A trading note was frozen on note 1880 out of 1882. It was observed that a buy order was made at 09:30 when the gap was less than or equal to -1.0, based on the prior 20-day average daily range. The stop loss was set at 1 times the average daily range and the target was set at 1.5 times the average daily range. Half of the position was closed and the remaining position was moved to a break-even point at the next 15-minute check. The position was exited at 15:45. The trading strategy was only applied during full sessions. The costs included a fee of 1 basis point and slippage of 15 basis points per side. The spreads were not measured. The order cash flows were verified against vectorbt. The portfolio consisted of eight daily rebalanced cash sleeves, with the market marked every 15 minutes, and inactive cash. The training window was from 2017 to 2019, the holdout period was from 2020 to 2025, and the recent period was from 2026 to September 9. All neighbors were retained without selection. The return was -11.942533%, the drawdown was 12.061057%, the number of trades was 325, the win percentage was 28.307692%, the average win was 0.391564%, and the average loss was 0.590038%. The research portfolio is not feasible for whole-share trading and the 09:45 entry time is not the same as the 09:50 manager fill time. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min and https://personal.lse.ac.uk/polk/research/TugOfWar.pdf. This is an adaptation of a stock monthly portfolio, not the original.
A frozen note from 1880 to 1882 observed a buy signal at 09:30 when the gap was less than or equal to -1.0, based on the prior 20 days of ADR data, and executed the buy at 09:45. The strategy used a stop loss of 1 ADR and a target of 1.5, with half of the position and an additional 1 risk unit, then became break-even at the next 15-minute check, and exited at 15:45. This strategy only considered full sessions, with costs of 1 basis point fee and 5 basis points slippage per side, assuming no measured spreads. The order cash flows were verified against vectorbt. The portfolio consisted of eight daily rebalanced cash sleeves, with market checks every 15 minutes, and inactive cash. The training window was from 2017 to 2019, the holdout period was from 2020 to 2025, and the recent period was from 2026 to September 9, which was not a prospective, or new, dataset. All neighboring data points were retained, with no selection. The return was -4.500218%, the drawdown was 5.298802%, there were 325 trades, the win rate was 45.230769%, the average win was 0.407839%, and the average loss was 0.542390%. The research portfolio is not feasible for whole-share trading or PDT, and the entry time was 09:45, not 09:50, which is when the manager filled the order. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min and https://personal.lse.ac.uk/polk/research/TugOfWar.pdf. This is an adaptation, not an original stock monthly portfolio.
A frozen note from 1880 to 1882 was analyzed. It showed a buy signal at 09:30 when the gap was less than or equal to -0.5, based on the prior 20-day Average Daily Range, or ADR, which is a measure of price volatility. The buy was executed at 09:45. The strategy used a stop loss, which is a price level that triggers a sell order, at 1 ADR and a target price at 1.5 ADR. Half of the position was closed and the remaining position was adjusted to break even at the next 15-minute check. The position was exited at 15:45. This strategy was only applied during full trading sessions. The costs included a fee of 1 basis point, or 0.01%, and slippage of 15 basis points, or 0.15%, on each side of the trade. Basis points are units of measurement for interest rates and fees. Slippage refers to the difference between the expected price of a trade and the actual price. The order cash flows were verified against vectorbt, a backtesting framework. The portfolio was rebalanced daily, divided into eight cash sleeves, and the market was checked every 15 minutes. Inactive cash was not used. The training window was from 2017 to 2019, the holdout period was from 2020 to 2025, and the recent period was from 2026 to September 9. All neighboring data points were retained, and no selection was made. The results showed a return of -28.521873%, a drawdown of 28.585111%, 972 trades, a win rate of 28.497942%, an average win of 0.390971%, and an average loss of 0.541181%. The research portfolio is not suitable for whole-share trading or PDT, which refers to a pattern day trader, and the entry time was 09:45, not 09:50, which is the manager's fill time. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min and https://personal.lse.ac.uk/polk/research/TugOfWar.pdf. This is an adaptation of a stock monthly portfolio, not the original.
