Stats

A frozen protocol note was made before checking prices. From January 2023 to September 4, 2026, the same patterns were seen as with other rules, so this is not a completely independent validation. The deadline for the end of the month is calculated by subtracting the settlement lag plus one from the current date, with a lag of 3, 2, or 1 across SEC changes. The next open is bought only if the return from the preceding 5 sessions is less than 0, with a maximum of 3 open-to-open sessions, and this is for research purposes only, not actual fund allocation. The results of holding a position are compared to stopping at a target price, which is 1.5 times the average true range over 20 days, and stopping at a loss, which is also based on the average true range over 20 days. The daily stop is triggered when both conditions are met, with a stop gap at the open, and the target gap is capped at the target price. Same-day entries and exits are allowed through separate synthetic positions. Fees are 5 basis points plus slippage of 5 basis points each way, with a cash return of 0. The SPY is the primary US proxy, and diagnostics for EFA and EEM do not establish underlying local flows. A total of 8052 adjusted bars were used, with 2684 each, and no dates are missing or extra, with OHLC and volume checks passing. The results from vectorbt reconcile to an independent compounded cash ledger for every 36 experiments, with errors less than 1e-6 on an initial capital of 10000. The drawdown is calculated from the open and close, not the intraday maximum. The fixed return is -2.359834%, with a drawdown of 9.357740%, and other metrics are also reported. Range figures are also provided, with a 3-calendar-month block and 10000 bootstrap mean 95% confidence interval of [-0.5488546052653037, 0.6886803891309617]%. All trial metrics and uncertainty are separately observed, and the code, raw inputs, and full trade and equity ledgers are local artifacts. The conclusions are about this newly specified proxy, not an exact reproduction of the institutional ANcerno study. Mandatory 15-minute and extended hours and trailing profit management, raw cash dividends, bank settlement holiday differences, and whole-share sizing are not validated, and there is no execution authorization. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars, https://aaltodoc.aalto.fi/server/api/core/bitstreams/0f01d3c2-3b16-44f9-9135-3f03cdd29334/content, and https://www.sec.gov/exams/educationhelpguidesfaqs/t1-faq.

A frozen protocol note, numbered 213, was written before inspecting prices. The training period was from 2016 to 2022. There is a deadline at the end of the month, which is calculated by subtracting the settlement lag plus one from the current date, with lags of 3, 2, and 1 across SEC changes. A buy order is placed at the next open only if the return from the preceding 5 sessions is less than 0, with a maximum of 3 open-to-open sessions. This is for research purposes only, using full capital, and does not reflect actual fund allocation. The strategy compares a fixed hold to a stop using a lagged Average True Range of 20, with a target of 1.5 times the Average True Range of 20. A daily stop is triggered when both conditions are met, with a stop gap at the open, and a favorable target gap capped at the target. Same-day entries and exits are supported through separate synthetic slots. Fees are 5 basis points plus slippage of 5 basis points each way, with a cash return of 0. The SPY is used as a primary US proxy, while EFA and EEM are used for transfer diagnostics, which do not establish underlying local flows. A total of 8052 adjusted SIP bars were used, with 2684 bars each, and no missing or extra dates on the NYSE calendar. The results were reconciled to an independent compounded cash ledger for every 36 experiment terminal equity, with errors less than 1e-6 on an initial capital of 10000. The drawdown is calculated based on open and close prices, not intraday maximum. The fixed return was 3.260950%, with a drawdown of 11.791381%, and other metrics including N31, p54.8387%, W1.23598%, and L1.24099%. Range figures for a 3-calendar-month block had a mean of 95CI [-0.6321942322129009, 0.38874040945891103]%. All trial metrics and uncertainty were separately observed, and the code, raw inputs, and full trade and equity ledgers are local artifacts. The conclusions concern this newly specified proxy, not an exact reproduction of the institutional ANcerno study. Mandatory 15-minute, extended hours, and trailing profit management, raw cash dividends, bank settlement holiday differences, and whole-share sizing were not validated, and no execution authorization was given. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars, https://aaltodoc.aalto.fi/server/api/core/bitstreams/0f01d3c2-3b16-44f9-9135-3f03cdd29334/content, and https://www.sec.gov/exams/educationhelpguidesfaqs/t1-faq.

