Notes

A news scan found the headline 'XRP/USD C91202 CLI20day close-breakout diagnostic'. This is a diagnostic report for a trading strategy. The report was run on September 12 at 02:04 UTC. The strategy used was a breakout strategy, which is a type of trading strategy that involves buying a security when its price breaks out of a certain range, looking back 20 days, and considering 730 days of data. The report assumed the same close entry, zero explicit fees or slippage, and no holdout, with the current partial day included. It did not use a static high rule, and there was no trade authorization. The results showed 23 trades, 731 bars of data, a return percentage of 539.0275372544944, a buy and hold percentage of 142.26815503737987, a maximum drawdown percentage of 19.959541078284378, a win rate of 56.52173913043478, and an average trade percentage of 9.865736042897957 for the symbol XRP/USD. https://runway.adityaperswal.workers.dev/data/crypto-bars ; /app/agent/dist/backtest.js

A news scan found the headline 'UNI/USD C91202 CLI20day close-breakout diagnostic'. This is a diagnostic report for a trading strategy. The report is for the symbol UNI/USD, which is a currency pair. It was run on September 12 at 02:04 UTC. The strategy used was a breakout strategy, which means it looks for a stock or currency to break out of a certain price range. The lookback period was 20 days, and the strategy was tested over 730 days. The report includes the same close entry, with zero explicit fees or slippage, and no holdout. It does not use a static high rule, and there is no trade authorization. The results of the strategy are as follows: the symbol UNI/USD had 34 trades, 731 bars, a return percentage of -15.704623902105707, a buy and hold percentage of -11.018164963471433, a maximum drawdown percentage of 35.08588994044547, a win rate of 32.35294117647059, and an average trade percentage of -0.2840292657895688. https://runway.adityaperswal.workers.dev/data/crypto-bars ; /app/agent/dist/backtest.js

A news scan found the headline 'LINK/USD C91202 CLI20day close-breakout diagnostic'. This is a diagnostic report for a trading strategy. The report is for the symbol LINK/USD and was run on September 12 at 02:04 UTC. The strategy used was a breakout strategy with a lookback period of 20 days over 730 days. The report includes the current partial day and does not account for fees or slippage. It also does not include a static high rule or trade authorization. The results show 41 trades over 731 days with a return percentage of -13.272619416715425, a buy and hold percentage of 7.07997405245111, a maximum drawdown percentage of 37.20074970390006, a win rate of 36.58536585365854, and an average trade percentage of -0.18909878689491377. https://runway.adityaperswal.workers.dev/data/crypto-bars ; /app/agent/dist/backtest.js

A news scan found the headline 'HYPE/USD C91202 CLI20day close-breakout diagnostic'. A diagnostic report for the HYPE/USD trading strategy was run on September 12 at 02:04 UTC. The strategy used was a breakout strategy with a 20-day lookback period and 730 days of data. The report included the current partial day and did not account for fees or slippage. It also did not use a static high rule or trade authorization. The results showed 17 trades, 208 bars, a return percentage of -1.6917646751823279, a buy and hold percentage of 168.21476323544613, a maximum drawdown percentage of 19.22060126476856, a win rate of 29.411764705882355, and an average trade percentage of 0.1603048616146881. https://runway.adityaperswal.workers.dev/data/crypto-bars ; /app/agent/dist/backtest.js

A news scan found the headline 'LTC/USD C91202 CLI20day close-breakout diagnostic'. This is a diagnostic report for a trading strategy. The report is for the symbol LTC/USD, which is a currency pair. It was run on September 12 at 02:04 UTC. The strategy used was a breakout strategy, which is a type of trading strategy that involves buying or selling when a stock or currency breaks out of a certain price range. The lookback period was 20 days, and the test was run over 730 days. The test included the current partial day and did not account for fees or slippage, which are costs associated with buying or selling. The test also did not include a static high rule, which is a rule that sets a maximum price, or trade authorization, which is approval to make a trade. The results of the test were as follows: the symbol LTC/USD had 40 trades, 731 bars of data, a return percentage of -19.35%, a buy and hold percentage of -14.44%, a maximum drawdown percentage of 47.22%, a win rate of 27.5%, and an average trade percentage of -0.34%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; /app/agent/dist/backtest.js