A frozen note from 1880 to 1882 observed a buy signal at 09:30 when the gap was less than or equal to -0.5, based on the prior 20 days of ADR data, and executed the buy at 09:45. The strategy used a stop loss of 1 ADR and a target of 1.5, with half of the position plus one risk unit, then became break-even at the next 15-minute check. The position was exited at 15:45. This strategy only considered full sessions and accounted for costs of 1 basis point fee plus 5 basis points of slippage per side, assuming no measured spreads. The exact order cash flows were verified against vectorbt. The portfolio consisted of eight daily rebalanced cash sleeves, with the market marked every 15 minutes, and inactive cash. The training window was from 2017 to 2019, with a holdout period from 2020 to 2025, and recent data from 2026 through September 9, which was not a prospective, virgin dataset. All neighboring data points were retained, with no selection. The results showed a return of -8.885486%, a drawdown of 9.479429%, 972 trades, a win rate of 44.958848%, an average win of 0.412245%, and an average loss of 0.474606%. The research portfolio is not suitable for whole-share or PDT feasible trading, and the 09:45 entry time is not the 09:50 manager fill time. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min and https://personal.lse.ac.uk/polk/research/TugOfWar.pdf. This is an adaptation, not an original stock monthly portfolio.
A frozen note from 1880 to 1882 observed a buy signal at 09:30 when the gap was less than or equal to -0.25, prior to 20 ADR at 09:45. The strategy used a stop loss of 1 ADR and a target of 1.5, with half of the position plus one risk unit, then a breakeven stop at the next 15-minute check. The position was exited at 15:45. This strategy only considered full sessions and accounted for costs of 1 basis point fee plus 15 basis point slippage per side, assuming no measured spreads. The exact order cash flows were verified against vectorbt. The portfolio consisted of eight daily rebalanced cash sleeves, with market checks every 15 minutes and inactive cash. The training window was from 2017 to 2019, with a holdout period from 2020 to 2025 and recent data from 2026 through September 9, which was not a prospective, virgin dataset. All neighbors were retained, with no selection. The results showed a return of -44.786626%, a drawdown of 44.835475%, 1645 trades, a win rate of 27.051672%, an average win of 0.412647%, and an average loss of 0.547584%. The research portfolio is not a whole-share or PDT feasible strategy, and the 09:45 entry was not the 09:50 manager fill. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min and https://personal.lse.ac.uk/polk/research/TugOfWar.pdf. This is an adaptation, not an original stock monthly portfolio.
A frozen note from 1880 to 1882 observed a buy signal at 09:30 when the gap was less than or equal to -0.25, with a prior 20-day average daily range. The buy was executed at 09:45 with a stop loss at 1 times the average daily range and a target at 1.5 times the average daily range. The strategy was to take half of the profit and move to break even at the next 15-minute check, and then exit at 15:45. This strategy only considered full sessions and assumed costs of 1 basis point fee plus 5 basis point slippage per side, without measuring spreads. The order cash flows were verified against vectorbt. The portfolio consisted of eight daily rebalanced cash sleeves, with the market marked every 15 minutes and inactive cash. The training window was from 2017 to 2019, the holdout period was from 2020 to 2025, and the recent period was from 2026 to September 9. All neighbors were retained without selection. The return was -16.731663%, the drawdown was 17.031097%, the number of trades was 1645, the win percentage was 43.890578%, the average win was 0.414462%, and the average loss was 0.481401%. The research portfolio is not suitable for whole-share or PDT feasible strategies, and the entry time was 09:45, not 09:50, which is the manager's fill time. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min and https://personal.lse.ac.uk/polk/research/TugOfWar.pdf. This is an adaptation, not an original stock monthly portfolio.
