Notes

The neighbor10_15_test model had a return of -14.35135055833916 percent from January 4, 2021, to September 10, 2026. The drawdown was 53.040254828000165 percent and the compound annual growth rate, or CAGR, was -2.6967938750516285 percent. The win rate was 44.18604651162791 percent. There were 258 samples, with a mean trade percentage of -0.027747137076297307 percent. The average win percentage was 2.2864677816784416 percent and the average loss percentage was 1.8598339477571322 percent. The mean confidence interval, or CI95, percentage was between -0.31033285590242954 and 0.26975274239519337 percent. The half Kelly, a measure of optimal bet size, was 0. The reasons for trades included gap target, stop, time, target, and gap stop. There were 35 partials and 23 same day trades. The vbt orders, or trades, were 551 and the trade months were 68. The market had a beta, or volatility measure, of 0.49006772582105573, an alpha, or excess return, of -8.656648149308538 percent per year, and a correlation, or corr, of 0.47171662385134105. This is a daily proxy report, with all 20 cells reconciled, based on training data from 2016 to 2020 and testing from 2021 to September 10, 2026. The history was inspected and no globally untouched claim was found. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

A controlled experiment using a trading strategy was completed from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The strategy had 20 high and 10 low settings, with a 2 average true range over 14 days, a fixed target of 3 times the risk, fees of 25 basis points, and a 20-day timeout. The strategy also included checks for the next timestamp open and an optional 2 times risk entry with a 2 average true range trail. The experiment used a 24-hour matching system to mask daily gaps and a 20-day window, but did not exactly replicate the prior strategy. The results of the experiment were as follows: the symbol was BTC, the source was native matched, the signal and execution were daily, and the part was holdout with a trail and a 20 basis point slip. The return was 6.3528700788678005, the drawdown was 12.748217160474052, the win rate was 40, the sample size was 5, the mean was 1.4996760242032914, the average win was 9.670485552752394, and the average loss was 3.9475303281627774. There were no zero volume entries or exits. The trade wealth was reconciled using vectorbt with a precision of 1e-10, and non-executable quotes were marked. https://runway.adityaperswal.workers.dev/data/crypto-bars

A controlled experiment using a trading strategy was completed from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The strategy had 20 high and 10 low settings, with a 2 ATR (Average True Range, a measure of price volatility) of 14, a net target of 3R (a risk-reward ratio) fixed at 10 basis points, and credited fees of 25 basis points. It also had a 20-day timeout and checked for the next timestamp open. Additionally, it had an optional 2RBE (Risk-Reward Ratio) entry with a 2 ATR trail. The experiment matched the BTC symbol with a native matched source, using a daily signal and daily execution on the holdout part, with a 10 basis point slip. The results showed a return of 7.421736699555104, a drawdown of 12.22313155668493, and a win rate of 40. The sample size was 5, with a mean of 1.7028789856190385, an average win of 9.890046523951035, and an average loss of 3.7552327066022926. There were no zero volume entries or exits. The trade wealth reconciles with vectorbt within a small margin of error, and marks quotes that are not executable. https://runway.adityaperswal.workers.dev/data/crypto-bars

The REV9110400 frozen cell neighbor10_15_train report shows the following results: the model used was neighbor10, the period was train, and the slippage in basis points was 15. The first trade was on 2016-01-04 05:00:00+00:00 and the last trade was on 2020-12-31 05:00:00+00:00. The return was -27.823110091738133, the drawdown was 51.516471177607514, and the compound annual growth rate, or CAGR, which is a measure of an investment's annual growth rate, was -6.317787585841284. The win rate, which is the percentage of winning trades, was 43.89140271493213. There were 221 samples, the mean trade percentage was -0.12208391241790083, the average win percentage was 1.847016078570865, and the average loss percentage was 1.6624282602075005. The mean confidence interval 95 percentage, which is a measure of the range of possible values, was between -0.3248899491009365 and 0.09155388045638681. The half Kelly, a formula used to determine the optimal bet size, was 0. The reasons for trades included time, which refers to a trade based on a specific time, 113 times, gap stop, which is a stop-loss order based on a price gap, 7 times, stop, which is a stop-loss order, 98 times, gap target, which is a take-profit order based on a price gap, 2 times, and target, which is a take-profit order, 1 time. There were 18 partials, which are partial trade executions, and 33 same-day trades. The vbt, or vector-based trading, reconciled and orders were 460. The trade months were 59, and the market had a beta, which measures volatility, of 0.45967623262438806, an alpha annual percentage, which measures excess return, of -12.160703786962777, and an alpha t HAC5, which is a measure of excess return with a specific statistical adjustment, of -2.080372716554282. The correlation, which measures the relationship between two variables, was 0.45896132281058627. This is a daily proxy report, and all 20 cells were reconciled. The training period was 2016-2020 and the test period was 2021-Sep10. The history was inspected and there were no globally untouched claims. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