A news scan found the headline about a change in the rules for 3350.T C91202 Metaplanet, which is related to a type of investment called a denominator-driven BTC yield. This change took effect on September 11 and includes several key points: the number of shares per right was changed from 696 to 410, and the number of underlying units was reduced from 319,464,000 to 188,190,000, which is a decrease of 41.1%. The remaining units were also reduced from 236,640,000 to 105,366,000, which is a decrease of 55.5%. Additionally, 82,824,000 units that were already issued will not be cancelled. The exercise price of 10 JPY remains unchanged, and unvested exercises will take place in 2029, 2030, and 2031, with a lockup period ending on August 17, 2031. As a result of these changes, the total number of diluted shares decreased from 1,631,543,380 on June 30 to 1,500,108,824 on September 11. The number of BTC, which stands for Bitcoin, a type of digital currency, remains unchanged at 43,000. The reported yield for the quarter, which is 8.8%, reflects the changes in the denominator, and not the 3,768 BTC that were purchased. If all remaining exercise proceeds are realized, they would total 1.05366 billion JPY, but this is not guaranteed to result in cash. https://contents.xj-storage.jp/xcontents/33500/d4cf9d42/889a/45f1/bd2c/dc39fe742609/140120260911535363.pdf

A news scan found the announcement that the issuer of 3350.T, also known as C91202 Metaplanet, has called a virtual Extraordinary General Meeting, or EGM, which is a special meeting of shareholders, to take place on December 18 at 10am. The record date for this meeting is September 30, and creditors can object by December 16. The changes are planned to be effective by December 30. The company's capital stock, which is the amount of money invested in the company, will be reduced from 27,803,501,261 Japanese yen to 1 Japanese yen. The capital reserve, which is a fund set aside for unexpected expenses, will be reduced from 27,803,501,260 Japanese yen to 0 Japanese yen. A total of 55,607,002,520 Japanese yen will be transferred to other capital surplus, which is a type of equity. Additionally, 1,805,472,467 Japanese yen will be used to offset the retained deficit, which is the amount of money the company has lost over time. The issuer states that the number of shares will remain unchanged and that there will be no reduction in assets or distribution of cash. https://contents.xj-storage.jp/xcontents/33500/d2e50092/f9f0/460d/84e0/cac7fc3966d3/140120260911535362.pdf

On September 11, the board approved a wholly owned Hong Kong asset management subsidiary, which is a company that manages investments, for Metaplanet, known by the ticker 3350.T and the code C91202. The subsidiary will initially have a planned capital of USD1 million and is scheduled to be established in September 2026. This subsidiary will help with executing trades and managing risks in Asia, and it will work together with the Miami office that uses a Multi-Account Management system, or MAM for short, which is a way to manage multiple investment accounts. The company that issued the announcement, called the issuer, expects that this new subsidiary will have a minimal impact on the financial results for the fiscal year 2026. The company has not yet received any investment inflows, which are payments into the investment fund, or purchased any Bitcoin, a type of digital currency. https://contents.xj-storage.jp/xcontents/33500/51596f91/9823/4191/a5f9/6ea19d289c1f/140120260911535285.pdf

A news scan found the following data: Completed daily returns from January 1, 2026, to September 11, and pairwise Bitcoin correlation, with annualization of 365 days. Hourly returns from March 1 to September 12 at 01 UTC, with 4634 overlapping complete 24-hour blocks, and 20000 paths for 48 hours by sampling 2 blocks. The exit strategy is based on hourly close, with a hypothetical spot 2 ATR stop and net 3R target, considering 25 basis points fees and 10 basis points lip, which is not a buy recommendation or forecast conditional on entry. All measured figures are as follows: the stock symbol is UNI, with 254 data points, annualized volatility of 82.30%, skew of 0.57, excess kurtosis of 3.18, daily percentage loss at 1% quantile of -8.80%, Bitcoin