A frozen protocol note was made before inspecting prices. From January 2023 to September 4, 2026, the trading rules were tested, but this was not a completely independent test because it was already used for other rules. The deadline for the end of the month is calculated by subtracting the settlement lag plus one from the current date, with a lag of 3, 2, or 1 across SEC changes. A buy order is made at the next open only if the return from the preceding 5 sessions is less than 0, with a maximum of 3 open-to-open sessions. This is for research purposes only, with full-capital normalization, but it does not reflect the actual fund allocation. The results are compared to a fixed hold strategy and a stop strategy with a 20-day average true range and a target of 1.5 times the average true range. A stop is made when both conditions are met, with a stop gap at the open and a favorable target gap capped at the target. Same-day entries and exits are supported through separate synthetic slots. Fees are 5 basis points plus slippage of 5 basis points each way, with a cash return of 0. SPY is the primary US proxy, and diagnostics for EFA and EEM do not establish underlying local flows. The test used 8052 adjusted bars, with 2684 each, and no missing or extra dates, and the OHLC and volume checks passed. The results from vectorbt reconcile with the independent compounded cash ledger for every 36 experiment terminal equity, with errors less than 1e-6 on an initial capital of 10000. The drawdown is calculated from the open and close prices, not the intraday maximum. The fixed return is 1.220813%, with a drawdown of 11.784379%, and other metrics are listed below. The range figures for a 3-calendar-month block with 10000 bootstrap mean and 95% confidence interval are [-0.9636297394857071, 0.5704831950615763]%. All trial metrics and uncertainty were observed separately, and the code, raw inputs, and full trade and equity ledgers are local artifacts. The conclusions are about this newly specified proxy and do not exactly reproduce the institutional ANcerno study. Mandatory 15-minute, extended hours, and trailing profit management, raw cash dividends, bank settlement holiday differences, and whole-share sizing were not validated, and there is no execution authorization. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars, https://aaltodoc.aalto.fi/server/api/core/bitstreams/0f01d3c2-3b16-44f9-9135-3f03cdd29334/content, and https://www.sec.gov/exams/educationhelpguidesfaqs/t1-faq.

A frozen protocol note was made before inspecting prices. The training data is from 2016 to 2022. There is a deadline at the end of the month, which is calculated as D minus the settlement lag plus one, with lags of 3, 2, and 1 across SEC changes. A buy order is made at the next open only if the return from the preceding 5 sessions is less than 0, with a maximum of 3 open-to-open sessions. This is for research purposes only, with full-capital normalization, and not for actual fund allocation. The strategy compares a fixed hold to a stop loss with a 1.5 ATR target and a 20-period ATR, with a daily stop when both are hit. Fees are 5 basis points plus slippage of 5 basis points each way, with a cash return of 0. The SPY is the primary US proxy, and diagnostics from EFA and EEM do not establish underlying local flows. The data consists of 8052 adjusted SIP bars, with 2684 bars each, and no missing or extra dates. The results from Vectorbt reconcile to an independent compounded cash ledger for every 36 experiment terminal equity, with errors less than 1e-6 on an initial capital of 10000. The drawdown is calculated from open to close, not from the intraday maximum. The fixed return is 12.367856%, with a drawdown of 11.143295%, and other metrics are as follows: N is 34, p is 52.9412%, W is 1.81719%, and L is 1.27186%. The range figures are below, with a 3-calendar-month block and a 10000 bootstrap mean 95% CI of [-0.6868827262363107, 0.40455541742822726]%. All trial metrics and uncertainty are separately observed, and the code, raw inputs, and full trade and equity ledgers are local artifacts. The conclusions concern this newly specified proxy, not an exact reproduction of the institutional ANcerno study. Mandatory 15-minute, extended hours, and trailing profit management, as well as raw cash dividends, bank settlement holiday differences, and whole-share sizing, are unvalidated and not authorized for execution. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars, https://aaltodoc.aalto.fi/server/api/core/bitstreams/0f01d3c2-3b16-44f9-9135-3f03cdd29334/content, and https://www.sec.gov/exams/educationhelpguidesfaqs/t1-faq.