A controlled experiment using a trading strategy was completed from 2024 to September 11, 2026. The strategy had 20 high and 10 low settings, with a 2 average true range over 14 periods, a fixed target of 10 basis points, and fees of 25 basis points. The experiment also had a 20-day timeout and checked for the next timestamp open. It used a daily matching system and a 20-day window, but did not exactly replicate the prior strategy. The results showed a return of 6.35%, a drawdown of 12.75%, and a win rate of 40%. The average win was 9.67% and the average loss was 3.95%. The trade wealth was reconciled using vectorbt, which is a backtesting library, and marked unexecutable quotes, meaning some quotes were not usable for trading. The symbol used was Bitcoin, or BTC, and the source was native matched, with a daily signal and daily execution, and it was part of a holdout, or test, group, with a slip of 20 basis points and no trailing stop, or trail, which is a strategy to limit losses. https://runway.adityaperswal.workers.dev/data/crypto-bars

A controlled experiment using a trading strategy was completed from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The strategy had 20 high and 10 low settings, with a 2 ATR 14, which means a 2 times average true range over 14 periods, and a fixed 10 basis point target, with 25 basis points credited for fees. It also had a 20-day timeout and checks for the next timestamp open, with an optional 2RBE, or risk-based entry, and a 2 ATR trail. The experiment matched the native strategy daily, with a full 24-hour matching that included daily gaps and a 20-day window, but it did not exactly replicate the prior strategy. The results were as follows: the symbol was BTC, the source was native matched, the signal and execution were daily, and the part was holdout, with no trail and a 10 basis point slip. The return was 7.421736699555104, the drawdown was 12.22313155668493, the win rate was 40, the sample size was 5, the mean was 1.7028789856190385, the average win was 9.890046523951035, and the average loss was 3.7552327066022926, with no zero volume entries or exits. The trade wealth reconciles with vectorbt within a very small margin of error, and marks quotes that are not executable. https://runway.adityaperswal.workers.dev/data/crypto-bars

A daily proxy report for the frozen cell neighbor10_5_test shows the following results: the model used was neighbor10, the period was test, and the slippage in basis points was 5. The first trade was on 2021-01-04 05:00:00+00:00 and the last trade was on 2026-09-10 04:00:00+00:00. The return was 48.30495919030964, the drawdown was 46.27596498361801, and the compound annual growth rate, or CAGR, which is a measure of an investment's annual growth rate, was 7.202260326614196. The win rate, which is the percentage of winning trades, was 48.44961240310077. There were 258 samples, the mean trade percentage was 0.18582287834812702, the average win percentage was 2.313173424406771, and the average loss percentage was 1.8135667326092457. The mean confidence interval 95 percentage, which is a measure of the range of values within which a result is likely to lie, was between -0.11811449501287959 and 0.5023963910054401. The half Kelly criterion, which is a formula used to determine the optimal bet size, was 0.040166222814828983. The reasons for the results include gap target, stop, time, gap stop, and target. There were 41 partials, 23 same day trades, and the vbt, or vector-based trading, was reconciled with 557 orders over 68 trade months. The market had a beta, which is a measure of volatility, of 0.49010809224665536, an alpha annual percentage, which is a measure of excess return, of 1.0473221624477715, and a correlation, which is a measure of the relationship between two variables, of 0.469273609004706. This is a daily proxy report, and all 20 cells have been reconciled. The training period was 2016-2020 and the test period was 2021-2026, and the history was inspected with no globally untouched claim found. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

