consistent risk control execution. Incorporating a trailing stop or trend-following mechanism in the future could further enhance profit retention during strong market rallies.

Figure 9: Example of ETH/USDT strategy exit point on May 26, 2025

Through the above case study, we clearly demonstrate the strategy’s entry and exit logic as well as its dynamic take-profit and stop-loss mechanism when price approaches the boundaries of the Bollinger Bands. The strategy enters long positions upon a downward breakout below the lower band and exits when price breaks above the upper band, effectively capturing short-term rebound opportunities. While controlling downside risk, it successfully locks in profits within the key price range. This example not only validates the practical applicability and disciplined execution of the Bollinger Band strategy, but also highlights its stability and risk control capabilities in high-volatility market conditions—providing empirical support for further parameter optimization and broader application.

4. Practical Backtesting Example

Backtesting Parameter Settings
To identify the optimal parameter combinations, we conducted a systematic grid search across the following ranges:

• bb_period: 5 to 25 (step size: 1)
• bb_dev: 1 to 5 (step size: 1)
• stop_loss_percent: 1% to 2% (step size: 0.5%)
• take_profit_percent: 10% to 16% (step size: 5%)

Using the top 10 cryptocurrencies by market capitalization (excluding stablecoins) as examples, we backtested 1-hour candlestick data from January to May 2025. A total of 630 parameter combinations were tested, from which the five top-performing strategies were selected based on annualized return.

Evaluation metrics included annualized return, Sharpe ratio, maximum drawdown, and ROMAD (Return Over Maximum Drawdown), providing a comprehensive assessment of each strategy’s performance stability and risk-adjusted returns under varying market conditions.

Figure 10: Performance comparison of the five best-performing strategy configurations.

Strategy Logic Description
When the system detects that the price has fallen below the lower Bollinger Band, it is interpreted as a short-term oversold signal, immediately triggering a buy order. This setup aims to capture rebound opportunities following price deviations from the mean, using the upper Bollinger Band as a dynamic take-profit reference to enhance profit locking. If the price rebounds to the upper band or reaches the predefined stop-loss or take-profit threshold, the system will automatically execute an exit to balance risk and return.

Using ETH as an example, the strategy is configured with the following parameters:

• bb_period = 23 (Bollinger Band calculation period, controlling the responsiveness of the band)
• bb_dev = 2 (Bollinger Band standard deviation multiplier, determining the sensitivity of the upper/lower bands)
• stop_loss_percent = 1%
• take_profit_percent = 15%

This logic combines statistical mean-reversion signals with fixed-percentage risk control mechanisms, making it well-suited for choppy or clearly corrective market conditions.

Performance and Results Analysis
The backtesting period spans from January 1 to May 26, 2025, during which the strategy delivered stable overall performance. The chart illustrates the cumulative return curves of the five best-performing parameter sets on a 1-hour timeframe. Most combinations significantly outperformed BTC and ETH buy-and-hold strategies. In particular, the strategies applied to ETH, ADA, and APT showed steadily rising returns, with peak cumulative gains exceeding 30% and annualized returns above 66%. In contrast, the spot performance of BTC and ETH remained flat or negative, with ETH experiencing a drawdown of nearly -50%.

Overall, these five strategies demonstrated solid capital management and risk control, especially during consolidation phases and early trend reversals. They effectively avoided medium- to long-term drawdowns while capturing short- to mid-term swing opportunities. The current strategy sets achieve a favorable balance between profitability and stability, making them suitable for real-world deployment. Going forward, enhancements such as dynamic Bollinger parameters, volume-based filters, or volatility screens may further improve adaptability across different market conditions and support broader application across multiple assets and timeframes.

Figure 11: Cumulative return comparison of the five best-performing strategies versus BTC and ETH buy-and-hold over the past year.

5. Strategy Summary

The “Bollinger Band Reversal Strategy” enters long positions when the price touches the lower Bollinger Band and exits when the price reaches the upper band, or predefined take-profit or stop-loss thresholds. This setup is designed to capture short-term rebound opportunities and is particularly effective during consolidation phases or at the early stage of trend reversals.