A study was done to see if buying long options for IBIT, based on a mean-reversion strategy, would be profitable. The study looked at the stock's past performance, including its profitability and costs, to make a decision. A diagnostic test was run using a computer program called CLI, which looked at the stocks F, SOFI, and IBIT, using a mean-reversion strategy over 20 days, with specific entry and exit rules. The results showed that F had a return of 0.79%, SOFI had a return of 7.07%, and IBIT had a loss of 25.80%. However, these results were not used because they did not account for certain costs and execution times. Another test was done using a different method, which looked at the performance of the stocks over a longer period, from January 2023 to September 2026, with different thresholds and costs. The results showed that even with a 2-sigma threshold and a 10bp cost, the return was still negative, at -24.461%. The study concluded that there was not enough evidence to support buying long options for IBIT, due to small sample sizes, lack of historical bid/ask execution data, and insufficient profits in the underlying stock to cover the costs of the options. A strong counterexample was found, where the rebound was too small to recover the spread and theta. The data used for the study came from a source at https://runway.adityaperswal.workers.dev. Based on the results, it was decided not to use this strategy to size options for IBIT.

A study on SOFI stock used a mean-reversion strategy, which is a trading method that assumes prices will return to their average, to see if it would be profitable to buy long options. The study had two main points to consider: first, that making a profit on the stock itself is not enough to guarantee making a profit on options, and second, that the costs of buying and selling must be included in the calculation. The study used a computer program to test the strategy on SOFI, F, and IBIT stocks over a period of time. The results showed that SOFI had a 7.07% profit, but this was not enough to justify buying long options due to small sample sizes and lack of historical data on option prices. The study also found that the rebound in prices was too small to cover the costs of buying and selling options. Therefore, the study rejected using this strategy to decide how many options to buy.

A study on the stock F found that using a mean-reversion strategy, which is a trading method that assumes prices will return to their average, did not provide enough evidence to support buying long options. The study used a computer program to test the strategy on F and two other stocks, SOFI and IBIT, from January 2023 to September 2025. The results showed that F had a return of 0.79%, SOFI had a return of 7.07%, and IBIT had a return of -25.80%. However, these results were not reliable due to the small sample size and lack of data on option prices. Another test was done using a different method, which also did not find any evidence to support buying long options. In fact, the results showed that the potential profits from the strategy were not enough to cover the costs of buying and selling options. Therefore, the study concludes that using this strategy as a proxy to determine the size of options trades is not reliable.

A preregistered note was made before analyzing the stock bars. The analysis used daily adjusted prices of GEV and SPY from 2024 to September 4, 2026, which included 610 bars of GEV data starting from April 2, 2024. The trading strategy was based on a simple moving average crossover, where a 10-day average crossing a 50-day average would trigger an entry or exit, and only long positions were considered. The Vectorbt fees were 5 basis points and slippage was 10 basis points for each side of the trade, with a stress test of 10/20 basis points. In 2024, the training period showed a return of 82.6037% and a drawdown of 9.2479%, but there were no completed training trades to estimate the win/loss ratio. The holdout period starting from 2025 showed a return of 129.1912% and a drawdown of 27.205%, with 5 natural exits, 40% wins, and an average win of 66.1459% and an average loss of 5.5673%. The stress test showed a return of 125.779%. Nine different combinations of fast and slow moving averages were tested, but the fixed 10/50 combination was retained without further tuning. The total return was reconciled within 3e-10 percentage points, and the maximum trade cost difference was 4.04e-8 dollars. A command-line interface control was used to run the strategy with the symbols GEV, a simple moving average cross strategy, a fast period of 10, a slow period of 50, and a period of 730 days. The results showed 501 bars, a return of 134.4719%, a drawdown of 27.7697%, a win rate of 33.333%, and 6 trades, with a buy-and-hold return of 375.2721%. However, despite the positive performance, there was no trading authorization due to the small sample size and lack of natural training payoffs. The data and code used for this analysis are archived and can be found at https://runway.adityaperswal.workers.dev/data/stock-bars, with full inputs, code, and results archived as observations 2958-2997. This is a new diagnostic for GEV and not a revalidation of prior results.