correlation of 0.67, 90% highest 20-day gain of 56.44%, spot price of 6.05, 14-day average true range of 0.64, stop price of 4.77, target price of 10.06, net risk percentage of 21.70%, 4634 historical blocks, 20000 simulated paths, stop first percentage of 0.01%, target first percentage of 0%, simulated return at 1% quantile of -13.18%, and simulated worst return of -24.74%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A news scan found the following data: Completed daily returns from January 1, 2026, to September 11, 2026, and pairwise Bitcoin correlation, with annualization of 365 days. Hourly returns from March 1 to September 12 at 01 UTC, with 4634 overlapping complete 24-hour blocks, and 20000 paths for 48 hours by sampling 2 blocks. The data includes hourly close exit and hypothetical spot 2 ATR stop and net 3R target, with 25 basis points fees and 10 basis points lip, which is not a buy recommendation or forecast conditional on entry. All measured figures are as follows: the stock symbol is LINK, the number of data points is 254, the annualized volatility percentage is 63.69334786266675, the skew is 0.24669117709870492, the excess kurtosis is 2.9845786704273447, the first percentile daily percentage is -7.4296945714373175, the correlation with Bitcoin is 0.8588838890319977, the 90th percentile high 20-day gain percentage is 37.260791648163924, the spot price is 11.53, the 14-day average true range is 0.7008925922500001, the stop price is 10.1282148155, the target price is 16.059618926048984, the net risk percentage is 12.771015578330303, the number of historical blocks is 4634, the number of simulations is 20000, the percentage of first stop is 0.21, the percentage of first target is 0, the first percentile simulated return percentage is -9.689645036581835, and the worst simulated return percentage is -14.817951548420039. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A news scan found the following data: Completed daily returns from January 1, 2026, to September 11, and pairwise correlation with Bitcoin, with annualization over 365 days. Hourly returns from March 1 to September 12 at 01 UTC, with 4634 overlapping complete 24-hour blocks, and 20000 paths over 48 hours by sampling 2 blocks. The data includes a hypothetical spot exit with a 2 ATR stop and a net 3R target, with 25 basis point fees and 10 basis point slippage, which is not a buy recommendation or forecast conditional on entry. All measured figures are as follows: the symbol is HYPE, the number of data points is 206, the annualized volatility is 86.10%, the skew is 0.38, the excess kurtosis is 1.36, the first percentile daily return is -9.15%, the correlation with Bitcoin is 0.53, the 90th percentile highest 20-day gain is 63.48%, the spot price is 78.77, the 14-day average true range is 5.02, the stop price is 68.72, the target price is 111.13, the net risk is 13.37%, the number of historical blocks is 4634, the number of simulations is 20000, the percentage of first stop is 1.52%, the percentage of first target is 0.015%, the first percentile simulated return is -13.68%, and the worst simulated return is -17.93%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A news scan found the following data: Completed daily returns from January 1, 2026, to September 11, with pairwise Bitcoin correlation and annualization of 365 days. Hourly returns from March 1 to September 12 at 01 UTC, with 4634 overlapping complete 24-hour blocks, and 20000 paths for 48 hours by sampling 2 blocks. The data includes hourly close exit, hypothetical spot 2 ATR stop, and net 3R target, with 25 basis points fees and 10 basis points lip, which is not a buy recommendation or forecast conditional on entry. All measured figures are as follows: the symbol is SOL, the number of data points is 254, the annualized volatility is 63.68%, the skew is -0.16, the excess kurtosis is 2.74, the 1st percentile daily return is -8.06%, the correlation with Bitcoin is 0.88, the 90th percentile 20-day high gain is 32.06%, the spot price is 102.04, the 14-day average true range is 4.97, the stop price is 92.09, the target price is 134.73, the net risk percentage is 10.37%, the number of historical blocks is 4634, the number of simulations is 20000, the percentage of stop first is 1.10%, the percentage of target first is 0%, the 1st percentile simulated return is -10.45%, and the worst simulated return is -13.04%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A news scan found the headline 'XRP/USD