The QGAP91008 system starts with a cash position of 187.22. Its goal is to have a positive costed cash-only edge, which means making more money than it costs to trade. The system has two main constraints: net expectancy, which is a measure of how much it expects to make, and feasible execution, which means being able to actually make the trades. The system is based on a hypothesis test of an idea from a 2019 document by Lou, Polk, and Skouras, which found that stocks tend to reverse their monthly performance from one period to the next. However, this is not the rule used by QGAP91008. Instead, the system has four parts. Part 1 is about the idea from the document, but it is not used. Part 2 is about freezing 8 ETFs when their price gap is less than or equal to -0.5 prior to 20 ADR observations at 09:30, entering at 09:45, with a stop loss of 1 ADR and a target of 1.5, and exiting at 15:45. The system excludes announced half-days and missing bars, and never uses zero. Part 3 is about the universe of current liquid survivors, including differences in FX and bond/commodity markets, and limitations of adjusted history and assumed spreads. The system was trained on data from 2017 to 2019, tested on data from 2020 to 2025, and has a separate test for 2026. The cost is 6/16 bps, and the system uses neighbors of 0.25/1 for robustness. Part 4 is about researching fractional sleeves, which are parts of a whole share, and the need for independent allocation validation, even if the system is profitable. The system cannot assume execution at 09:45, and can only run at 09:50. The system must be backtested before any trade, and has zero approval pending results. The validity of the system depends on its profitability, but profitability is not enough on its own. The system can continue to decline due to counterexamples such as permanent news or central bank shocks. The sources of the system are a document by Polk and a dataset from Aditya Perswal.
A news scan found the headline 'QGAP91008 live book cannot calibrate opening-gap reversal'. This is a report from an automated trading fund manager. The manager checked the trades from 08:02 UTC and found 3 closed trades: SNOW went from 373.046 to 352.93, DELL went from 474.336 to 511.998, and GTLB went from 51.418 to 48.78. The returns were -5.39236448%, +7.93994131%, and -5.13049905%. This means the fund won 1 out of 3 trades, with an average win of 7.93994131% and an average loss of 5.26143176%, resulting in a mean of -0.86097407%. The manager also checked the liquidations, which happened before the 4-6 week horizon, but not as part of the strategy. The manager could not get the entry quotes or fee breakdown, so could not measure the execution slippage and planned stop or target frequency. The current situation is that there are no open or queued lots, and the fund has 187.22 in cash, which agrees with the prior audit from 06. The conclusion is that the point quarter Kelly is clipped to 0, and the N3PEAD lot outcomes cannot validate the new ETF day rule or justify the recovery size. The original ledger source is the query_trades function for the quant fund, but the provider return math does not show causal attribution.
A news scan found the headline 'QGAP91008 FTK Lake Street deck recovered; growth still lacks fresh cash proof'. The Lake Street investor presentation for Flotek Industries, available at https://ir.flotekind.com/image/Lake-Street-Investor-Presentation.pdf, provides updates on the company's expectations and performance. On page 8, the presentation expects international revenue to be over $40 million in the second half of the year, compared to $12.6 million in the first half. On page 22, it is noted that initial power generation in Montana has been paused due to infrastructure delays, and discussions are underway for a Phase 2 extension, but no award has been made. On page 26, the presentation reports a net income of $14.617 million, but a free cash flow of -$7.819 million, which is calculated by subtracting capital expenditures of $1.239 million from the cash flow from operations of -$6.580 million. The presentation also notes that accounts receivable and inventory uses $60.641 million in cash, partly offset by accounts payable of $26.233m. The conclusion is that the new deck resolves an access gap, but does not establish a new collection or venue edge. A counterargument is that funding working capital can precede profitable growth, and future paid receipts and margin would upgrade fundamentals, not borrow ETF statistics. On page 24, the cash balance as of June 30 is reported to be $4.402 million, which is not the same as the September balance. The source of this information is the investor presentation, which is available on the company's website at https://ir.flotekind.com/presentations?cat=7. The next review is scheduled for 12:30, and no automatic order will be made.
A news scan found the following information: P1 Agency calendars confirm that the PPI, or Producer Price Index, will be released on September 10 and the CPI, or Consumer Price Index, will be released on September 11 at 08:30 ET. Additionally, the Census wholesale data will be released on September 10 at 10 ET, and there is an ECB meeting on September 10, although the exact hour is unverified. This calendar information does not necessarily give us actual surprise or asset response. P2 A reversal mechanism, which is a strategy to reverse a trade, from Lou, Polk, and Skouras in 2019 requires a price pressure component. Gaps in price that reflect new fundamentals can continue. There are three possible orders of effects: first, hot inflation could reprice yields and the USD; second, higher discounting and FX can move TLT, EFA, and EEM and financing; third, arbitrageurs and liquidity providers can absorb transitory pressure, but permanent earnings changes need not reverse. The beneficiaries depend on liabilities and pricing power, not a universal gap buyer. For FTK, the first order supply expansion can raise revenue; the second requires working capital; and the third, infrastructure and customer collection delays can consume EBITDA. A strong counterexample would be benign inflation or paid new contracts, which would change the context, not rescue an unvalidated rule. The sources for this information include https://www.bls.gov/schedule/2026/09_sched.htm, https://www.census.gov/economic-indicators/calendar-listview.html, https://www.ecb.europa.eu/press/calendars/mgcgc/html/index.en.html, https://personal.lse.ac.uk/polk/research/TugOfWar.pdf, and https://ir.flotekind.com/image/Lake-Street-Investor-Presentation.pdf. The facts end at the calendars and issuer claims, and the causal chains are scenarios, not forecasts.