A controlled experiment to understand how a trading strategy works was completed from January 1, 2024, to September 11, 2026, at 00 UTC. This experiment had 20 high and 10 low points, with a strategy that used a 14-day average true range, a fixed target of 10 basis points, and fees of 25 basis points. The experiment also had a 20-day time limit and checked for the next timestamp open. The results showed a return of -6.7946%, a drawdown of 35.8052%, and a win rate of 41.1765%. The average win was 10.3560 and the average loss was 7.0545. A news scan found the data line with symbol BTC, source native_matched, signal daily, execution daily, part train, and trail true, with a slip of 20 basis points. This experiment used a 24-hour matching system to compare daily gaps and 20-day windows, but did not exactly replicate the prior strategy. The results were checked against Tradewealth and vectorbt, and were found to be accurate within a very small margin of error, with no issues related to unexecutable quotes. https://runway.adityaperswal.workers.dev/data/crypto-bars

A controlled experiment to understand how a trading strategy works was completed from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The strategy had 20 high and 10 low points, with a risk management system called 2 ATR, which stands for Average True Range, a measure of price volatility. The goal was to make a 3% return, with a 0.25% fee, and a 20-day limit to close a trade. The system checked for the next open time and had an optional feature to adjust the strategy based on a 2% return and 2 ATR trail. The experiment matched the market 24 hours a day, considering daily gaps and a 20-day window, but did not exactly copy the previous strategy. The results for the BTC symbol, using a native matching system, with daily signals and daily execution, and a training set, showed a return of -3.57%, a drawdown of 33.78%, and a win rate of 41.18%. The average win was 10.58% and the average loss was 6.87%. There were no trades with zero volume entries or exits. The results were verified to be accurate within a very small margin of error, and any unexecutable quotes were marked as such. https://runway.adityaperswal.workers.dev/data/crypto-bars

A daily proxy report for the frozen cell neighbor10_5_train shows the following results: the model used was neighbor10, the period was train, and the slippage in basis points was 5. The first trade was on 2016-01-04 05:00:00+00:00 and the last trade was on 2020-12-31 05:00:00+00:00. The return was 23.4844213638959, the drawdown was 32.10100999192349, and the compound annual growth rate, or CAGR, which is a measure of an investment's annual return, was 4.312652122160232. The win rate, which is the percentage of winning trades, was 49.7737556561086. There were 221 samples, the mean trade percentage was 0.12159184490584596, the average win percentage was 1.8961583958732682, and the average loss percentage was 1.636987620016824. The mean confidence interval 95 percentage, which is a measure of the range of possible values, was between -0.08560103192469003 and 0.33188233049926. The half Kelly, which is a formula used to determine the optimal bet size, was 0.03206268135891874. The reasons for trades included time, gap stop, stop, gap target, and target, with 123, 8, 87, 2, and 1 occurrences respectively. There were 23 partials and 24 same day trades. The vbt, or vectorized backtest, was reconciled and there were 465 vbt orders. The trade months were 59. The market had a beta, which is a measure of volatility, of 0.4941443867156505, an alpha annual percentage, which is a measure of excess return, of -1.9427376337290452, and a correlation, which is a measure of how much two variables move together, of 0.4907980284351738. This is a daily proxy report, and all 20 cells were reconciled. The training period was 2016-2020 and the test period was 2021-2021-09-10. The history was inspected and there were no globally untouched claims. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

A controlled experiment using a trading strategy was completed from 2024 to 2025 and reused from January 1, 2026, to September 11, 2026. The strategy used a 20-day high and 10-day low, with a 2 ATR (Average True Range, a measure of price volatility) of 14 days, a fixed target of 3R (three times the risk), and fees of 25 basis points (a unit of measurement for interest rates and fees). The strategy also had a 20-day timeout and checked for the next timestamp open. Additionally, it had an optional 2R (two times the risk) entry with a 2 ATR trail (a type of stop-loss order). The experiment used a 24-hour matching mask to match daily gaps and a 20-day window, but did not exactly replicate the prior strategy. The results of the experiment were as follows: the symbol was BTC, the source was native_matched, the signal and execution frequencies were daily, and the part of the experiment was the train trail with no slip, and a slip of 20 basis points. The return was 0.8466368397613167, the drawdown (the peak-to-trough decline) was 31.545607956352217, the win rate was 41.17647058823529, the sample size was 17, the mean return was 0.6486432853019342, the average win was 11.653202479303745, and the average loss was 7.0545481504993335. There were no zero-volume entries or exits. The results were reconciled with Tradewealth and vectorbt, with a tolerance of 1e-10, and marked any non-executable quotes. https://runway.adityaperswal.workers.dev/data/crypto-bars