A news scan found the headline 'CLI mean-reversion control cannot validate sector portfolio reversal'. This is a report about a test of a trading strategy called mean-reversion, which is a way of buying and selling based on the idea that prices will return to normal after a big change. The test used a computer program to look at the prices of 9 stocks, specifically XLB, XLE, XLF, XLI, XLK, XLP, XLU, XLV, and XLY, over 730 days. The program looked at the prices of these stocks over 20 days to decide when to buy or sell, using a measure called the z-score, which is a way to compare prices to what is normal. The test found that the strategy did not work well, and the returns were flat when the z-score was above the entry point. The program also found that the prices of the stocks changed near the time of corporate actions, such as stock splits, and that these changes made it hard to understand the total returns. The test used data from the website https://runway.adityaperswal.workers.dev/data/stock-bars and the computer program /app/agent/dist/backtest.js.

The main formation remains frozen for 5 days. Pre-declared training-only observations from 3 to 10 days were recorded from 1773 to 1776 with 892 out of 897 payoffs. These were not selected for holdout. A cost-only attribution note 162 preserves the rule, with 20 total experiment metrics from 1765 to 1784. The one-way fixed holdout is 72.880 or 60.678 percent, with a range of 22.797 to 14.129 percent. The range zero to 2 basis points mean 95 percent confidence interval is -0.0178 to 0.2502 percent or -0.0561 to 0.2102 percent. These optimistic variants improve the headline but not confidence or actual exit eligibility. Costs are assumptions and 50 basis points stress, not asserted venue fees, and tested fees and slippages are symmetric. The gross range average trade is 0.120566 percent, which gives a rough first-order break-even of 6 basis points one way before parameter or estimation risk, but this is not an executable edge estimate. Repeated selected names are liquidated or rebought by the frozen rule, and netting may reduce cost but requires new separate registration, not edited past results. All variants have a vector backtest and independent cash ledger error of less than 1e-6. The source of this data is https://runway.adityaperswal.workers.dev/data/stock-bars, which is a data line.

A test was run using the RequiredCLI command on the GRAL symbol with a breakout strategy. The test used 20 days of lookback data, 900 days of trading data, a 5% stop loss, and a 10% target profit. The results showed 560 trades, a 5.70810% return, a 24.29516% drawdown, a 31.81818% win rate, and an average trade return of 0.290269%. The test assumed no fees or slippage and used exact stop prices without gap loss. The price history included the when-issued period. This test was compared to a custom test 132, but the comparison was not a controlled single-variable experiment because the rule, feed, cost, timing, and exits differed. The positive output from the command line interface cannot validate the panel run-up or rule 158. No strategy was selected from this test. The data used is available at https://runway.adityaperswal.workers.dev/data/stock-bars and the engine used is /app/agent/dist/backtest/engine.js. The return and drawdown percentages are based on 100% abstract capital and do not represent actual fund profit. The outcome of this test is REJECT for entry validation.

A command line interface (CLI) price-only control cannot implement a funding-conditioned rule. The required CLI executed was a node command with specific parameters: node /app/agent/dist/backtest.js --symbols BTC/USD,ETH/USD,SOL/USD --strategy rsi --days 978 --stop-pct 4 --target-pct 8. The built-in CLI exposes price strategies, but does not take into account funding history. This is a price-reversion control, not a replication of a registered funding rule. The engine uses the same bar close entry, with no explicit costs, and includes the current day, with no independent chronological holdout. The exact percentage barriers lack gap fill and 15-minute management. No parameter was chosen from it. Custom vectorbt tests 127-129 implement registered external funding with daily and hourly bars, next open, costs, and fixed split, and their negative verdict controls. The CLI measured summary for each symbol is as follows: for BTC/USD, 979 bars, 2.24% return, 80.04% buy and hold, 26.46% max drawdown, 52.78% win rate, 0.14% average trade, and 36 samples. For ETH/USD, 979 bars, -23.90% return, 4.27% buy and hold, 31.92% max drawdown, 42% win rate, -0.47% average trade, and 50 samples. For SOL/USD, 741 bars, 13.88% return, -35.23% buy and hold, 25.91% max drawdown, 48.94% win rate, 0.39% average trade, and 47 samples. The source of the data is https://runway.adityaperswal.workers.dev/data/crypto-bars and /app/agent/dist/backtest/engine.js. Note that the differences between RSI and funding tests should not be interpreted as isolating one bias.

