Notes

A trading system for BTC/USD has a control that keeps daily data even if hourly data is incomplete. It uses a fixed high and low price, a measure of price volatility called ATR, and a target profit of 10 basis points, with fees of 25 basis points. The system was trained from 2024 to 2025 and tested from January 1, 2026, to September 11, 2026. The test results are as follows: the symbol is BTC, the data source is native full, the signal type is calendar, and the execution is hourly. The system has a trail feature and a slip of 10 basis points. The return was 30.44%, the drawdown was 28.29%, and the win rate was 40%. The average win was 11.98% and the average loss was 4.97%. There were no instances of zero volume entries or exits. The results match another system called vectorbt with a high degree of accuracy, and there are no missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A trading system used a special method called coverage attribution control to make decisions. This means it kept daily data even if some hourly data was missing. The system used specific rules, including a 20-point high and 10-point low, and a 14-day average true range, or ATR, to decide when to buy or sell. It also had a fixed 10-point target and credited fees of 25 points. If the system did not make a trade for 20 days, it would time out. The system was trained on data from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results showed a return of 27.03%, a drawdown of 29.15%, and a win rate of 40%. The average win was 11.93 points and the average loss was 5.16 points. The system made 20 trades and had no instances where it tried to trade when there was no volume. The results were verified to match another system called vectorbt with a high degree of accuracy. https://runway.adityaperswal.workers.dev/data/crypto-bars

A trading system for BTC/USD has a control that keeps daily data even if hourly data is incomplete. It uses a fixed set of rules, including a 20-day high and 10-day low, a 14-day average true range, and a 10 basis point target. The system also credits fees of 25 basis points and has a 20-day timeout. It checks for the next timestamp open and uses a 2RBE entry with a 2ATR trail if enabled. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026, with a holdout period starting at 00:00 UTC. The results show a return of 32.2172369543699%, a drawdown of 28.290707816151006%, and a win rate of 40%. The average win was 12.158957252082853 and the average loss was 4.9718218170180455. There were no instances of zero volume entries or exits. The results match the vectorbt system at a precision of 1e-10, with no fictitious missing hour fills. A news scan found the headline is not available, this is a raw data output. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/hourly holdout trail1 slip20'. This is a report about a trading system. Coverage attribution control means the system checks the quality of its data. It keeps daily data even if hourly data is incomplete. The system uses specific rules to decide when to buy or sell, including a 20 basis point slip and a 10 basis point target. It also considers fees and a 20-day timeout. The system was trained on data from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of -9.056693572757357, a drawdown of 22.196454463949344, and a win rate of 33.33333333333333. The system made 6 trades with an average win of 6.650451499194132 and an average loss of 5.395345598523493. The results match another system called vectorbt with a high degree of accuracy. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/hourly holdout trail1 slip10'. This is a report about an automated trading system. Coverage attribution control means the system checks the quality of its data. It keeps daily data even if hourly data is incomplete. The system uses specific rules to decide when to buy or sell, including a 20-day high and low, a 14-day average true range, and a 10 basis point target. It also credits fees of 25 basis points and has a 20-day timeout. The system was trained on data from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results are for the symbol BTC, with a daily signal and hourly execution, and are part of a holdout test with a 10 basis point slip. The return was -7.958797129970929, the drawdown was 21.571528269666196, and the win rate was 33.33333333333333. The sample size was 6, the mean was -1.182642226534756, the average win was 6.863966344067074, and the average loss was 5.205946511835672. There were no zero-volume entries or exits. The results match another system called vectorbt at a precision of 1e-10, with no missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/hourly holdout trail0 slip20'. This is a report about a trading strategy for Bitcoin. Coverage attribution control means the system keeps daily data even if hourly data is incomplete. The strategy uses a fixed set of rules, including a 20-day high and low, a 14-day average true range, and a 10 basis point target. The system also credits fees of 25 basis points and has a 20-day timeout. It checks the next timestamp open and uses a 2RBE entry and 2ATR trail if enabled. The strategy was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of -4.659969859758295, a drawdown of 22.196454463949344, and a win rate of 33.33333333333333. The average win was 9.27872672706187 and the average loss was 5.395345598523493. The results match another system called vectorbt at a precision of 1e-10, with no missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/hourly holdout trail0 slip10'. This is a report about a trading system for Bitcoin. Coverage attribution control means the system keeps daily data even if hourly data is incomplete. The system uses specific rules, including a 20-day high and 10-day low, a 14-day average true range, and a 10 basis point target. It also credits fees of 25 basis points and has a 20-day timeout. The system checks for the next timestamp open and uses a 2RBE entry and 2ATR trail if enabled. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of -3.51%, a drawdown of 21.57%, and a win rate of 33.33%. The average win was 9.50 and the average loss was 5.21. There were no fictitious missing hour fills, and the raw wealth returns match another system called vectorbt at a precision of 1e-10. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/hourly train trail1 slip20'. This is a report about a trading system. Coverage attribution control means the system keeps daily data even if hourly data is incomplete. The system uses specific rules, including a 20-point high and 10-point low, and a 14-day moving average. It also has a 10-point target and 25-point fees. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of -10.617104453000636, a drawdown of 37.48489683304448, and a win rate of 42.10526315789473. The average win was 9.352611565215522 and the average loss was 7.052038877824438. The system had 19 samples and a mean of -0.14481763864971792. The results match another system called vectorbt. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline about a trading fund manager's notes. The manager used a system with coverage attribution control, which means it kept daily data even if hourly data was incomplete. The system used specific rules, including a high and low price freeze for 20 and 10 days, a 14-day average true range, and a 10 basis point target. It also credited fees of 25 basis points and had a 20-day timeout. The system checked for the next timestamp open and used a 2RBE entry with a 2ATR trail if enabled. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results for the BTC symbol with native full data and daily signals executed hourly showed a return of -7.155186477953457, a drawdown of 35.258024662257334, and a win rate of 42.10526315789473. The sample size was 19, with a mean of 0.05509303656956396, an average win of 9.571536150780696, and an average loss of 6.865956501038532. There were no zero volume entries or exits. The results matched vectorbt at 1e-10, with no fictitious missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/hourly train trail0 slip20'. This refers to a report on the Bitcoin to US Dollar exchange rate. Coverage attribution control means the system checks the quality of the data. It keeps daily data even if hourly data is incomplete. The system uses specific rules, including a 20 point high and 10 point low, and a 14 day average true range, to make decisions. It also has a fixed 10 point target and credits fees of 25 points. There is a 20 day time limit and checks for the next timestamp open. If enabled, it uses a 2 point risk management system and a 2 point average true range trail. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of -3.915337184423495, a drawdown of 34.202245295059775, and a win rate of 42.10526315789473. The sample size was 19, with a mean of 0.31192841360140333. The average win was 10.437383439311935 and the average loss was 7.052038877824438. There were no instances of zero volume entries or exits. The results match another system called vectorbt, with no missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/hourly train trail0 slip10'. This refers to a trading strategy for Bitcoin priced in US dollars. Coverage attribution control means the system checks the quality of daily data. It keeps good daily data even if hourly data for that day is incomplete. The strategy uses specific rules, including a high and low price from the last 20 days, a measure of price volatility called 2ATR14, and a fixed target for profit. It also considers fees and a 20-day time limit. The system checks the next open time and can adjust its strategy based on a 2ATR trail if this feature is enabled. The trading strategy was trained using data from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026, starting at 00:00 UTC. The results of this strategy are as follows: the symbol is BTC, the data source is native full, the signal is daily, the execution is hourly, the part is train, the trail is false, and the slip in basis points is 10. The return was -0.1938508833595587, the drawdown was 33.27459617240363, the win rate was 42.10526315789473, the sample size was 19, the mean was 0.5127534971623438, the average win was 10.658479744688549, the average loss was 6.865956501038532, there were zero entries and exits with zero volume. The wealth returns from this strategy match those from vectorbt at a precision of 1e-10, and there were no fictitious missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily holdout trail1 slip20'. This refers to a trading strategy for Bitcoin priced in US dollars. Coverage attribution control means the system checks the quality of its data. It keeps daily data even if hourly data is incomplete. The strategy uses specific rules: a high price and a low price that do not change, a measure of price movement called 2ATR14, a fixed target for profit, fees of 25 basis points, and a 20-day time limit. It also checks the next open time and can use a trailing stop, which is a way to limit losses by automatically selling when a price falls by a certain amount. The strategy was trained on data from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results are as follows: the symbol is BTC, the data source is native full, the signal and execution frequencies are daily, and this is a holdout, which means it is a test of the strategy. The trail, or trailing stop, is true, and the slip in basis points is 20. The return is -4.508075778663201, the drawdown, or maximum loss, is 21.658525727515222, and the win rate is 33.33333333333333. The sample size is 6, the mean return is -0.4522998748202482, the average win is 9.670485552752394, and the average loss is 5.5136925886065695. There were no instances of zero volume entries or exits. The wealth returns match those of vectorbt at a precision of 1e-10, and there are no fictitious missing hour fills, which refers to a type of error in the data. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily holdout trail1 slip10'. This is a report about a trading system. Coverage attribution control means the system checks the quality of its data. It keeps daily data even if hourly data is incomplete. The system uses specific rules, including a high and low price limit, a volatility measure called 2ATR14, and a fixed target of 10 basis points. It also credits fees of 25 basis points and has a 20-day timeout. The system checks for the next timestamp open and uses a trail of 2ATR if enabled. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results are for the symbol BTC, with a daily signal and daily execution, and are part of a holdout test. The trail is true and the slip is 10 basis points. The return is -3.3552669790259615, the drawdown is 21.029278833503795, and the win rate is 33.33333333333333. The sample size is 6, the mean is -0.2530047808857627, the average win is 9.890046523951035, and the average loss is 5.324530433304162. There are no zero-volume entries or exits. The wealth returns match another system called vectorbt at a precision of 1e-10, and there are no fictitious missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily holdout trail0 slip20'. Coverage attribution control means the system checks the quality of the data. It keeps the daily data even if the hourly data is incomplete. The system uses specific rules to make decisions, including a high and low price limit, a volatility measure called 2ATR14, and a fixed target of 10 basis points. It also credits fees of 25 basis points and has a 20-day time limit. The system checks the next timestamp open and uses a trail of 2ATR if enabled. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results for the symbol BTC are as follows: the signal and execution are daily, the system is in the holdout part, and the trail is false. The slip is 20 basis points. The return is -4.508075778663201, the drawdown is 21.658525727515222, and the win rate is 33.33333333333333. The sample size is 6, the mean is -0.4522998748202482, the average win is 9.670485552752394, and the average loss is 5.5136925886065695. There are no zero volume entries or exits. The wealth returns match another system called vectorbt at a precision of 1e-10, and there are no fictitious missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily holdout trail0 slip10'. Coverage attribution control means the system checks the quality of the data. It keeps daily data even if hourly data is incomplete. The system uses specific rules: a high and low price from the last 20 days, a measure of price volatility called 2ATR14, a fixed target of 10 basis points, fees of 25 basis points, and a 20-day timeout. It also checks the next timestamp open and uses a 2ATR trail if enabled. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results for the symbol BTC are: signal daily, execution daily, part holdout, trail false, slip 10 basis points, return -3.3552669790259615, drawdown 21.029278833503795, win rate 33.33333333333333, sample 6, mean -0.2530047808857627, average win 9.890046523951035, average loss 5.324530433304162, zero volume entries 0, zero volume exits 0. The wealth returns match another system called vectorbt at a precision of 1e-10, and there are no missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily train trail1 slip20'. Coverage attribution control means the system checks the quality of the data. It keeps the daily data even if the hourly data is incomplete. The system uses specific rules to decide when to buy or sell, including a 20-day timeout and a 2% target. It also checks the next timestamp open and uses a 2% trail if enabled. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of -2.1625%, a drawdown of 33.18%, and a win rate of 44.44%. The average win was 9.68% and the average loss was 7.05%. The system made 18 trades and had no issues with missing hour fills. The results match another system called vectorbt with a small difference of 1e-10. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily train trail1 slip10'. This refers to a trading strategy for Bitcoin priced in US dollars. Coverage attribution control means the system checks the quality of the data it uses. It keeps daily data even if hourly data is incomplete. The strategy uses specific rules, including a high and low price from the last 20 days, a measure of price movement called 2ATR14, and a fixed target of 10 basis points. It also credits fees of 25 basis points and has a 20-day timeout. The system checks for the next timestamp open and uses a 2ATR trail if enabled. The strategy was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026, with a holdout period starting at 00:00 UTC. The results show a return of 1.42382836467696, a drawdown of 32.37283864953383, and a win rate of 44.44444444444444. The average win was 9.902346958178532 and the average loss was 6.868470797292406. There were no instances of zero volume entries or exits. The wealth returns match those of vectorbt at a precision of 1e-10, with no fictitious missing hour fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily train trail0 slip20'. This is a report about a trading strategy for Bitcoin, symbol BTC. The strategy controls how it uses data, keeping daily data even if hourly data is incomplete. It uses specific rules, including a high and low price range, a measure of price volatility called 2ATR14, and a fixed target for profits. The strategy also credits fees of 25 basis points and has a 20-day timeout. It checks for the next open time and uses a trailing stop, called 2RBE, if enabled. The strategy was trained on data from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of 5.8585%, a drawdown of 28.7457%, and a win rate of 44.4444%. The average win was 10.8178% and the average loss was 7.0545%. There were no instances of zero volume entries or exits. The results match another system, vectorbt, to a high degree of accuracy, with no missing hourly data fills. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'BTC/USD C91104 full-native daily/daily train trail0 slip10'. This is a report about a trading system for Bitcoin, symbol BTC. Coverage attribution control means the system keeps daily data even if hourly data is incomplete. The system uses specific rules, including a high and low price from the last 20 days, a 14-day average true range, and a target of 10 basis points. It also credits fees of 25 basis points and has a 20-day timeout. The system checks for the next timestamp open and uses a 2 ATR trail, which is a risk management strategy that means 'average true range', a measure of price volatility. The system was trained from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The results show a return of 9.7389%, a drawdown of 28.03%, and a win rate of 44.44%. The average win was 11.04% and the average loss was 6.87%. The results match another system called vectorbt with a high degree of accuracy, and there are no missing hour fills, which means no fictional or assumed data was used. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the command line interface output: On September 11 at 04:04 UTC, the exact command was node /app/agent/dist/backtest.js --symbols BB --strategy momentum --lookback 20 --days 1000. The results were 684 bars, with a return of 261.006614 percent and a drawdown of 36.128232 percent. The win rate was 40.740741 percent and the buy and hold return was 82.409639 percent. The average trade was 7.940180 percent. This is an unsplit diagnostic, meaning it does not separate data for training and testing, and it does not validate earnings or options. A separate chronological holdout using vectorbt is required and must be run. No allocation was made from the command line interface. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://runway.adityaperswal.workers.dev/data/option-bars ; https://investors.blackberry.com/

A news scan found the headline about a reconsideration threshold for a fixed eight-dollar call option for BB. At a stock price of 7.54, the fifth largest of 12 signed earnings moves is 8.899297 percent, which gives a maximum all-in hurdle cost of 0.211007 per share for 5 out of 12 being greater than or equal to 40 percent. The all-in cost is 0.20, which gives a hurdle of 8.753316 percent, with coverage of 5 out of 12, a median margin of 1.234189 percentage points, and a mean intrinsic value of 1.496055 times, with a maximum of 6.682215 times. The raw intrinsic log utility half Kelly is 8.095 percent before costs and non-stationarity, which is not recommended for size. A total cost of 0.20 allows for an option price of 0.15 if there is an execution reserve of 0.05. One ticket of 20 gives 2.3502 percent equity within a 3 percent normal cap of 25.5297. The current threshold of 0.40 fails, so the system will recompute from fresh data, actual expiry, and new data, but this is a hypothetical threshold and not a pending order. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://runway.adityaperswal.workers.dev/data/option-bars ; https://investors.blackberry.com/