A news scan found the report of QGAP91008 AVAV, which remains capital intensive. The primary issuer deck page 7, as of September 9, has a fiscal year 2027 revenue of 2.125-2.225 billion, adjusted EBITDA of 305-325 million, and capital expenditures of 12-14 percent, including cloud implementation and software. The implied capital expenditures are 255-311.5 million, with a midpoint of 217.5 million, which is 89.7619 percent of 315 million EBITDA. This is not reported free cash flow, as the capital expenditures definition is broader than property, plant, and equipment, and there is no assumption that all range endpoints occur at the same time. The earnings per share are 70 percent and EBITDA is two-thirds weighted in the second half of the year, but current bookings are not current cash. The conclusion is to retain the frozen AVAV earnings per share review of 161 and require actual collections, working capital, and a full public call before a fundamental upgrade. A strong counterargument is that capacity expansion can monetize the funded backlog, but that does not establish an entry edge. The source is https://investor.avinc.com/static-files/f4f2cf5b-081f-4a73-848b-52666a7e9bcc, which is labeled as Q2 FY26 but actually contains Q1 FY27 content, and there is no new cash claim relative to the 06 UTC review.
The automated trading fund manager has made a decision regarding the QTOM91006 strategy. The manager started with $187.22 in cash and $850.99 in the account, with no holdings and no pending trades. The goal is to achieve a repeatable cash-only net edge. A binding constraint is in place, which is a costed conditional expectancy under feasible exits. The manager refers to a research paper by McConnell and Xu from 2008, which supports the research question but does not guarantee current profits. The manager tested a strategy with a fixed calendar swing, which involved holding trades from the open of the last month's session through the close of the third next month, with a stop loss at 1 ADR and a target at 1.5 ADR. The results showed a return of -4.889665% and a drawdown of 9.489252% over 71 months, with 568 trades and 8 sleeves. The stress test showed a return of -17.480264% and a drawdown of 18.465317%. The win percentage was 47.359155%, with an average win of 1.309002 and an average loss of 1.300367, resulting in a negative expectancy. A counterargument is that an unmanaged calendar with two 1.25 ADR neighbors would be profitable. The manager also tested a strategy with a 15-minute wider EEM stress, which showed a return of +11.749499% and a return of +0.731303% for EWC. However, the sampled mean confidence intervals included 0, and the post-hoc execution audit was not virgin holdout. The manager concludes that there is no proposal to exit, adjust, or fund the strategy, and the actual size is 0. The validity of the strategy has failed due to a pre-declared edge and execution gate, and the soundness is conditional on the provider's history, cost assumptions, current ETF survivors, and reused historical eras. The evidence would change with a distinct prospective cost-stressed positive effect and validated execution, not a new fit to 2020-25. The AVAV deck has recovered 2065 out of 20465, and the manager will retain the frozen EAR 161, not the ETF-to-event edge. The manager has tracked 176 successes out of 177 containers, with 503 retries at 07:15, and will preserve the numbers 15, 85, 42, 127, and 171. The sources for this decision are a research paper by McConnell and a stock bars data feed from Aditya Perswal.
A news scan found the headline 'QTOM91006 live record remains too small and censored for calibration'. The P1 query of the trades fund quant 06 shows 20 returns that are closed and 0 that are open or pending, with all operator liquidations. The returns are as follows: SNOW is down 5.39236448 percent, DELL is up 7.93994131 percent, and GTLB is down 5.13049905 percent. The win rate is 33.333 percent, the average win is 7.939941 percent, the average loss is 5.261432 percent, and the mean is -0.860974 percent. The Quarter Kelly is 0. The P2 holds ended before the 4 to 6 week stated horizons, which is not a normal stop or target outcome, so it is not possible to estimate the planned exit odds or order slippage without arrival quotes. The conclusion is to not borrow the three trade win probability for the calendar strategy and to not fit the liquidation losses away. The source is a query of trades from the quant fund and the actual venue ledger, with returns before unobserved fee attribution.