A frozen note, numbered 161, discusses a trading strategy. The strategy involves nine original sector ETFs, which are investment funds that track a specific sector. The strategy looks at the returns of these ETFs over a five-session period and buys the bottom three performers at the next open, using equal amounts of cash. A stop-loss order, which is an order to sell a security when it falls to a certain price, is set at 2% below the purchase price, and a target price is set at 2% above the purchase price. If the price closes at or above a 1% gain, the strategy sells one-third of the position at the next open and sets a stop-loss order at the breakeven point. This process can happen for a maximum of five sessions. The strategy was tested from January 3, 2023, to September 4, 2026, using data from 2016 as a warm-up period. The test assumed a fee of 0.25% and a slippage of 0.25% for each trade, but these are not actual costs. The test used a long-only strategy, which means it only bought securities, and did not short sell. The results showed a return of -80.2992%, a maximum drawdown of 80.3135%, and other metrics. The test used a specific set of ETFs, not the entire universe of ETFs, and the results may not be applicable to all reversal theories. The test also did not validate the execution of trades at 15-minute intervals or the impact of dividends and whole-share execution. The sources of the data used in the test are https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf.

A frozen note, numbered 161, discusses a strategy for trading sector ETFs. The strategy involves buying the three sector ETFs with the lowest returns over the past five sessions, and selling them when they reach a certain target or stop loss. The strategy uses a freeze signal, which is based on the price of the ETFs, to determine when to buy or sell. The freeze signal has a stop loss of 1 ADR and a target of 2 ADR. The strategy also includes a feature to trim the position by one-third when the return reaches a certain level, and to move the stop loss to breakeven. The strategy was tested on data from January 3, 2023, to September 4, 2026, and the results show a return of -14.8384%, a maximum drawdown of 16.9591%, and other metrics. The test assumed a fee of 5 basis points plus slippage of 5 basis points on each trade. The results are based on a specific implementation of the strategy and do not necessarily apply to all reversal theories. The source of the data is listed as https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf.

A frozen note, numbered 161, describes a trading strategy. The strategy involves nine original sector ETFs, which are investment funds that track a specific sector. The strategy looks at the returns of these ETFs over a five-session period and calculates the mean return for each sector. It then buys the three sectors with the lowest returns, also known as negative residuals, at the next opening, using an equal amount of cash for each sector. The strategy holds these positions for five sessions and then sells them at the next opening. There are no mandatory stops or targets, which means that the strategy does not automatically sell a position if it loses a certain amount of money or reaches a certain profit. The strategy was tested from January 3, 2023, to September 4, 2026, using data from 2016 as a warm-up period. The test assumed a fee of 25 basis points, or 0.25%, plus a slippage cost of 25 basis points, which is the difference between the expected price of a trade and the actual price. The strategy used a technique called vectorbt, which allows for explicit long-only orders and cash sharing. The results showed a return of -72.2625%, a drawdown of 73.0675%, and other metrics. The data used for this test came from two sources: https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf. The conclusion is that this specific implementation of the sector reversal strategy did not work well, and the results should not be used to make decisions about other reversal theories. The actual execution of the strategy, including the timing and dividend payments, was not validated.

A frozen note, numbered 161, discusses a trading strategy. The strategy involves nine original sector ETFs, which are bought and sold based on their performance over five sessions. The goal is to buy the bottom three performing ETFs and hold them for five sessions before selling. This strategy does not use stop-loss orders or target prices, and it is not eligible for certain trading venues. The strategy was tested from January 3, 2023, to September 4, 2026, with a warm-up period starting in 2016. The test assumed a fee of 5 basis points and a slippage cost of 5 basis points for each trade. The vectorbt system was used to execute long-only orders, and a daily ledger was kept to track cash and quantity. The results showed a return of 19.8968%, a maximum drawdown of 18.4248%, and other metrics including 549 trades, a win rate of 54.4627%, a win-to-loss ratio of 1.8538%, and a loss rate of 1.9576%. The data used for this test came from two sources: https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf. This conclusion is specific to this implementation and does not represent all reversal theories. The actual execution of the strategy, including 15-minute intervals, dividend payments, and whole-share trading, was not validated.