C91202 daily moments and 48-hour block-bootstrap risk'. Completed daily returns from January 1, 2026, to September 11, and pairwise BTC correlation, with annualization of 365 days. Hourly returns from March 1 to September 12 at 01 UTC, with 4634 overlapping complete 24-hour blocks, and 20,000 paths for 48 hours by sampling 2 blocks. The exit strategy is based on hourly close, with a hypothetical spot 2 ATR stop and net 3R target, considering 25 basis points fees and 10 basis points lip, which is not a buy recommendation or forecast conditional on entry. All measured figures are for XRP, with the following results: the stock symbol is XRP, the number of data points is 254, the annualized volatility is 65.13%, the skew is 0.56, the excess kurtosis is 11.75, the first percentile daily return is -6.01%, the correlation with BTC is 0.88, the 90th percentile of the highest 20-day gain is 40.18%, the current spot price is 1.3605, the 14-day average true range is 0.0676, the stop price is 1.2254, the target price is 1.8043, the net risk is 10.56%, the number of historical blocks is 4634, the number of simulations is 20,000, the percentage of stop losses in the first position is 0.29%, the percentage of target hits in the first position is 0.04%, the first percentile simulated return is -8.16%, and the worst simulated return is -12.87%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A news scan found the following data: Completed daily returns from January 1, 2026, to September 11, and pairwise correlation with Bitcoin, annualized over 365 days. Hourly returns from March 1 to September 12, 01 UTC, were calculated over 4634 overlapping complete 24-hour blocks, with 20000 paths simulated over 48 hours by sampling 2 blocks. The strategy used an hourly close exit, with a hypothetical spot stop set at 2 ATR and a net target of 3R, considering 25 basis point fees and 10 basis point slippage, which is not a buy recommendation or forecast conditional on entry. The measured figures for LTC are as follows: the symbol is LTC, the number of data points is 254, the annualized volatility is 49.665170102993144 percent, the skew is -0.26118321363956926, the excess kurtosis is 6.165896267931569, the first percentile daily return is -6.745787803728555 percent, the correlation with Bitcoin is 0.805171854243774, the 90th percentile 20-day high is 30.019072323794592 percent, the current spot price is 53.69345, the 14-day ATR is 2.490703183071428, the stop price is 48.71204363385714, the target price is 70.1477141271181, the net risk is 9.910897640282645 percent, the number of historical blocks is 4634, the number of simulations is 20000, the percentage of stops hit first is 0.255, the percentage of targets hit first is 0, the first percentile simulated return is -7.713471659078926 percent, and the worst percentile simulated return is -12.522325850711201 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A news scan found the following data: Completed daily returns from January 1, 2026, to September 11, 2026, and pairwise Bitcoin correlation, with annualization of 365 days. Hourly returns from March 1, 2026, to September 12, 2026, at 01 UTC, with 4634 overlapping complete 24-hour blocks, and 20,000 paths for 48 hours by sampling 2 blocks. The data includes hourly close exit, and a hypothetical spot with a 2 ATR stop and a net 3R target, with 25 basis point fees and 10 basis point slippage, which is not a buy recommendation or forecast conditional on entry. All measured figures are as follows: the symbol is ETH, with 254 data points, an annualized volatility of 63.38%, a skew of 0.18, excess kurtosis of 5.17, a daily percentage loss at the first quartile of -8.59%, a correlation with Bitcoin of 0.91, a 20-day high percentage gain at the 90th percentile of 35.06%, a spot price of 2514.15, an ATR of 95.36, a stop price of 2323.43, a target price of 3157.02, a net risk percentage of 8.23%, 4634 historical blocks, 20,000 simulated paths, a first stop percentage of 3.09%, a first target percentage of 0.04%, a simulated return at the first percentile of -9.06%, and a worst simulated return of -12.32%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A news scan found the following data: Completed daily returns from January 1, 2026, to September 11, and pairwise BTC correlation, with annualization of 365 days. Hourly returns from March 1 to September 12 at 01 UTC, with 4634 overlapping complete 24-hour blocks, and 20000 paths for 48 hours by sampling 2 blocks. The data includes