In this backtest, the strategy was applied to the top 10 cryptocurrencies by market capitalization (excluding stablecoins), using 1-hour candlestick data. A total of 630 parameter combinations were tested. The best-performing configuration was found on ETH, with settings of bb_period = 23, bb_dev = 2.0, take_profit_percent = 10%, and stop_loss_percent = 1%. This setup yielded an impressive annualized return of 66%, significantly outperforming ETH’s buy-and-hold return of -32% over the same period.

Notably, the strategy’s win rate was only 35%, with a 65% loss rate. However, the average profit per winning trade far exceeded the average loss, resulting in superior overall returns. This highlights that return and win rate are not always positively correlated—what matters more is a well-designed risk-reward ratio.

Parameter distribution analysis showed that mid-range Bollinger periods, wider bands, high take-profit levels, and low stop-loss thresholds contributed to better performance by capturing larger price swings and reducing premature exits. The strategy features clear entry and exit logic and stable risk management, making it well-suited for deployment across multiple assets and timeframes. Incorporating volume filters, trailing stops, or trend-detection mechanisms could further enhance its robustness and real-world performance.

Conclusion

From May 13 to May 26, 2025, BTC showed strong performance, breaking above 110,000 USDT and reaching a new all-time high, reflecting market recognition of its role as a hedge and store of value. In contrast, ETH entered a consolidation phase after a brief rally, with significantly higher volatility than BTC, indicating its greater sensitivity to short-term sentiment and news. Although the open interest for BTC and ETH continued to rise—signaling increased trading activity—the long/short ratio (LSR) did not strengthen accordingly. This suggests that despite upward price movement, the market remains cautious at high levels, with visible hedging activity. Moreover, long liquidations significantly outpaced shorts, implying that while capital inflows support the rally, there is an elevated risk of long-position chasing in a highly leveraged and volatile environment marked by structural divergences.

The quantitative section focuses on the applicability and stability of a “Bollinger Band Reversal Strategy” during choppy or reversing market conditions, especially for high-volatility assets like ETH. The strategy enters on price dips below the lower Bollinger Band, exits on either an upper band breakout or predefined stop-loss/take-profit triggers, effectively balancing risk and reward. In backtests from January to May 2025, the best-performing parameter set on ETH achieved an annualized return of 66%, far outperforming the buy-and-hold approach. Results showed that mid-range Bollinger periods, wider band settings, and higher take-profit thresholds enhance strategy performance. Overall, the strategy features clear entry and exit logic with strong execution discipline, making it suitable for volatile markets and scalable across multiple assets. However, real-world applications may still be affected by market noise, extreme conditions, or signal failure. It is recommended to combine the strategy with other quantitative factors and robust risk controls to improve stability and adaptability, while maintaining rational judgment and caution.


References:

1. CoinGecko, https://www.coingecko.com/
2. Gate, https://www.gate.com/trade/BTC_USDT
3. Gate, https://www.gate.com/trade/ETH_USDT
4. Coinglass, https://www.coinglass.com/global-assets
5. Coinglass, https://www.coinglass.com/LongShortRatio
6. Coinglass, https://www.coinglass.com/BitcoinOpenInterest?utm_source=chatgpt.com
7. Gate, https://www.gate.com/futures_market_info/BTC_USD/capital_rate_history
8. Gate, https://www.gate.com/futures/introduction/funding-rate-history?from=USDT-M&contract=ETH_USDT
9. Coinglass, https://www.coinglass.com/pro/futures/Liquidations

Gate Research is a comprehensive blockchain and cryptocurrency research platform that provides readers with in-depth content, including technical analysis, trending insights, market reviews, industry research, trend forecasts, and macroeconomic policy analysis.

Disclaimer
Investing in the cryptocurrency market involves high risk. Users are advised to conduct independent research and fully understand the nature of the assets and products before making any investment decisions. Gate is not responsible for any losses or damages arising from such investment decisions.