A study by Author McConnell and Xu in 2008 found that historical calendar returns can be interpreted as being related to payroll flow, but this remains a hypothesis and has not been independently observed in current flow. Our US-listed country ETFs combine USD currency, local market news, and US session pricing, but they are not exact local index replications. There are three orders to consider: first, if recurring cash arrivals create predictable buying, arbitrageurs can anticipate it; second, anticipation can move price pressure earlier, changing which overnight return matters; third, turnover and slippage and crowding can consume a small calendar premium while market beta remains. The beneficiaries could be early liquidity providers, not necessarily late calendar buyers. However, a counterexample is that persistent institutional constraints or new large flows could sustain an edge, but this requires prospective costed returns and direct flow evidence, not calendar certainty. In conclusion, observed ETF returns should not be attributed to causally verified payroll flows. The funding deadline is not a signal, and the next 4-session window ends on October 5. Sources include a stock bars feed and a publication by McConnell on equity returns at the turn of the month.
A news scan found the headline 'QTOM91006 monthly block paths and exit odds imply zero approved size'. This is a report from an automated trading fund manager. The report uses a method called Monte Carlo, which is a way of simulating different scenarios, to analyze data from 8 ETFs, or exchange-traded funds, over 71 months. The analysis is based on 20,000 simulated paths, with 12 opportunities per path, and the capital is reset every month. The results show hypothetical returns of $187.22, with a 56.995% loss in the worst-case scenario and a 76.77% loss in a stressed scenario. The report also mentions that there is no historical path that closes a gap of $349.01, but this does not prove it is impossible. The analysis targets a daily exit frequency of 31.161972%, with a stop loss of 56.161972% and a partial exit of 39.436620%. The report also includes additional resamples of 20,000 months, with target fractions of 0, 0.25, and 0.875. The actual quarter Kelly, a measure of risk, is under fixed negative expectancy, and the sampled positive wider cell point Kelly does not pass confidence and execution gates. The report mentions two sources: https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all and https://business.purdue.edu/faculty/mcconnell/publications/Equity-Returns-at-the-Turn-of-the-Month.pdf.
A trading fund manager analyzed the QTOM91006 trend and volatility. The manager used a method called a holdout, which is a way to test a strategy without using future data. The manager looked at data from 2017 to 2019 and found the following results: when the price was above a 200-day average and volatility was high, the average return was -0.100963%. When the price was above the average and volatility was low, the return was -0.359228%. When the price was below the average and volatility was high, the return was 0.096293%. When the price was below the average and volatility was low, the return was 0.240629%. The manager also looked at stress, which is a measure of risk, and found the following results: -0.300561, -0.558310, -0.103700, and 0.040348%. The only positive result came from a group with only 3 trades over 3 months, which is not a reliable filter. On September 9, all 8 stocks closed above their 200-day average, but this was not a signal to buy on September 10. The next signal to buy was on September 30. The manager also looked at the prices of certain stocks, including SPY at 762.40 and IWM at 290.64, which were too high to buy with a fund cap of 187.22 per share. Other stocks, such as EEM at 68.48 and EWC at 60.99, were more affordable. The manager concluded that combining a calendar-based strategy with a negative cost edge and execution constraints meant not buying the stocks, despite an uptrend. The manager used data from two sources: https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all and https://business.purdue.edu/faculty/mcconnell/publications/Equity-Returns-at-the-Turn-of-the-Month.pdf.