hourly close exit, and a hypothetical spot with a 2 ATR stop and a net 3R target, with 25bp fees and 10bps lip, which is not a buy recommendation or forecast conditional on entry. All measured figures are as follows: the symbol is BTC, the number of data points is 254, the annualized volatility is 46.46%, the skew is -0.31, the excess kurtosis is 7.55, the 1st percentile daily return is -5.82%, the correlation with BTC is 1, the 90th percentile 20-day high gain is 20.51%, the spot price is 77249.81, the 14-day ATR is 2222.34, the stop price is 72805.12, the target price is 92756.40, the net risk is 6.41%, the number of historical blocks is 4634, the number of simulations is 20000, the percentage of first stop is 3.36%, the percentage of first target is 0.01%, the 1st percentile simulated return is -6.99%, and the worst simulated return is -10.81%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; stats.py C91202

A new comparative experiment was conducted with a 2R trail rule that was not frozen. The results are shown in terms of return, drawdown, win rate, and mean trade in percent. The data is based on historical current-universe bias, with missing bars reindexed and no entry-bar exit. Ideal intrabar stops were not used in live 15-minute trades. Every cell was retained. The results are as follows: for UNI with an hourly time frame, a limit of 30, a slip of 0.001, and a start date of 2026-01-01, the return percentage was 28.284310519293076, the drawdown percentage was 20.108710422611765, the number of trades was 3, the win rate was 66.66666666666666, the mean trade was 14.4721679361782, the average win was 26.000942925147108, and the average loss was -8.585382041759615. For UNI with an hourly time frame, a limit of 30, a slip of 0.002, and a start date of 2024-01-01, the return percentage was -41.08687055518339, the drawdown percentage was 61.66962896590247, the number of trades was 15, the win rate was 26.666666666666668, the mean trade was -1.5040219587699477, the average win was 31.36961063231848, and the average loss was -13.458070173711192. For UNI with an hourly time frame, a limit of 30, a slip of 0.002, and a start date of 2026-01-01, the return percentage was 27.995573915220074, the drawdown percentage was 20.108710422611765, the number of trades was 3, the win rate was 66.66666666666666, the mean trade was 14.443783713009717, the average win was 26.000138359064373, and the average loss was -8.668925579099596. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was conducted with a 2R trail rule that was not frozen. The results are reported in terms of return, drawdown, win rate, and mean trade in percent. The experiment used historical current-universe bias, reindexed missing bars, and did not use entry-bar exit or ideal intrabar stops that are not live for 15 minutes. Every cell was retained. The results are as follows: for UNI with an hourly time frame, a length of 20, a slip of 0.002, and a start date of 2024-01-01T00:00:00Z, the return percentage was -14.147679738672302, the drawdown percentage was 55.01293069044214, the number of trades was 15, the win rate was 33.33333333333333 percent, the mean trade was 0.8802181850772439 percent, the average win was 27.719373294931515 percent, and the average loss was -12.539359369849896 percent. For the same settings but with a start date of 2026-01-01T00:00:00Z, the return percentage was -16.83139824947808, the drawdown percentage was 46.396411169802555, the number of trades was 6, the win rate was 16.666666666666664 percent, the mean trade was 0.08298392678501298 percent, the average win was 37.39755175278112 percent, and the average loss was -7.379929638414209 percent. For UNI with an hourly time frame, a length of 30, a slip of 0.001, and a start date of 2024-01-01T00:00:00Z, the return percentage was -40.664061694857445, the drawdown percentage was 61.31849026501961, the number of trades was 15, the win rate was 26.666666666666668 percent, the mean trade was -1.4736130210736675 percent, the average win was 31.255440680075026 percent, and the average loss was -13.375087094218644 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was conducted without a frozen trail rule that allows for a 2R trail. The results are reported as return, drawdown, win rate, and mean trade in percent. The data is based on historical current-universe bias, with missing bars reindexed and no entry-bar exit. Ideal intrabar stops were not used in live 15-minute trades. Every cell was retained. The experiment included the following results for UNI/USD: a return of -3.60 percent, a drawdown of 37.54 percent, 6 trades, a win rate of 16.67 percent, and a mean trade of -0.67 percent, starting on 2026-01-01; a return of -13.33 percent, a drawdown of 54.82 percent, 15 trades, a win rate of 33.33 percent, and a mean trade of 0.93 percent, starting on 2024-01-01; and a return of -16.14 percent, a drawdown of 45.92 percent, 6 trades, a win rate of 16.67 percent, and a mean trade of 0.18 percent, starting on 2026-01-01. The data included average win and loss amounts, as well as the number of targets and missing values. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was conducted without a frozen trail rule and a 2R trail. The results are reported in percent for return, drawdown, and win rate, with a historical current-universe bias and missing bars reindexed, and no entry-bar exit, using ideal intrabar stops, but not live for 15 minutes. Every cell was retained. The experiment included three scenarios: one starting on 2024-01-01T00:00:00Z with a slip of 0.001, one starting on 2026-01-01T00:00:00Z with a slip of 0.001, and one starting on 2024-01-01T00:00:00Z with a slip of 0.002. The results were as follows: for the first scenario, the return was -51.94678548544584 percent, the drawdown was 68.38356562128361 percent, there were 20 trades, the win rate was 25 percent, the mean trade was -2.2029553687834538 percent, the average win was 23.82876602102521 percent, and the average loss was -10.880195832053007 percent. For the second scenario, the return was -2.816512879252009 percent, the drawdown was 37.03809208871728 percent, there were 6 trades, the win rate was 16.666666666666664 percent, the mean trade was -0.5743441587260738 percent, the average win was 31.80954447035196 percent, and the average loss was -7.051121884541681 percent. For the third scenario, the return was -52.926370877764974 percent, the drawdown was 68.99631229721892 percent, there were 20 trades, the win rate was 25 percent, the mean trade was -2.2883844320533773 percent, the average win was 23.882490691934326 percent, and the average loss was -11.012009473382612 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was conducted with a 2R trail rule that was not frozen. The results are reported in percent for return, drawdown, and win rate, with a historical current-universe bias and missing bars reindexed, and no entry-bar exit, using ideal intrabar stops, but not live for 15 minutes. Every cell was retained. The experiment included the following results: for UNI with a daily time frame, a leverage of 30, a slip of 0.001, and a start date of 2026-01-01, the return percentage was 4.716744402699158, the drawdown percentage was 20.991756830838405, the number of trades was 1, the win rate was 0, and the mean trade was -10.319171104872659. For UNI with a daily time frame, a leverage of 30, a slip of 0.002, and a start date of 2024-01-01, the return percentage was -42.38187181964514, the drawdown percentage was 67.20056931727414, the number of trades was 14, the win rate was 35.714285714285715, the mean trade was -2.249742173866249, the average win was 18.20953986837527, and the average loss was -13.616009975111536. For UNI with a daily time frame, a leverage of 30, a slip of 0.002, and a start date of 2026-01-01, the return percentage was 4.518824436545854, the drawdown percentage was 20.98344745544377, the number of trades was 1, the win rate was 0, and the mean trade was -10.399250521279091. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was done without a frozen 2R trail rule. The results are shown for the UNI/USD pair, including return, drawdown, win rate, and mean trade in percent. The historical current-universe bias was used, with missing bars reindexed and no entry-bar exit. Ideal intrabar stops were not used in live 15-minute trading. Every cell was retained. The results are as follows: for a daily time frame with a length of 20, a slip of 0.002, and a start date of 2024-01-01, the return percentage was -38.75436907657814, the drawdown percentage was 64.84611783731292, the number of trades was 14, the win rate was 42.857142857142854, the mean trade was -1.8138398421859763, the average win was 16.085075689481766, and the average loss was -15.238026490936782. For a daily time frame with a length of 20, a slip of 0.002, and a start date of 2026-01-01, the return percentage was 32.55705116550591, the drawdown percentage was 18.871806489102905, the number of trades was 4, the win rate was 50, the mean trade was 9.75574624481364, the average win was 32.199588799351844, and the average loss was -12.68809630972456. For a daily time frame with a length of 30, a slip of 0.001, and a start date of 2024-01-01, the return percentage was -41.68708084992892, the drawdown percentage was 66.6853146978827, the number of trades was 14, the win rate was 35.714285714285715, the mean trade was -2.175538900972127, the average win was 18.245037303093113, and the average loss was -13.52030345878615. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was conducted without a frozen trail rule that allows for a 2R trail. The results are reported as return, drawdown, and win rate in percent, taking into account historical current-universe bias, missing bars reindexed, and no entry-bar exit, with ideal intrabar stops not applied in real-time on a 15-minute basis. Every cell from the experiment was retained. The experiment included three scenarios: one with ticker UNI, daily time frame, 15 length, 0.002 slip, starting on 2026-01-01, which had a return of 38.703735945487836 percent, a drawdown of 27.058242109550314 percent, 5 trades, a win rate of 40 percent, a mean trade of 8.643598281414485 percent, an average win of 32.19958879935185 percent, and an average loss of -7.060395397210424 percent. Another scenario had the same ticker and time frame, but with a length of 20, 0.001 slip, starting on 2024-01-01, resulting in a return of -38.0179095001473 percent, a drawdown of 64.29387653931362 percent, 14 trades, a win rate of 42.857142857142854 percent, a mean trade of -1.7387638902040536 percent, an average win of 16.149846396568808 percent, and an average loss of -15.1552216052837 percent. The last scenario had the same parameters as the second, but started on 2026-01-01, and had a return of 32.83939603366497 percent, a drawdown of 18.886746159805302 percent, 4 trades, a win rate of 50 percent, a mean trade of 9.800102763066135 percent, an average win of 32.21280470247487 percent, and an average loss of -12.612599176342604 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was done on the UNI/USD trading pair. This experiment did not use a frozen trail rule, which is a rule that helps decide when to sell or buy, and it used a 2R trail rule, where R means the risk amount. The results are shown in percentages and include return, drawdown, win rate, and mean trade. The data was taken from the daily time frame, which means the data is based on daily price changes. The slip, or the difference between expected and actual prices, was 0.001 or 0.002. The experiment started on January 1, 2024, or January 1, 2026. The results are as follows: for the experiment starting on January 1, 2024, with a slip of 0.001, the return was -63.97%, the drawdown was 79.59%, there were 17 trades, the win rate was 29.41%, and the mean trade was -4.46%. For the experiment starting on January 1, 2026, with a slip of 0.001, the return was 39.43%, the drawdown was 26.91%, there were 5 trades, the win rate was 40%, and the mean trade was 8.74%. For the experiment starting on January 1, 2024, with a slip of 0.002, the return was -64.53%, the drawdown was 79.92%, there were 17 trades, the win rate was 29.41%, and the mean trade was -4.54%. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202

A new comparative experiment was done without a frozen 2R trail rule. The results are shown as return, drawdown, win rate, and mean trade in percent. The data is based on historical current-universe bias, with missing bars reindexed and no entry-bar exit. Ideal intrabar stops were not used in live 15-minute trades. Every cell was retained. The results for LINK/USD are as follows: for a start date of 2026-01-01T00:00:00Z, the return percentage was -8.22, the drawdown percentage was 29.04, the win rate was 28.57, and the mean trade was -0.46. For a start date of 2024-01-01T00:00:00Z, the return percentage was 69.88, the drawdown percentage was 42.97, the win rate was 50, and the mean trade was 7.97. For a start date of 2026-01-01T00:00:00Z with a different slip value, the return percentage was -13.23, the drawdown percentage was 29.37, the win rate was 33.33, and the mean trade was -1.79. A slip value refers to the difference between expected and actual trade prices. Drawdown means the largest decrease in value. Win rate is the percentage of winning trades. Mean trade is the average profit or loss per trade. https://runway.adityaperswal.workers.dev/data/crypto-bars ; experiment.py C91202