Notes

A news scan found the headline 'CENX R91806 backtests 3/3'. Methods note3577 is a set of rules for testing, used for a September 2018 calculation on completed September 2017 bars. The results are given in percent units and the simulations are conditional, using a Kelly diagnostic, which is a tool to measure risk. The results are as follows: the columns are split, lookback, filter, cost in basis points per side, bars, return percent, closed return percent, drawdown percent, win rate percent, number of trades, mean trade percent, median trade percent, open, buy and hold percent, targets, and gaps. The rows show three different scenarios for forward2026 with different filters: none, sma200, which is a simple moving average over 200 days, and spy200, which is a filter based on the SPY stock over 200 days. The numbers are: for no filter, 20 lookback, 6 cost, 178 bars, -27.83 percent return, -27.83 percent closed return, 45.34 percent drawdown, 20 percent win rate, 5 trades, -5.35 percent mean trade, -11.66 percent median trade, 0 open, 1.98 percent buy and hold, 1 target, 1 gap; for sma200 filter, 20 lookback, 6 cost, 178 bars, -28.41 percent return, -28.41 percent closed return, 45.79 percent drawdown, 20 percent win rate, 5 trades, -5.49 percent mean trade, -11.66 percent median trade, 0 open, 1.98 percent buy and hold, 1 target, 1 gap; for spy200 filter, 20 lookback, 6 cost, 178 bars, -34.17 percent return, -34.17 percent closed return, 50.15 percent drawdown, 16.67 percent win rate, 6 trades, -5.94 percent mean trade, -11.23 percent median trade, 0 open, 1.98 percent buy and hold, 1 target, 0 gaps. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CENX&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'Methods note3577'. This is about a calculation done in September 2018 on completed bars from September 2017. The results are given in percent units and the simulations are conditional, which means they depend on certain conditions. A Kelly diagnostic was also used, which is a way to measure the performance of an investment strategy. The results are given in a table with the following columns: split, lookback, filter, cost_bps_side, bars, return_pct, closed_return_pct, drawdown_pct, win_rate_pct, n, mean_trade_pct, median_trade_pct, open, buyhold_pct, targets, gaps. The rows of the table show the results of different simulations, including the holdout method and the forward2026 method. For example, one row shows that the holdout method with a lookback of 20 and a filter of sma50 had a return of 51.20150056 percent and a drawdown of 51.92307979 percent. Another row shows that the forward2026 method with a lookback of 10 and a filter of sma50 had a return of -6.7405399 percent and a drawdown of 34.79397751 percent. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CENX&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'CENX R91806 backtests 1/3'. Methods note3577 is a calculation method used for this backtest, which was done in September 2018 on completed data from September 2017. The calculation is based on percent units and includes simulations that are conditional, with a Kelly diagnostic, which is a measure used to determine the optimal bet size. The results are shown in a table with the following columns: split, lookback, filter, cost_bps_side, bars, return_pct, closed_return_pct, drawdown_pct, win_rate_pct, n, mean_trade_pct, median_trade_pct, open, buyhold_pct, targets, gaps. The rows of the table show the results for different combinations of these parameters, including train and holdout sets, with lookback periods of 10, 20, and 40, and filters such as sma50, none, sma200, and spy200. For example, the first row shows a train set with a lookback period of 10, a filter of sma50, a cost_bps_side of 6, and 2012 bars, with a return_pct of 91.60207924 and a closed_return_pct of 102.45573162. The results also include a holdout set with a lookback period of 10, a filter of sma50, a cost_bps_side of 6, and 502 bars, with a return_pct of 89.45243595 and a closed_return_pct of 44.47279863. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CENX&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written. It is about calculations done in September 2018 using data from September 2017. The calculations are in percent units and the simulations are conditional, using a Kelly diagnostic, which is a tool to measure risk. The results are shown in a table with the following columns: split, regime, above, number of times, mean percent, and win percent. The rows in the table are: train own200 with 26 times above, mean percent 3.74570402, and win percent 34.61538462, train own200 with 16 times not above, mean percent 3.89943614, and win percent 31.25, train spy200 with 33 times above, mean percent 3.45805072, and win percent 33.33333333, train spy200 with 9 times not above, mean percent 5.07373434, and win percent 33.33333333, holdout own200 with 12 times above, mean percent 4.56499205, and win percent 41.66666667, holdout own200 with 1 time not above, mean percent -10.41764264, and win percent 0, holdout spy200 with 13 times above, mean percent 3.41248169, and win percent 38.46153846, forward2026 own200 with 5 times above, mean percent -5.49377278, and win percent 20, forward2026 spy200 with 4 times above, mean percent -3.13998352, and win percent 25, and forward2026 spy200 with 1 time not above, mean percent -14.90892983, and win percent 0. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CENX&timeframe=1Day&feed=sip&adjustment=all