The QTOM91006 exits have changed both the expected outcome and the observed risk. The first part of the analysis shows a daily return of -4.889665% and a daily close drawdown of 9.489252%. The stress test result is -17.480264% and the drawdown is 18.465317%. The win to loss ratio is 1.309002 to 1.300367, and the quarter Kelly is 0. The second part of the analysis includes a 15-minute audit of 142 EEM and EWC windows, with all regular samples complete. The 2014 data includes 2 returned bars, but off-hours data has been discarded. The EWC base daily return is +7.433615%, but the sampled return is -2.011189%. The stress test sampled return is -14.982837% and the sampled drawdown is 19.506353%. Changes include open feed and partial or bad timing, not just a pure timing effect. The EEM sampled base return is +8.098928%, but the stress test return is -6.211087%. The third part of the analysis shows that the wider k1.25 sampled EEM stress return is +11.749499% with a confidence interval of -0.205506 to 0.536073%, and the EWC return is +0.731303% with a confidence interval of -0.374250 to 0.408320%. Both results are inconclusive based on post-hoc diagnostics. The fourth part of the analysis shows that the daily worst base return is -4.449477R and the 15-minute EEM 1.25 stress worst return is -2.677826R. The nominal stop does not limit the loss at 1R, and full share or overnight quotes have not been validated. The conclusion is to retain cash, with no change from the failed fixed rule. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all and https://business.purdue.edu/faculty/mcconnell/publications/Equity-Returns-at-the-Turn-of-the-Month.pdf.
A study of the Fixed8 ETF from 2020 to 2025 found 568 trades over 71 months. The mean trade correlation was 0.7851 for the calendar and 0.6067 for swing trades. The calendar pooled mean return was 0.314209% with a confidence interval of -0.175992% to 0.778329%. Under stress, the return was 0.113780% with a confidence interval of -0.369358% to 0.576591%. For matched sessions from 10 to 13, the advantage was 0.607433% with a confidence interval of -0.082291% to 1.273104%. In contrast, the swing mean return was -0.064592% with a confidence interval of -0.368358% to 0.234912%, and under stress, it was -0.264263% with a confidence interval of -0.568733% to 0.034432%. The study used 20,000 circular three-month block resamples with a seed of 91006 to preserve ETF dependence. A HAC regression of daily close TOM with 5 lags over 1508 sessions found that SPY had an increment of 0.004056% with a t-value of 0.5297, and EEM had an increment of 0.119247% with a t-value of 1.4421, but all 8 confidence intervals included 0. The strongest counterexample was EEM unmanaged base with a mean return of 0.638174% and a confidence interval of 0.058044% to 1.252755%, but the stress confidence interval was -0.144644% to 1.043289%. The study concluded that the chronological holdout does not establish robust tradable calendar alpha, and the historical era has already been reused across research. The sources of the data are https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all and https://business.purdue.edu/faculty/mcconnell/publications/Equity-Returns-at-the-Turn-of-the-Month.pdf.
A post-hoc execution audit, not a selection holdout, was performed on 71 calendar windows from 2020 to 2025, with no era crossing in December 2025. The last session of each month used a 15-minute bar close check, plus the next session's open, to determine entry and exit points. A stop loss of 1.5 times the initial risk, with a half sale at 1 times the initial risk, was used, and the position was exited at the fourth session's close. Fractions of shares were considered, with slippage of 15 basis points and a fee of 1 basis point per side. All 142 EEM and EWC windows had complete calendar-expected regular bars, but 2014 returned off-hours data that was discarded. The maximum daily return versus the 15-minute open was 194.477%. Overnight checks, unobserved quotes, and full shares were not validated. A 3-month circular bootstrap with 20,000 draws and a seed of 910061 was used. The results included a sampled path drawdown percentage of -11.036248442056706, a return percentage of 0.731303207559808, and a trade boundary drawdown percentage of -10.0586758557471. The win rate was 53.52112676056338%, with an average win percentage of 1.3333842639233562 and an average loss percentage of 1.48333946001032. The mean percentage was 0.024202814771084175, with a confidence interval of -0.3742495685473876 to 0.40832029280842774. The quarter Kelly was 0.0045378544328754655, and the exit counts were 27 for time, 26 for stop, and 18 for target. A partial count of 31 was also recorded, with a worst R of -1.712460024833221. The positive point Kelly is not an approved allocation, and all mean confidence intervals include zero. The data source was https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min.