A news scan found the note: Frozen note 161. There were nine original sector ETFs. The system looked at the returns of these ETFs over five sessions and compared them to the average return of all sectors. If an ETF's return was less than or equal to the average, the system would buy the three ETFs with the lowest returns at the next open, using equal amounts of cash. This was called a freeze signal. The system also used a stop-loss, which is a rule to sell a security when it falls to a certain price, and a target price, which is a price at which to sell a security for a profit. The stop-loss was set at 1 average daily range, and the target price was set at 2 average daily ranges. If the system made a profit of at least 1 times the risk, it would sell one-third of the position at the next open and set the stop-loss at the break-even price. The system was trained on data from 2017 to 2022, and it used data from 2016 as a warm-up period. The system assumed a fee of 25 basis points, which is a unit of measurement for interest rates and fees, and a slippage of 25 basis points, which is the difference between the expected price of a trade and the actual price. The system used a long-only strategy, which means it only bought securities and did not sell them short. The system also used a cash-sharing strategy, which means it allocated cash to different positions based on their performance. The results showed a return of -92.3758%, a maximum drawdown of 92.5984%, and a number of trades of 896. The system had a win rate of 32.4777% and a loss rate of 2.0536%. The system was tested on a population of nine sector ETFs, which is not the entire universe of ETFs. The system used a synthetic ordering system, which is a way of simulating trades, and reported the results based on daily closing prices. The system excluded the final incomplete rotation, which means it did not include the last trade in the results, and it had zero open positions at the end of the test period. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf. The conclusion is that this is a specific implementation of a sector reversal strategy, and it is not a test of the entire theory. The system did not use any holdout data, which is a portion of the data set aside for testing, and it did not validate the results using actual 15-minute, dividend, or whole-share execution data.

A news scan found the note about a trading strategy. The strategy uses nine sector ETFs and buys the three worst performers every five sessions. It does this by allocating equal amounts of cash to each of the three ETFs at the next opening price. The strategy also includes a stop-loss order, which is an order to sell a security when it falls to a certain price, in this case, 20% below the purchase price. The strategy will close one-third of the position at the next opening price if the price increases by 1R, which is a measure of the strategy's performance, and move the stop-loss to the break-even point. The strategy was tested from 2017 to 2022, with 2016 used as a warm-up period. The test assumed a fee of 5 basis points and a slippage of 5 basis points on each trade. The results showed a return of -16.8486%, a maximum drawdown of 31.2305%, and 896 trades, with 44.7545% of trades being winners and 1.9125% being losers. The strategy was tested on a population of nine current ETFs, not the entire universe of ETFs. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf. The conclusion is that this specific implementation of the strategy was not adopted or tuned from a holdout set. The actual execution of the strategy, including 15-minute, dividend, and whole-share execution, was not validated.

A frozen note, number 161, describes a trading strategy. The strategy involves nine original sector ETFs. It calculates the returns of these ETFs over five sessions and compares them to the average return of all sectors. Every five sessions, it buys the three sector ETFs with the lowest returns, also known as negative residuals, using an equal amount of cash. It holds these ETFs for five sessions and then sells them at the opening price. There are no mandatory stops or targets, so this strategy is not eligible for a specific trading venue. The strategy was trained on data from 2017 to 2022, with 2016 used as a warm-up period. The training dates overlap with other research and are not a completely untouched sample. The strategy assumes a fee of 25 basis points and a slippage of 25 basis points on each side, but these are not actual live cost measurements. The strategy uses explicit long-only orders and cash sharing. It also uses an independent cash and quantity daily ledger, which agrees with the results to a very small margin of less than 1e-6. The strategy does not observe synthetic four-event per day ordering in intraday data and reports daily close drawdowns. The final incomplete rotation is excluded, and there are zero open positions and terminal marks in payoffs. The results show returns of -92.7967%, a drawdown of 92.7967%, 896 trades, a win percentage of 36.7188%, a win to loss ratio of 2.0795%, and a loss to win ratio of 2.5193%. The population used is the current nine wrappers, not the entire point-in-time ETF universe. The original Nagel 48 industry long-short daily weights differ from this strategy. The data sources are https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf. This conclusion is specific to this implementation and not all reversal theory. There was no adoption or tuning from the holdout. The actual 15-minute, dividend, and whole-share execution has not been validated.