A news scan found the trade statistics for CENX. The statistics were calculated on September 18, using data from completed trades in September 2017. The calculations are based on percent units and include simulations that are conditional, with a Kelly diagnostic, which is a measure of risk. The statistics are presented in a table with the following columns: split, n, mean percent, mean confidence interval 95 percent, win percent, average win percent, average loss percent, half Kelly stock fraction, half Kelly dollars, minimum trade percent, and maximum trade percent. The rows of the table are divided into three categories: train, holdout, and forward 2026. The train category has 42 data points, with a mean percent of 3.80426864, a win percent of 33.33333333, and an average win percent of 32.92887047. The holdout category has 13 data points, with a mean percent of 3.41248169, a win percent of 38.46153846, and an average win percent of 25.33071584. The forward 2026 category has 5 data points, with a mean percent of -5.49377278, a win percent of 20, and an average win percent of 23.18306881. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CENX&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was made. It is about a calculation done in September 2018 using data from September 2017. The calculation is shown in percent units and includes simulations that are conditional, with a Kelly diagnostic, which is a tool used to measure performance. The data is presented in a table with the following information: the type of data set, the number of items, and various percentages, including minimum gap percentage, and percentages of closing prices. The table has three sections: train, holdout, and forward2026, with the following values: for train, 2011 items, minimum gap percentage of -14.4379845, and minimum and maximum closing percentages of -20.37037037 and 21.1409396; for holdout, 502 items, minimum gap percentage of -7.84946237, and minimum and maximum closing percentages of -14.99490835 and 21.30406714; and for forward2026, 178 items, minimum gap percentage of -8.99253163, and minimum and maximum closing percentages of -10.85144631 and 11.79859185. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CENX&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written for the stock CENX, with a specific calculation done in September 2018, using data from completed bars in September 2017. The calculations are shown in percent units and the simulations are conditional, using a Kelly diagnostic, which is a tool to measure the performance of a trading strategy. The data for the stock CENX on September 17, 2026, is as follows: the date is September 17, 2026, the number of days is 2692, the close price is 40.75, the prior 20-day average is 47.95, the 50-day simple moving average, or sma50, is 45.3155, the 200-day simple moving average, or sma200, is 50.11945, the 20-day average true range, or atr20, is 2.175605, the average true range percent, or atrpct, is 5.33890798, the 20-day realized volatility, or rv20, is 0.44464254, and the 20-day average volume, or volume20, is 1570131.4. The signal is false, the rule stop is 36.39879, the rule target is 53.80363, and the shares capital, or shares_cap, is 2. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=CENX&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written in September 2018, based on calculations from September 2017 data. The calculations are shown in percent units and are based on simulations that are conditional, using a Kelly diagnostic, which is a tool to measure risk. The results are shown in a table with the following columns: tail replacement probability, equity weight, number of paths, number of trades, mean terminal percent, fifth percentile terminal percent, ninety-fifth percentile drawdown percent, and probability of a 10 percent loss. The table has the following rows: [0, 0.11999953, 20000, 20, 8.19440572, -6.85528474, 11.20700063, 1.9], [0, 0.02999988, 20000, 20, 1.99079343, -1.71123831, 2.90464533, 0], [0.01, 0.11999953, 20000, 20, 7.2541789, -8.02767907, 12.12620312, 2.705], [0.01, 0.02999988, 20000, 20, 1.76835234, -2.01861219, 3.15440486, 0], [0.05, 0.11999953, 20000, 20, 3.9100826, -11.49047706, 15.08602542, 7.18], [0.05, 0.02999988, 20000, 20, 0.96327046, -2.93831468, 3.96551652, 0]. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A news scan found the simulation results for AA R91806. The results are based on a calculation from September 2018 using data from September 2017. The results are given in percent units and the simulations are conditional, using a Kelly diagnostic, which is a method to determine the optimal size of a bet. The results are as follows: the simulation was run on two different time periods, pre2024 and last252, with different block sizes, 5 and 10. The levels used were 2ATR6ATR, the number of paths was 20000, and the horizon was 60. The stop and target prices were 42.85013 and 59.32961 respectively. The results also include the stop first percentage, target first percentage, timeout percentage, mean percentage, median percentage, and p05 and p01 percentages. Additionally, the results include the mean number of sessions and sizing information, which is a way to determine the amount of money to invest. The sizing information includes the dollars invested and the probability of account loss greater than 3 percent. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'AA R91806 backtests 3/3'. Methods note3577 is a note about methods used, from September 2018, calculating on completed September 2017 bars. Percent units are explicitly named, simulations are conditional, and a Kelly diagnostic is used, which is a measure of risk. The results are: forward2026 with 20 lookback, no filter, 6 cost basis points per side, 178 bars, -12.14531608 percent return, -12.14531608 percent closed return, 27.99662872 percent drawdown, 25 percent win rate, 4 trades, -2.59943922 percent mean trade, -7.68590258 percent median trade, 0 open, -12.76002972 percent buy and hold return, 0 targets, 0 gaps. Another result is forward2026 with 20 lookback, sma200 filter, 6 cost basis points per side, 178 bars, -4.7618966 percent return, -4.7618966 percent closed return, 21.94537373 percent drawdown, 33.33333333 percent win rate, 3 trades, -0.88172236 percent mean trade, -7.61921536 percent median trade, 0 open, -12.76002972 percent buy and hold return, 0 targets, 0 gaps. The last result is forward2026 with 20 lookback, spy200 filter, 6 cost basis points per side, 178 bars, -13.14868649 percent return, -13.14868649 percent closed return, 26.22099215 percent drawdown, 25 percent win rate, 4 trades, -2.85226181 percent mean trade, -7.68590258 percent median trade, 0 open, -12.76002972 percent buy and hold return, 0 targets, 1 gap. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'AA R91806 backtests 2/3'. Methods note3577 is a calculation method used for this backtest. The calculation was done in September 2018, using data from completed bars in September 2017. The results are given in percent units. Simulations are conditional, which means they depend on certain conditions, and a Kelly diagnostic, a tool used to determine the optimal size of a bet, was used. The results are shown in a table with the following columns: split, lookback, filter, cost_bps_side, bars, return_pct, closed_return_pct, drawdown_pct, win_rate_pct, n, mean_trade_pct, median_trade_pct, open, buyhold_pct, targets, gaps. The table has 12 rows of data, each representing a different set of parameters for the backtest. The data includes the results of the backtest for different lookback periods, filters, and cost_bps_side values. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'AA R91806 backtests 1/3'. Methods note3577 is a calculation method used in this context. A calculation was done on September 18, using completed data from September 17, which is referred to as Sep17 and Sep18. The results are given in percent units. Simulations were done conditionally, which means they were done based on certain conditions, and a Kelly diagnostic was used, which is a tool to measure the performance of an investment strategy. The results are given in a table format with the following columns: split, lookback, filter, cost_bps_side, bars, return_pct, closed_return_pct, drawdown_pct, win_rate_pct, n, mean_trade_pct, median_trade_pct, open, buyhold_pct, targets, gaps. The rows of the table are: train with 10 lookback and sma50 filter, train with 10 lookback and sma50 filter, train with 20 lookback and sma50 filter, train with 20 lookback and sma50 filter, train with 40 lookback and sma50 filter, train with 40 lookback and sma50 filter, train with 20 lookback and no filter, train with 20 lookback and sma200 filter, train with 20 lookback and spy200 filter, holdout with 10 lookback and sma50 filter, holdout with 10 lookback and sma50 filter, holdout with 20 lookback and sma50 filter. The specific values for each row are given in the table. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written. It described a calculation done in September 2018 using data from September 2017. The calculation used something called percent units, which are units of measurement that show a value as a percentage of something else. The simulations, which are like practice runs, were conditional, meaning they depended on certain conditions being met. A Kelly diagnostic, which is a tool used to figure out the best size of a bet, was also used. The results of the calculation are shown in a table with the following columns: split, regime, above, n, mean percentage change, and win percentage change. The rows in the table show the results for different groups, including train, holdout, and forward2026, and for two regimes, own200 and spy200. The results are as follows: for the train group and own200 regime where the value was above a certain point, there were 19 instances with a mean percentage change of 7.71717392 and a win percentage change of 47.36842105, for the train group and own200 regime where the value was not above a certain point, there were 18 instances with a mean percentage change of -1.38336122 and a win percentage change of 22.22222222, for the train group and spy200 regime where the value was above a certain point, there were 33 instances with a mean percentage change of 3.82778821 and a win percentage change of 36.36363636, for the train group and spy200 regime where the value was not above a