Over a year, do robot trades beat just holding a big coin after fees?

"We find that the breakeven transaction costs are substantially higher than those typically found in cryptocurrency markets."

"Nevertheless, a large proportion of technical trading rules do generate breakeven transaction costs greater than 50 basis points indicating that performance of technical trading rules in cryptocurrencies are not wiped out by appropriate transaction costs."

"For instance, 36.06% of oscillator rules generate breakeven transaction costs greater than 50 basis points for CoinDesk while 47.35% of channel breakout rules generate rules with breakeven transaction costs greater than 50 basis points for Ethereum."

"However there is no predictability for Bitcoin in the out-of-sample period, although predictability remains in other cryptocurrency markets."

"We find that a long-short portfolio strategy based on the predictions of the employed LSTM and GRU ensemble models yields an annualized out-of-sample Sharpe ratio after transaction costs of 3.23 and 3.12, respectively."

"In comparison, the buy-and-hold benchmark market portfolio strategy only yields a Sharpe ratio of 1.33."

"We find that a long-short portfolio strategy based on the predictions of the employed LSTM and GRU ensemble models yields an annualized out-of-sample Sharpe ratio after transaction costs of 3.23 and 3.12, respectively. In comparison, the buy-and-hold benchmark market portfolio strategy only yields a Sharpe ratio of 1.33."

"However, this outperformance is conditioned by ex post selection of the “best” models and neglect of transaction costs, and therefore it must be interpreted with caution as an illustration of the potential and limits of deep neural networks in short-term prediction of cryptocurrencies."

"The Halloween-based investment strategy outperforms both the No-Halloween and Buy-and-Hold strategies for most assets, suggesting that, depending on risk tolerance, investors may benefit from a Halloween-focused approach."

"The empirical evidence shows that cryptocurrencies exhibit higher returns during the Halloween period (October 31–April 30) compared to the rest of the year, although this period is associated with increased risk."

"Volatility analysis reveals significant differences between Halloween and non-Halloween periods, with most cryptocurrencies (except BTC) showing outperformance during Halloween."

"By default, this study applies a 0.05% transaction fee per execution, aligning with the standard USDT taker fee rate for regular users on the Binance for USD-MFutures4. However,torigorouslytestthestrategy’sresilienceagainstunmodeledfrictions, the assumptions verification phase includes a 0.10% friction rate. This elevated rate serves as a composite proxy, encapsulating both the baseline 0.05% comm"

"As shown in Table 15, the differences in both Sharpe and Sortino ratios between Agent 2 and the Baseline are statistically significant at the 10% level (p-value ≈ 0.06)."

"i represents the proportionate transaction costs of trading US equities with 50 basis points, bonds and risk-free rates with 17 basis points, and cryptocurrency factors with 50 basis points (DeMiguel et al. 2009; Platanakis and Urquhart 2020)."

"we observe that the aggregated market share of prevalent cryptocurrencies such as Bitcoin and Ethereum dropped from roughly 90% in January 2014 to 63% in June 2021 (see, e.g., CoinM arket Cap)."

"For instance, the volume-based factors offer negligible diversification benefits, indicating that the volume-related factors c"

"The results demonstrate that the innovative combination of CUSUM-filtered data with Triple Barrier labeling outperforms traditional time bars and next-bar prediction, achieving consistently positive trading performance even after accounting for transaction costs."

"Their profitability analysis suggested that strategies based on LSTM model predictions could outperform the traditional buy-and-hold strategy. However, the analysis did not account for transaction costs, which can significantly diminish gains in high-frequency trading."