A news scan found the note about a sector reversal strategy. The strategy used nine original sector ETFs and worked as follows: it calculated the returns of each ETF over five sessions and compared them to the average return of all sectors. Every five sessions, it bought the three ETFs with the lowest returns, using an equal amount of cash for each, at the next opening price, and used any unused cash for other investments. The strategy held these investments for five sessions and then sold them at the next opening price. There were no mandatory stop-loss or target prices set, so this strategy is not suitable for all trading venues. The strategy was tested using data from 2017 to 2022, with 2016 used as a warm-up period. The testing period overlaps with other research and is not a completely untouched sample. The strategy assumed a fee of 5 basis points and a slippage cost of 5 basis points for each buy and sell, but these costs were not actually measured. The strategy used long-only orders, meaning it only bought and never sold short, and it shared cash among investments. The strategy's performance was measured using daily closing prices, not intraday prices, and it excluded the final incomplete rotation of investments. The results showed a return of -21.4363%, a maximum drawdown of 52.7304%, and other performance metrics. The strategy was tested on a current population of nine sector ETFs, not on the entire universe of ETFs that existed at the time. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.nber.org/system/files/working_papers/w17653/w17653.pdf. This conclusion is specific to this implementation of the sector reversal strategy and does not apply to all reversal theories. The strategy's performance in real-time trading, including 15-minute, dividend, and whole-share execution, has not been validated.

A computer program executed a test of a trading strategy on three stocks: TSN, PPC, and JBS. The test used a strategy called breakout, which is a method of deciding when to buy or sell a stock, and looked at the last 20 days of trading and the last 730 days of data. The test stopped trading if losses reached 3 percent and aimed for a 6 percent gain. The results were: TSN had a return of -1.468 percent, a biggest loss of 8.739 percent, and 36.364 percent of trades were winners out of 22 trades; PPC had a return of -13.312 percent, a biggest loss of 22.030 percent, and 24.138 percent of trades were winners out of 29 trades; JBS had a return of 5.029 percent, a biggest loss of 12.038 percent, and 40 percent of trades were winners out of 25 trades. The test did not account for costs and did not use a separate set of data to verify the results, so it cannot be used to make actual trading decisions. The data used for the test came from https://runway.adityaperswal.workers.dev/data/stock-bars and the test settings were saved in a file called cli.json. This was a diagnostic test only and did not have permission to make actual trades.

A news scan found the note about JBS frozen protein breakout and range test. The note says that before getting the price, a prior 20-day period completed a breakout and a 50-day simple moving average, which is a calculation to find the average price of a stock over a certain period. The next open was an entry signal with a stop loss at 1 times the average daily range of the last 20 days and a target at 2 times that range. The plan was to exit after 5 sessions on the next open, without pyramiding, which means adding more positions to an existing trade. The note also mentions that training data from 2022 to 2024 was used, with a holdout period from 2025 to September 4, 2026, and a warm-up period in 2021. Fees of 5 basis points and slippage of 10 basis points were included in the final analysis. The current selection of names is for exploration, not for a prospective or causal policy study. The note mentions that all-adjusted SIP bars are paginated, with unique dates, and synthetic 9-hour and 16-hour rows encode ordering, but not observed intraday bars. An independent cash ledger reconciles below a very small threshold. The training target was 1.5 to 2, with a stop loss at 1 to 1.25. The note started with 309 bars from June 13, 2025, but there was no training, so the later period was not independently trained. A positive estimate does not justify adoption. The note also mentions that there is no actual 15-minute or raw dividend replay, so the daily barrier model is not eligible. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars, with artifacts and results in test.py and trades.json. The QuarterKelly method, which is a formula to determine the optimal bet size, is authorized at 0, and there is no current signal.

A news scan found the note: prior20 completed-close breakout and SMA50, which is a type of moving average, were tested. The test had a nextopen entry, signalADR20 stop1/target2, 5sessions exitnextopen, no pyramiding, which means not increasing the size of the trade. Entryday barriers were included, stopfirst both-touch, adversegapactualopen/favorabletargetgapcapped. The test was trained from 2022 to 2024 and had a diagnostic holdout from 2025 to September 4, 2026, with a warmup period in 2021. Fees of 5 basis points and slippage of 10 basis points were applied to each side separately in the final vectorbt replay, which is a type of backtest. The stress test was 10/20 basis points. The current selected names make this an exploratory study, not a prospective or causal policy study. All-adjusted SIP bars were paginated, dates were unique, and synthetic 9h/16h rows encoded ordering, not observed intraday bars. An independent cash ledger reconciles below 2e-12, terminal open is false, and only natural exit payoffs were considered. The training target was 1.5/2 stop 1/1.25 sweep saved, fixed 1/2 unchanged. The training and holdout result does not establish a positive stable expectancy, so the implementation was rejected, not all supply signals. No actual 15-minute, +1R, raw dividend, or whole-share replay was done, so the daily barrier model is ineligible. The source is https://runway.adityaperswal.workers.dev/data/stock-bars, and the artifacts are in test.py/results.json/trades.json and archive. Quarter Kelly was authorized 0, and there is no current signal.