certain point, there were 4 instances with a mean percentage change of -1.14780212 and a win percentage change of 25, for the holdout group and own200 regime where the value was above a certain point, there were 6 instances with a mean percentage change of 8.68023553 and a win percentage change of 50, for the holdout group and own200 regime where the value was not above a certain point, there were 5 instances with a mean percentage change of -2.35329242 and a win percentage change of 20, for the holdout group and spy200 regime where the value was above a certain point, there were 11 instances with a mean percentage change of 3.66499556 and a win percentage change of 36.36363636, for the forward2026 group and own200 regime where the value was above a certain point, there were 3 instances with a mean percentage change of -0.88172236 and a win percentage change of 33.33333333, for the forward2026 group and spy200 regime where the value was above a certain point, there were 2 instances with a mean percentage change of 4.40334061 and a win percentage change of 50, and for the forward2026 group and spy200 regime where the value was not above a certain point, there was 1 instance with a mean percentage change of -11.4518483 and a win percentage change of 0. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written in September 2018, based on calculations from September 2017 data. The calculations are shown in percent units and the simulations are conditional, using a Kelly diagnostic, which is a tool to measure risk. The data is presented in a table with the following columns: split, n, mean percent, mean confidence interval 95 percent, win percent, average win percent, average loss percent, half Kelly stock fraction, half Kelly dollars, minimum trade percent, and maximum trade percent. The table has three rows: train, holdout, and forward 2026, with the following values: train has 37 data points, a mean percent of 3.29, a win percent of 35.14, an average win percent of 25.90, an average loss percent of 8.96, a half Kelly stock fraction of 0.5, half Kelly dollars of 101.99, a minimum trade percent of -14.35, and a maximum trade percent of 40.03. Holdout has 11 data points, a mean percent of 3.66, a win percent of 36.36, an average win percent of 24.13, an average loss percent of 8.03, a half Kelly stock fraction of 0.5, half Kelly dollars of 101.99, a minimum trade percent of -9.58, and a maximum trade percent of 26.83. Forward 2026 has 3 data points, a mean percent of -0.88, a win percent of 33.33, an average win percent of 16.43, an average loss percent of 9.54, a half Kelly stock fraction of 0, half Kelly dollars of 0, a minimum trade percent of -11.45, and a maximum trade percent of 16.43. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'AA R91806 tails 1/1'. This is a data line. Methods note3577 is about a calculation done in September 2018 on completed data from September 2017 bars. The calculation uses percent units and simulations that are conditional, with a Kelly diagnostic, which is a tool to measure risk. The data has several columns, including split, n, min_gap_pct, p01_gap_pct, p001_gap_pct, below10, below20, below30, min_close_pct, and max_close_pct. The rows of data are divided into three categories: train, holdout, and forward2026, with the following values: train has 1801, -13.24651408, -5.39021852, -12.04425573, 4, 0, 0, -21.05162524, and 28.2108389; holdout has 502, -6.44948996, -4.70822703, -5.86236677, 0, 0, 0, -11.38532412, and 14.56834532; forward2026 has 178, -7.61117446, -6.42767366, -7.46427783, 0, 0, 0, -9.47834509, and 8.65777778. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written. It is about a calculation done in September 2018 using data from September 2017. The calculation uses percentages and is based on certain conditions, including a diagnostic test called Kelly. A simulation was run using the following data: date, number of days, closing price, average price for the prior 20 days, simple moving average for 50 and 200 days, average true range for 20 days, average true range as a percentage, realized volatility for 20 days, average volume for 20 days, a signal to buy or sell, a stop loss rule, a target price rule, and the maximum number of shares to buy. The data from September 17, 2026, is as follows: the closing price was 46.97, the average price for the prior 20 days was 51.85, the simple moving average for 50 days was 48.346, the simple moving average for 200 days was 57.8794, the average true range for 20 days was 2.059935, the average true range as a percentage was 4.38563977, the realized volatility for 20 days was 0.40473797, the average volume for 20 days was 3627137.25, the signal was false, the stop loss price was 42.85013, the target price was 59.32961, and the maximum number of shares to buy was 2. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AA&timeframe=1Day&feed=sip&adjustment=all