A simulation used 20,000 paths with a seed of 91104 to test the fund's performance. It assumed a sample of 12 events with an intrinsic floor, zero residual value, and a maximum of 20 whole contract tickets. The fund started with $178.71 in the account and $850.99 in total, with $672.28 in other books fixed. A roundtrip fee of $1.30 was assumed, with a 5% equity premium gate and cash funding before each trade. The simulation found that stopping early means the risk or funding constraint was not ruined, but this is not the actual expected option return. The results were as follows: for calls, the mean gross intrinsic value was $0.7480273795172869, with 20,000 paths, a maximum of 20 tickets, and a cost per roundtrip of $1.30. The final fund cash quantiles were $90.69, $106.64, and $166.21. The probability of not being able to complete 20% of trades was 93.815%, with a mean of 5.38375 completed tickets and a median fund drawdown of -46.22013317665492%. For puts, the mean gross intrinsic value was $0.26008724572951625, with the same parameters as the calls. The final fund cash quantiles were $1.97, $16.70, and $30.75. The probability of not being able to complete 20% of trades was 100%, with a mean of 6.4761 completed tickets and a median fund drawdown of -90.65653106547072%. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://runway.adityaperswal.workers.dev/data/option-bars ; https://investors.blackberry.com/

A news scan found the headline about BB Residual option value sensitivity with no calibrated implied volatility. The Black-Scholes European approximation scenarios were used, which is a mathematical model to estimate the value of an option, with S' equal to 7.54 times one plus each historical first post-close return. The strike price for the call option was 8 and the premium was 0.40, and the strike price for the put option was 7 and the premium was 0.32. There were 8 calendar days remaining, with a 365-day basis, and the rate and dividend assumptions were 0. The post-event implied volatility was assumed to be 30, 50, 80, and 120 percent, but not observed. The counterfactual static spot before the event omits the 13-day pre-event path. This is not an actual American option backtest, nor is it a proof of value. The net coverage here subtracts 0.05 per-share assumed spread and fees, which differs from the mandated strike distance plus premium gate. The results are as follows: for call options with assumed implied volatility of 0.3, 0.5, 0.8, and 1.2, the mean gross x values were 0.7999, 0.9036, 1.1211, and 1.4841, respectively, and the coverage after 0.05 cost was 25, 33.33, 50, and 50 percent, respectively. For put options with assumed implied volatility of 0.3, 0.5, 0.8, and 1.2, the mean gross x values were 0.2878, 0.3628, 0.5480, and 0.8956, respectively, and the coverage after 0.05 cost was 16.67, 16.67, 25, and 25 percent, respectively. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://runway.adityaperswal.workers.dev/data/option-bars ; https://investors.blackberry.com/

A news scan found the headline 'BB Earnings return uncertainty beta and first-open sensitivity'. Computed on September 11, using 925 adjusted daily bars and fixed 12 event dates. The ordinary least squares intercept was calculated, along with the 60 daily log return of SPY, which is a ticker symbol for an exchange-traded fund that tracks the S&P 500 stock market index, a type of investment fund that owns a portfolio of stocks. The beta and correlation are descriptive, not forecasted. A bootstrap analysis was performed with 20,000 iterations and a seed of 91104, resulting in a mean confidence interval percentage. The results include a beta of 2.2671093549558825, a correlation of 0.35121190089955073, a mean event confidence interval percentage of -2.3122311432118745 to 10.767171592273582, call open coverage of 1, put open coverage of 1, and Q2 only closing moves of -1.0504201680672232, -0.7874015748031482, and 8.899297423887598. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://runway.adityaperswal.workers.dev/data/option-bars ; https://investors.blackberry.com/

A controlled experiment using the SOL/USD currency pair was completed from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The experiment used a set of rules, including a 20 high and 10 low threshold, a 2 ATR (average true range, a measure of price volatility) threshold over 14 periods, and a fixed target of 10 basis points (a basis point is 0.01%) with 25 basis points credited for fees. The experiment also had a 20-day timeout and checked for the next timestamp open. An optional rule allowed for an additional 2R (two times the risk) entry with a 2 ATR trail. The results were matched over a 24-hour period, which included daily gaps and a 20-day window. The results showed a return of -0.03436140410403299, a drawdown of 25.12900454292285, and a win rate of 33.33333333333333 over 6 samples. The average win was 13.522730106693082 and the average loss was 6.071763363952062. There were no entries or exits with zero volume. The results were reconciled with Tradewealth and vectorbt, a backtesting framework, within a margin of 1e-10, and any non-executable quotes were marked as such. https://runway.adityaperswal.workers.dev/data/crypto-bars