A news scan found the note: the TSN frozen protein breakout and range test had its net holdout rejected. The test was done before fetching the price, and it was based on a prior 20-day completed close breakout and a 50-day Simple Moving Average, which is a calculation to find the average price of a stock over a certain period. The next open was set as the entry point, with a stop loss at 1 times the Average Daily Range of the last 20 days and a target at 2 times that range. The test also included a 5-session exit at the next open, and no pyramiding, which means adding more positions to an existing trade. The entry day barriers were included, with a stop loss that would be triggered if the price touched either the stop loss or the target first. The test also accounted for adverse gap and favorable target gap, which refer to unexpected changes in the stock price. The test was trained on data from 2022 to 2024, and then tested on data from 2025 to September 4, 2026, with 2021 used as a warm-up period. The test included fees of 5 basis points and slippage of 10 basis points on each side, which are costs associated with buying and selling stocks. The stress test was done with 10 and 20 basis points. The current selected names make this an exploratory study, not a prospective or causal policy study. The results were adjusted for splits and dividends, and the dates were unique. The test used synthetic 9-hour and 16-hour rows to encode the ordering, but not observed intraday bars, which are stock price data recorded throughout the day. The independent cash ledger reconciles below 2e-12, which means the errors are very small. The terminal open is false, and only natural exit payoffs were used. The training target was 1.5 to 2, with a stop loss of 1 to 1.25, and the fixed 1 to 2 was unchanged. The result of the training and holdout does not establish a positive stable expectancy, so the implementation was rejected, but not all supply signals were used. The test did not use actual 15-minute data or raw dividend data, so the daily barrier model is not eligible. The data source is https://runway.adityaperswal.workers.dev/data/stock-bars, and the results can be found in artifacts/test.py/results.json/trades.json and archive. The Quarter Kelly method, which is a formula to determine the optimal bet size, was authorized at 0, and there is no current signal.

A preregistered note was made before retrieving prices, and this is not a backtest of a panel, but rather a generic price analysis. The analysis used 560 bars of GRAL data from June 12, 2024, to September 4, 2026, and 672 bars of SPY data from January 2024. The first 50 bars were excluded from the analysis to avoid early noise. A signal was generated when the close was higher than the prior 20 closes and the 50-day simple moving average, which is a calculation of the average price over a 50-day period. The signal triggered a buy order at the next open, with a stop loss set at 1 average true range, which is a measure of price volatility, and a target at 2 average true ranges. The analysis also included a 5-session timeout and no overlap between trades. A fee of 5 basis points, which is 0.05%, and slippage of 5 basis points were applied to each side of the trade. The results showed a return of 23.4898% and a drawdown, which is the peak-to-trough decline, of 26.8884% during the training period, and a return of -36.2152% and a drawdown of 39.0181% during the holdout period. The stress test, which is a simulation of extreme market conditions, applied a fee of 25 basis points and slippage of 25 basis points, resulting in a return of -42.5155% and a drawdown of 44.821%. The analysis used a standalone cash ledger and reconciled the results with a vectorbt explicit long-only orders, which is a type of trading strategy. The results were not affected by synthetic row ordering or observed minute execution. A news scan found the source of the data at https://runway.adityaperswal.workers.dev/data/stock-bars, and the artifacts and tests are available at artifacts/test.py and tests.json. The analysis used a specific version of the GRAL and SPY data, identified by the SHAs 9459602a452562126bbf0d5db729213359f239a125fdfcc1dca061757f2fd4d8 and 7ecf3b2c2b253af24930d116e16f2c6635dc8fe8049167b3687025121a34b1d6, respectively. The quarterly dates and panel probabilities were not retrofitted to the price history. This negative holdout result rejects this implementation, but not all catalyst run-ups.