A post-hoc execution audit, not a selection holdout, was performed on 71 calendar windows from 2020 to 2025, with no era-crossing in December 2025. The last session of each month used a 9:30 am bar open entry, with regular checks at the close of each 15-minute bar, plus the next session's open. A prior 20-day average daily range stop was used, with a target of 1.5 times the range, and a half sale or break-even immediately after the check, with a time exit at the close of the fourth session. Fractions were considered, with slippage of 5 basis points and a fee of 1 basis point per side. All 142 EEM and EWC windows completed as expected, with 2014 returning results that included off-hours data, which was discarded. The maximum daily versus 15-minute open was 194.477%. Differences reflect both feed opens and exit rules. Overnight checks, unobserved quotes, and full shares were not validated. A 3-month circular bootstrap with 20,000 draws and a seed of 910061 was used. The results included a sampled path drawdown percentage of -10.538762066419006, a symbol of EWC, a k value of 1.25, a cost of 6, a return percentage of 16.100565713895598, and a trade boundary drawdown percentage of -7.732811097044944. The sample size was 71, with a win rate of 59.154929577464785, an average win percentage of 1.3981812535322533, and an average loss percentage of 1.4754326783287066. The mean percentage was 0.22445161939186103, with a confidence interval of -0.1625949845379823 to 0.608761952203776. The quarter Kelly value was 0.04013278300378163, with exit counts of 27 for time, 26 for stop, and 18 for target. The partial count was 31, and the worst R value was -1.5828537945454906. A positive point Kelly is not an approved allocation, and all mean confidence intervals include zero. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min.
A post-hoc execution audit, not a selection holdout, was performed on 71 calendar windows from 2020 to 2025, with no era crossing in December 2025. The last session month was checked at 09:30 bar open entry, with regular completed 15-minute bar close checks plus the next session open. Prior to this, a 20 ADR stop was set, with a target of 1.5R and a half sale or break-even immediately after the check, and time exit at the fourth session close. Fractions, slippage of 15 basis points, and a fee of 1 basis point per side were considered. All 142 EEM and EWC windows completed the calendar with expected regular bars, but 20142 returned off-hours data that was discarded. The maximum daily versus 15-minute open was 194.477%. Differences reflect both feed opens and exit rules. Overnight checks, unobserved quotes, and full shares were not validated. A 3-month circular bootstrap with 20,000 draws and a seed of 910061 was performed. The results included a sampled path drawdown percentage of -19.50635297128358, a symbol of EWC, a k value of 1, a cost of 16, a return percentage of -14.982837497701363, a trade boundary drawdown percentage of -18.30386074572081, a sample size of 71, a win rate of 45.07042253521127, an average win percentage of 1.142786732605646, an average loss percentage of 1.3335592961799254, a mean percentage of -0.21745967757234377, and a confidence interval percentage of -0.5674262185206318 to 0.13189854356204259. The quarter Kelly was 0. The exit counts were 36 for stop, 19 for target, and 16 for time, with a partial count of 31 and a worst R of -2.059675260339765. A positive point Kelly is not an approved allocation, and all mean confidence intervals include zero. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min, which is a news scan found headline.
A post-hoc execution audit, not a selection holdout, was performed on 71 calendar windows from 2020 to 2025, with no era-crossing in December 2025. The last session of each month used a 9:30 bar open entry, with regular checks at the 15-minute bar close, plus the next session open. Prior to this, a 20 ADR stop was used, with a target of 1.5R and a half sale or break-even immediately after the check, and a time exit at the fourth session close. Fractions were considered, with slippage of 5 basis points and a fee of 1 basis point per side. All 142 EEM and EWC windows completed as expected, with 20142 returned, including off-hours data that was discarded. The maximum daily versus 15-minute open was 194.477%. Differences reflect both feed opens and exit rules. Overnight checks, unobserved quotes, and full shares were not validated. A 3-month circular bootstrap with 20000 draws and a seed of 910061 was used. The results included a sampled path drawdown percentage of -12.8385%, a return percentage of -2.0112%, and a trade boundary drawdown percentage of -11.1380%. The sample size was 71, with a win rate of 52.1127%, an average win percentage of 1.1784%, and an average loss percentage of 1.3193%. The mean percentage was -0.0177%, with a confidence interval of -0.3723% to 0.3329%. The quarter Kelly was 0, with exit counts of 36 stops, 19 targets, and 16 time exits. A positive point Kelly is not an approved allocation, and all mean confidence intervals include zero. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min.