A controlled experiment using a trading strategy was completed from 2024 to September 11, 2026. The strategy used 20 high and 10 low values, with a 2 average true range over 14 periods, a fixed 10 basis point target, and credited fees of 25 basis points. It also had a 20-day timeout and checked for the next timestamp open. The results showed a return of 4.334386391934331, a drawdown of 29.53125475873297, and a win rate of 41.17647058823529. The average win was 11.876732861705525 and the average loss was 6.868470797292406. A news scan found the symbol BTC, with a source of native_matched, a daily signal, and daily execution. The trade results were matched using a full 24-hour period, with a 20-day window, and a 10 basis point slip. The return and drawdown values were reconciled with tradewealth and vectorbt, with a difference of less than 1e-10. https://runway.adityaperswal.workers.dev/data/crypto-bars

The neighbor3_15_test model had a return of -34.50132234748037 percent from January 4, 2021, to September 10, 2026. The model's drawdown was 44.78913223596377 percent and its compound annual growth rate, or CAGR, was -7.195198353857846 percent. The win rate was 42.5287356321839 percent, based on 261 samples. The average trade percentage was -0.12975630614881453 percent, with an average win of 2.2839040779989435 percent and an average loss of 1.9158649904181555 percent. The reasons for trades included 118 stops, 15 gap stops, 120 time stops, 7 targets, and 1 gap target. The model had 556 orders and traded for 68 months. The market had a beta of 0.4771055470893486, an alpha of -13.17171048137182 percent per year, and a correlation of 0.4560924381158696. A news scan found the headline 'Daily proxy. All 20 cells reconciled. Training 2016–2020/test 2021–Sep 10; inspected history, no globally untouched claim.' https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

The neighbor3_15_train model, also known as a type of automated trading strategy, had a return of -5.0356698911354965 from January 4, 2016, to December 31, 2020. This model, which is a part of the neighbor3 model, experienced a drawdown of 38.241508926979485 and a compound annual growth rate, or CAGR, of -1.028868105718761. The win rate was 43.75 percent out of 224 samples. The average trade percentage was 0.00351529877604099, with an average win percentage of 1.995409086128383 and an average loss percentage of 1.5457354247202248. A news scan found the headline 'Daily proxy. All 20 cells reconciled. Training 2016–2020/test 2021–Sep 10; inspected history, no globally untouched claim.' https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

A daily proxy report for the frozen cell neighbor3_5_test shows the following results: the model used was neighbor3, the period was test, and the slippage in basis points was 5. The first trade was on 2021-01-04 05:00:00+00:00 and the last trade was on 2026-09-10 04:00:00+00:00. The return was 2.281326189643984, the drawdown was 29.391231411424634, and the Compound Annual Growth Rate, which is a measure of investment return, was 0.39885680643896304. The win rate, or percentage of winning trades, was 45.97701149425287. There were 261 trades in the sample, with a mean trade percentage of 0.04138504655656646. The average win percentage was 2.303508964395379 and the average loss percentage was 1.8838267984126356. The mean confidence interval at 95 percent, which is a measure of statistical accuracy, was between -0.2562198984201958 and 0.33990779442520025. The half Kelly criterion, which is a formula for bet sizing, was 0.008983044389286599. The reasons for trades were: stop losses, which is a strategy to limit losses, at 114, gap stop losses at 14, time-based trades at 125, target-based trades at 7, and gap target-based trades at 1. There were 36 partial trades and 32 same-day trades. The report also shows that the trades were reconciled, meaning they were verified for accuracy, and there were 558 orders. The trades spanned 68 months. The market analysis shows a beta, which is a measure of volatility, of 0.4810762109771532, an alpha, which is a measure of investment return, of -5.361733763793577 percent per year, and a correlation, which is a measure of relationship between variables, of 0.4588651848265258. This is a daily proxy report, and all 20 cells were reconciled. The training period was from 2016 to 2020 and the test period was from 2021 to September 10, 2026. The history was inspected and no globally untouched claim was found. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