A news scan found the simulation note: Methods note3577 used a September 2018 calculation on completed September 2017 bars. The simulation used percent units and was conditional, with a Kelly diagnostic. The simulation results were as follows: the columns included period, block, levels, paths, horizon, stop, target, stop first percentage, target first percentage, timeout percentage, mean percentage, median percentage, fifth percentile, first percentile, mean sessions, and sizing. The rows of data were for the last 252 periods, with a block size of 5 and 10, levels of 2ATR6ATR, 20000 paths, and a horizon of 60. The stop was 11.60766 and the target was 21.69702. The stop first percentage was 50.49 and 51.82, the target first percentage was 43.405 and 42.86, the timeout percentage was 6.105 and 5.32, the mean percentage was 14.63000731 and 13.98798011, the median percentage was -18.11341239, the fifth percentile was -19.7320444 and -19.58416788, and the first percentile was -23.23799989. The mean sessions were 22.55105 and 21.4061. The sizing data included dollars and account loss percentages, with probabilities of account loss greater than 3 percent. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=ALM&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'ALM R91806 backtests 2/2'. Methods note3577 is a set of rules for the automated trading fund manager. The calculation was done in September 2018 using data from completed September 2017 bars. The results are shown in percent units and the simulations are conditional, using a Kelly diagnostic, which is a measure of risk. The results are as follows: the columns are split, lookback, filter, cost in basis points per side, number of bars, return percent, closed return percent, drawdown percent, win rate percent, number of trades, mean trade percent, median trade percent, open, buy and hold percent, targets, and gaps. The rows show the results for different scenarios, including forward2026 with different lookback periods, filters, and costs. For example, one scenario used a 20-day lookback period, an sma50 filter, and a cost of 16 basis points per side, resulting in a return of -8.58735446 percent and a drawdown of 51.04614685 percent. The results for all scenarios are: forward2026 with 20-day lookback, sma50 filter, and 16 basis points cost per side, return -8.58735446 percent, drawdown 51.04614685 percent; forward2026 with 40-day lookback, sma50 filter, and 6 basis points cost per side, return -4.98736678 percent, drawdown 49.473138 percent; forward2026 with 40-day lookback, sma50 filter, and 16 basis points cost per side, return -6.12070598 percent, drawdown 49.72519232 percent; forward2026 with 20-day lookback, no filter, and 6 basis points cost per side, return -6.3143824 percent, drawdown 50.17883295 percent; forward2026 with 20-day lookback, sma200 filter, and 6 basis points cost per side, return -14.44223712 percent, drawdown 27.46509156 percent; forward2026 with 20-day lookback, spy200 filter, and 6 basis points cost per side, return -6.3143824 percent, drawdown 50.17883295 percent. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=ALM&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written in September 2018, using data from completed bars in September 2017. The calculations are in percent units and the simulations are conditional, with a Kelly diagnostic, which is a measure of risk. The results are shown in a table with the following columns: split, lookback, filter, cost in basis points per side, number of bars, return percent, closed return percent, drawdown percent, win rate percent, number of trades, mean trade percent, median trade percent, open, buy and hold percent, targets, and gaps. The rows show the results for different combinations of these factors, including holdout and forward 2026, with various lookback periods and filters, such as sma50, none, and spy200. For example, one row shows a holdout with a 10-day lookback, sma50 filter, and 6 basis points cost per side, resulting in a return of 84.67 percent and a drawdown of 17.37 percent. Another row shows a forward 2026 with a 10-day lookback, sma50 filter, and 6 basis points cost per side, resulting in a return of 1.07 percent and a drawdown of 46.25 percent. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=ALM&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written. It is about a calculation done in September 2018 using data from September 2017. The calculation is measured in percent units and the results are explicitly named. The simulations are conditional, which means they depend on certain conditions, and a Kelly diagnostic was used, which is a tool to measure the performance of a trading strategy. The results are shown in a table with the following information: the type of data split, the regime, whether the result is above a certain threshold, the number of times the condition was met, the average percent change, and the percent of times the result was a win. The table shows the following results: for the holdout data and the own200 regime, when the result is not above the threshold, the average percent change was 17.02461883 and the win percent was 66.66666667, for the holdout data and the spy200 regime, when the result is above the threshold, the average percent change was 17.02461883 and the win percent was 66.66666667, for the forward2026 data and the own200 regime, when the result is above the threshold, the average percent change was -15.51019388 and the win percent was 0, for the forward2026 data and the own200 regime, when the result is not above the threshold, the average percent change was 4.85169414 and the win percent was 40, and for the forward2026 data and the spy200 regime, when the result is above the threshold, the average percent change was 1.45804614 and the win percent was 33.33333333. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=ALM&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'ALM R91806 trade_stats 1/1'. Methods note3577 is about a calculation done in September 2018 on completed data from September 2017 bars. The calculation uses percent units and simulations that are conditional, with a Kelly diagnostic, which is a tool to measure risk. The results are shown in a table with the following columns: split, n, mean_pct, mean_CI95_pct, win_pct, avgwin_pct, avgloss_pct, halfKelly_stock_fraction, halfKelly_dollars, min_trade_pct, and max_trade_pct. The table has two rows: one for holdout and one for forward2026. The holdout row shows 3, 17.02461883, [-10.05909858,31.2774184], 66.66666667, 30.56647754, 10.05909858, 0.5, 101.99, -10.05909858, and 31.2774184. The forward2026 row shows 6, 1.45804614, [-16.4636408,20.9267719], 33.33333333, 36.18036961, 15.9031156, 0.1245, 101.99, -19.00350414, and 40.59248545. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=ALM&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written. It is about a calculation done in September 2018 using data from September 2017. The calculation is measured in percent units and the results are explicitly named. The simulations are conditional, which means they depend on certain conditions, and a Kelly diagnostic was used, which is a tool to measure the performance of a trading strategy. The results are shown in a table with the following columns: split, n, min_gap_pct, p01_gap_pct, p001_gap_pct, below10, below20, below30, min_close_pct, and max_close_pct. The table has two rows: one for the holdout period and one for the forward period until 2026. The holdout period has 119 data points, with a minimum gap percentage of -13.41772152, and the forward period has 178 data points, with a minimum gap percentage of -14.07156673. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=ALM&timeframe=1Day&feed=sip&adjustment=all

A methods note, number 3577, was written. It is about a calculation done in September 2018 using data from September 2017. The calculation uses percentages and is based on certain conditions, including a diagnostic called Kelly. A news scan found the following data: a table with information about a stock, including its date, closing price, and other technical indicators. The table has columns for date, number of days, closing price, and other indicators, and one row of data for September 17, 2026, with a closing price of 14.13, and other values, including a signal that is false, a stop rule of 11.60766, and a target rule of 21.69702, with 7 shares. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=ALM&timeframe=1Day&feed=sip&adjustment=all

A news scan found the simulation results for PSIX R91806. The results are based on a September 18 calculation using completed September 17 bars. The results are in percent units and the simulations are conditional, using a Kelly diagnostic, which is a measure of risk. The results are as follows: the columns are period, block, levels, paths, horizon, stop, target, stop first percentage, target first percentage, timeout percentage, mean percentage, median percentage, 5th percentile percentage, 1st percentile percentage, mean sessions, and sizing. The rows are last 252 days, with a block size of 5 and 10, using 2ATR6ATR levels, 20000 paths, and a 60 day horizon. The stop is 41.02339 and the target is 59.12983. The stop first percentage is 73.045 and 74.57, and the target first percentage is 26.72 and 25.19. The timeout percentage is 0.235 and 0.24. The mean percentage is -0.8900247 and -1.82886788, the median percentage is -10.22541191, the 5th percentile percentage is -23.07726015 and -30.22118816, and the 1st percentile percentage is -34.48402394 and -38.81617366. The mean sessions are 10.6661 and 10.4411. The sizing results are also provided, with dollars and probabilities of account loss. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=PSIX&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'PSIX R91806 backtests 2/2'. Methods note3577 is a set of rules for the automated trading fund manager. The Sep18 calculation was done on completed Sep17 bars, which are groups of data from September 2017. The results are shown in percent units, which are units of measurement that show a value as a percentage. The simulations, or pretend trades, were conditional, meaning they were based on certain conditions, and included a Kelly diagnostic, which is a tool used to determine the best size for a trade. The results are as follows: a table with columns for split, lookback, filter, cost in basis points per side, number of bars, return as a percentage, closed return as a percentage, drawdown as a percentage, win rate as a percentage, number of trades, mean trade return as a percentage, median trade return as a percentage, whether the trade is open, buy and hold return as a percentage, targets, and gaps. The rows of the table show the results of different simulations, including forward2026 with different parameters, such as lookback of 20 or 40, filter of sma50 or none, and cost of 6 or 16 basis points per side. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=PSIX&timeframe=1Day&feed=sip&adjustment=all