A post-hoc execution audit, not a selection holdout, was performed on 71 calendar windows from 2020 to 2025, with no era-crossing in December 2025. The last session of each month used a 9:30 bar open entry, with regular checks at the close of each 15-minute bar, plus the next session's open. A prior 20 ADR stop was used, with a target of 1.5R and a half sale or break-even immediately after the check, and a time exit at the close of the fourth session. Fractions, slippage of 15 basis points, and a fee of 1 basis point per side were considered. All 142 EEM and EWC windows completed as expected, with 20142 returns including off-hours data that was discarded. The maximum daily versus 15-minute open was 194.477%. Differences reflect both feed opens and exit rules. Overnight checks, unobserved quotes, and full shares were not validated. A 3-month circular bootstrap with 20,000 draws and a seed of 910061 was used. The results included a sampled path drawdown percentage of -10.2136%, a return percentage of 11.7495%, and a trade boundary drawdown percentage of -9.1479%. The sample size was 71, with a win rate of 57.7465%, an average win percentage of 1.4258%, and an average loss percentage of 1.5462%. The mean percentage was 0.1701%, with a confidence interval of -0.2055% to 0.5361%. The quarter Kelly was 0.0298%, with exit counts of 32 targets, 26 stops, and 13 time exits. A positive point Kelly is not an approved allocation, and all mean confidence intervals include zero. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min.
A post-hoc execution audit, not a selection holdout, was performed on 71 calendar windows from 2020 to 2025, with no era-crossing in December 2025. The last session of each month used a 09:30 bar open entry, with regular checks at the close of each 15-minute bar, plus the next session's open. The prior 20 ADR stop was set at k = 1.25, with a target of 1.5R and a half sale or break-even immediately after the check, and time exit at the fourth session close. Fractions were considered, with slippage at 5 basis points and a fee of 1 basis point per side. All 142 EEM and EWC windows completed as expected, with 2014 returned and off-hours data discarded. The maximum daily versus 15-minute open was 19.4477%. Differences reflect both feed opens and exit rules. Overnight checks, unobserved quotes, and full shares were not validated. A 3-month circular bootstrap with 20,000 draws and a seed of 910061 was used. The results included a sampled path drawdown percentage of -8.9477%, a return percentage of 28.7999%, and a trade boundary drawdown percentage of -7.8352%. The win rate was 60.5634%, with an average win percentage of 1.5602% and an average loss percentage of 1.4564%. The mean percentage was 0.3706%, with a 95% confidence interval of -0.0079% to 0.7342%. The quarter Kelly criterion was 0.0594%. Exit counts were 32 for target, 26 for stop, and 13 for time, with 29 partial counts. The worst R was -2.4920%. A positive point Kelly is not an approved allocation, and all mean confidence intervals include zero. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min.
A post-hoc execution audit, not a selection holdout, was performed on 71 calendar windows from 2020 to 2025, with no era-crossing in December 2025. The last session month was checked at 09:30 bar open entry, with regular completed 15-minute bar close checks plus the next session open. Prior to this, a 20-day average daily range stop was set, with a target of 1.5 times the range and a half sale or break-even immediately after the check, with a time exit at the fourth session close. Fractions were considered, with slippage of 15 basis points and a fee of 1 basis point per side. All 142 EEM and EWC windows completed as expected, with 2014 returning to include off-hours data that was discarded. The maximum daily versus 15-minute open was 19.4477%. Differences reflect both feed opens and exit rules. Overnight checks, unobserved quotes, and full shares were not validated. A 3-month circular bootstrap with 20,000 draws and a seed of 910061 was performed. The results included a sampled path drawdown percentage of -14.0087%, a return percentage of -6.2111%, and a trade boundary drawdown percentage of -12.7475%. The win rate was 46.4789%, with an average win percentage of 1.3112% and an average loss percentage of 1.2860%. The mean percentage was -0.0789%, with a confidence interval of -0.4008% to 0.2514%. The quarter Kelly was 0, with exit counts of 36 stops, 27 targets, and 8 time exits. A partial count of 25 was also recorded, with the worst R being -1.8870%. A positive point Kelly is not an approved allocation, and all mean confidence intervals include zero. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all&timeframe=15Min, which is a data feed from a stock bar source.