A daily proxy report for the frozen cell neighbor3_5_train shows the following results: the model used was neighbor3, the period was train, and the slippage in basis points was 5. The first trade was on 2016-01-04 05:00:00+00:00 and the last trade was on 2020-12-31 05:00:00+00:00. The return was 53.19094900422727, the drawdown was 23.00661059110285, and the compound annual growth rate, or CAGR, which is a measure of investment return, was 8.912078617204822. The win rate, or percentage of winning trades, was 47.767857142857146. There were 224 samples, or trades, with a mean trade percentage of 0.2176281277769781. The average win percentage was 2.053633872720923 and the average loss percentage was 1.46145404922304. The mean confidence interval 95 percentage, which is a measure of statistical accuracy, was between -0.062302955218307435 and 0.5202556609315512. The half Kelly, a measure of optimal bet size, was 0.05298610688784461. The reasons for trades were: gap stop 14 times, time 117 times, stop 84 times, target 5 times, and gap target 4 times. There were 30 partials, 28 same day trades, and vbt reconciled was true, with 478 vbt orders. The trade months were 59, and the market had a beta, which measures volatility, of 0.4016246252040945, an alpha annual percentage, which measures excess return, of 3.4911761643260117, and a correlation, which measures relationship between variables, of 0.45041213245668027. This is a daily proxy, all 20 cells were reconciled, and the training period was 2016-2020, with testing from 2021 to September 10. The history was inspected and no globally untouched claim was found. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

A news scan found the headline from PrimaryCision issuer index quotes on September 6, 2023, at 17:30 ET, which was a preliminary report for the second quarter of the fiscal year. This report mentioned a definitive agreement with Cylance, which was classified as held for sale. The issuer's product page, as of February 3, 2025, now shows assets as part of Arctic Wolf. Excluding the September 2023 final report gives 11 events, with a median absolute value of 10.859729%, and a classification of 3 out of 11 as positive and 2 out of 11 as negative. Also, excluding a deal-dependent event in December 2024 gives 10 events, with a median of 9.987505%, and a classification of 2 out of 10 as positive and 2 out of 10 as negative. Post-close earnings were classified as 2 positive out of 1 total. These are sensitivity labels, not the primary sample, which has not been retuned. https://www.newswire.ca/news/blackberry-limited/?page=3&pagesize=25 ; https://www.accesswire.com/957744/blackberry-reports-third-quarter-fiscal-year-2025-results ; https://www.blackberry.com/us/en/campaigns/intelligent-security

The issuer reported on April 9 that the fiscal year 2026, which ended on February 28, had the following results as of September 11: the estimated royalty backlog for QNX was 950 million dollars. The total revenue for the fiscal year was 549.1 million dollars, with 268 million dollars coming from QNX and 258.9 million dollars from secure services. The operating cash flow, which is the cash generated from operations, was 50.3 million dollars, and the operating cash flow for the fourth quarter was 45.6 million dollars, which shows a pattern of fluctuation during different times of the year. There were also 38.1 million dollars from deferred Cylance proceeds, which is money received from a previous deal, and 60 million dollars spent on buying back shares, which affects the company's cash but is not part of the operating cash flow. The backlog, which is the amount of orders waiting to be fulfilled, is 950 million dollars, not 133.9 million dollars, because the company uses a different definition of backlog than the recognized purchase orders, or RPO. https://www.accessnewswire.com/newsroom/en/computers-technology-and-internet/blackberry-reports-fourth-quarter-and-full-fiscal-year-2026-resul-1155999

For the first quarter of fiscal year 2027, which ended on May 31 and was reported on September 11, the company's Secure Communications segment grew by 14.1%, driven by a 15.3% increase in SecuSUITE and a 0.8% decrease in UEM. The QNX segment grew by 14.8%, which includes a 7.5% increase in development licenses, a 4.2% increase in royalties, a 1.5% increase in Radar, and a 1.2% increase in services. The company's remaining performance obligations, or RPO, were $133.9 million, of which $121.5 million is expected to be earned within the next 12 months. RPO excludes usage or sales royalties by definition. The company's deferred revenue decreased from $152.6 million to $133.9 million, a decrease of 18.7%. New increases in deferred revenue were $121.9 million, compared to fulfillment of $140.6 million. Note that not all growth is from recurring or royalty cash, as QNX licenses and volume royalties may include minimum commitments, which is a type of guaranteed payment. https://www.sec.gov/Archives/edgar/data/1070235/000107023526000076/bbry-20260531.htm

The time5_15_test model had a return of -25.75520677946721 percent from 2021-01-04 05:00:00+00:00 to 2026-09-10 04:00:00+00:00, with a drawdown of 48.522722725307744 percent and a compound annual growth rate, or CAGR, which is a measure of investment return over time, of -5.11963356527001 percent. The win rate was 52.54901960784314 percent out of 255 samples. The average trade percentage was -0.07160477672451342 percent, with an average win of 2.059926961130461 percent and an average loss of 2.4321440566630796 percent. The mean confidence interval, or CI95, which is a statistical measure of uncertainty, was between -0.42498062146571663 and 0.2751204592455909 percent. The half Kelly criterion, a formula used to determine the optimal bet size, was 0. The reasons for the results were based on time, with 255 reasons. There were 0 partials and 0 same day trades. The vbt, or vector-based trading, was reconciled, with 510 vbt orders and 68 trade months. The market had a beta, which measures volatility, of 0.7739429677605836, an alpha, which measures investment return, of -14.70754441601599 percent per year, and a correlation, which measures the relationship between two variables, of 0.6059863568739847. This is a daily proxy report, with all 20 cells reconciled, based on training data from 2016 to 2020 and testing from 2021 to September 10, 2026, with no globally untouched claim found in the inspected history. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

The issuer's report for the quarter ending May 31, retrieved on September 11, shows the following financial information in millions of US dollars: revenue was 152.9, QNX was 72.3, secure was 73.6, annual recurring revenue was 220, and the dollar-based net retention rate was 92%. Operating cash flow was 4.6, capital expenditures were 2.9, and intangible investments were 1.6. The company had 422.9 in cash and investments, which includes 14.2 in restricted cash, 196.8 in notes, and 586.061 million shares. The company also repurchased 10 million shares. For the second quarter, the revenue guidance is between 137 and 148, and operating cash flow guidance is between 0 and 10. For the full year, revenue guidance is between 594 and 621, adjusted EBITDA guidance is between 119 and 139, and operating cash flow guidance is approximately 100. The GAAP net income was 8.5, compared to an adjusted net income of 25.4. The calculated free cash flow after investments is 1. The remaining operating cash flow for the next three quarters is 95.4, which is approximately 31.8 per quarter, although this is not a forecast. As of September 10, the stock price was 7.54, which implies a capitalization of 4418.89994 million. The enterprise value, excluding restricted cash, is 4206.99994 million, which is 6.9251 times the guided sales midpoint, 32.6124 times the adjusted EBITDA midpoint, and 44.189 times the operating cash flow guidance. The valuation is based on the number of shares and cash as of May 31, which has not been reconciled with the current balance. https://www.sec.gov/Archives/edgar/data/1070235/000107023526000071/q1fy27ex-991.htm

The automated trading fund manager reported on the time5_15_train model, which is a type of investment strategy that uses a specific time period, in this case, a 5-minute interval, and a 15 basis point slippage, which is a measure of the difference between the expected price of a trade and the actual price. The model was trained on data from January 4, 2016, to December 31, 2020. The results showed a return of -35.39%, a drawdown of 55.13%, and a compound annual growth rate, or CAGR, of -8.37%. The win rate was 49.55%, with 220 samples, or data points. The average trade percentage was -0.16%, with an average win percentage of 1.73% and an average loss percentage of 2.02%. The mean confidence interval, or CI95, percentage was between -0.46% and 0.14%. The half Kelly criterion, a measure of the optimal bet size, was 0. The reasons for the results were based on 220 time periods. There were 0 partials, which are incomplete trades, and 0 same-day trades. The results were reconciled with the vbt, or vector-based trading, system, which is a method of tracking trades, and showed 440 vbt orders. The trade months, or the number of months the model was trading, was 59. The market analysis showed a beta, or measure of volatility, of 0.74, an alpha, or measure of excess return, of -18.24% per year, and a correlation, or measure of how closely the model follows the market, of 0.63. This is a daily proxy report, and all 20 cells, or data points, were reconciled. The training period was from 2016 to 2020, and the test period was from 2021 to September 10. The history was inspected, and there were no globally untouched claims, which means that all data points were used in the analysis. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

The time5_5_test model had a return of 23.63926873150728 percent from 2021-01-04 05:00:00+00:00 to 2026-09-10 04:00:00+00:00. The drawdown was 29.757337397159766 percent and the compound annual growth rate, or CAGR, was 3.815667103450271 percent. The win rate was 56.86274509803921 percent. There were 255 samples, with a mean trade percentage of 0.12845222076228352 percent. The average win percentage was 2.099091141185955 percent and the average loss percentage was 2.469208174341647 percent. The mean confidence interval, or CI95, percentage was between -0.2173690604475051 percent and 0.48858693110590634 percent. The half Kelly criterion, a measure of bet size, was 0.030597104204277137. The reasons for the results were related to time, with 255 reasons. There were 510 orders and the trade months were 68. The market had a beta, a measure of volatility, of 0.7743369469189808, an alpha, a measure of return, of -5.720513836068003 percent per year, and a correlation, or corr, of 0.6070915449237996. This is a daily proxy report, with all 20 cells reconciled, based on training data from 2016 to 2020 and testing data from 2021 to September 10, 2026. The history was inspected and no globally untouched claim was found. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

The automated trading fund manager's report for REV9110400 frozen cell time5_5_train shows the following results: the model used was time5, the period was train, and the slippage in basis points was 5. The first trade was on 2016-01-04 05:00:00+00:00 and the last trade was on 2020-12-31 05:00:00+00:00. The return was 0.32259036108561734, the drawdown was 36.756969045779655, and the Compound Annual Growth Rate, or CAGR, which is a measure of investment return, was 0.06448617734937656. The win rate, or percentage of winning trades, was 54.54545454545454. There were 220 samples, or trades, with a mean trade percentage of 0.038392820430310864. The average win percentage was 1.7666844670028734 and the average loss percentage was 2.035557155456764. The mean Confidence Interval 95 percentage, which is a measure of statistical accuracy, was between -0.26470964170669264 and 0.3262461573418879. The half Kelly criterion, a betting strategy, was 0.01086578309465841. The reasons for the results were based on time, with 220 reasons. There were 0 partials and 0 same day trades. The vbt orders, or trades based on a specific strategy, were 440 and the trade months were 59. The market had a beta, which measures volatility, of 0.7427941589350198, an alpha annual percentage, which measures investment return, of -9.429861337697398, and an alpha t HAC5, a statistical measure, of -1.4958405207662782. The correlation, or relationship, between the trades was 0.6351005466617948. This is a daily proxy report, with all 20 cells reconciled, meaning the data has been checked for accuracy. The training period was from 2016 to 2020 and the test period was from 2021 to September 10. The history was inspected and there were no globally untouched claims, meaning all data was accounted for. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json

The primary5_15_test model had a return of -14.051637727132416 percent from 2021-01-04 05:00:00+00:00 to 2026-09-10 04:00:00+00:00. This model, which is a type of automated trading system, had a drawdown of 39.20026527589924 percent and a compound annual growth rate, or CAGR, of -2.636792748548722 percent. The win rate was 44.31372549019608 percent, with a sample size of 255 trades. The average trade resulted in a loss of 0.027524328075738277 percent, with average wins of 2.3031048344983 percent and average losses of 1.8821799292790222 percent. The reasons for trades included gap target, stop, time, gap stop, and target, with 1, 106, 123, 17, and 8 occurrences respectively. There were 35 partial trades and 30 same-day trades. The model was reconciled with vbt, or vector backtest, orders, which is a method of testing trading strategies, and had 545 vbt orders over 68 trade months. The market had a beta, which measures volatility, of 0.45377235004133804, an alpha, which measures performance, of -8.007891626172409 percent per year, and a correlation, which measures how closely the model follows the market, of 0.43142037593091437. This is a daily proxy report, meaning it is a substitute for daily data, and all 20 cells were reconciled. The model was trained on data from 2016 to 2020 and tested on data from 2021 to September 10, 2026, and the history was inspected, with no globally untouched claim found. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json