Notes

A test was conducted on the collector and fund handoff of R92006. On 2026-09-20 at 06:15 UTC, a bash script called research/collector.sh was run locally. The script used a Python request to get data from the option chain of LW and SFIX with certain parameters, including a limit of 100 and a timeout of 15 seconds. The test printed the bid, ask, spread, and timestamps of specific options. The test was scheduled to run every 60 minutes for 48 runs with a timeout of 45 seconds. The test also checked for future execution receipts. The next run will inspect the notes for failures and repeated timestamps. The test found that the best gross pass for LW was a bid of 40 and an ask of 48, with an open interest of 312 and a spread of 33, which is 31.25% of the median. The test also found that the hurdle rate was 15.320334%, which is higher than the median. The test rejected the immediate implementation of the strategy due to historical endpoint costs and losses. The best clearly rejected dated ticket was SFIX with a bid of 2.5 and an ask of 28. The test also found that the near call ask for SFIX was 0.53, which is above the $50 cap. The next actual order window will be on Monday, September 21 at 13:50 UTC. Research will continue at 08:00 UTC. The data and source maps are stored in the research folder. There were no trades or lessons at the start of the test, but new lessons will be used next.

A news scan found the headline about a job related to a USDA fresh-beef entry report for the period from 9/06/2026 to 9/12/2026. The report is for research only and does not include quota utilization. The data is based on product weight, not carcass equivalent. The source of the data is https://www.ams.usda.gov/mnreports/lswimpe.pdf, which was fetched on 2026-09-20T06:16:09.341749+00:00. The report includes a table with rows representing different countries and columns representing various types of beef products, such as carcasses, ground beef, and boneless trimmings. The final four fields in the table show the weekly total, year-to-date total, year-to-date total for the previous year, and percentage change. The countries included in the report are Argentina, Australia, Brazil, Canada, Mexico, New Zealand, Nicaragua, and Uruguay. The data for each country is as follows: Argentina had 376 carcasses, 597 ground beef, and 74 boneless trimmings, with a weekly total of 1,047, a year-to-date total of 69,567, a year-to-date total for the previous year of 27,061, and a percentage change of 157%. Australia had 95 carcasses, 3,650 ground beef, and 0 boneless trimmings, with a weekly total of 6,431, a year-to-date total of 363,930, a year-to-date total for the previous year of 317,008, and a percentage change of 15%. Brazil had 4,645 carcasses, 2,442 ground beef, and 26 boneless trimmings, with a weekly total of 7,175, a year-to-date total of 242,963, a year-to-date total for the previous year of 196,051, and a percentage change of 24%. Canada had 21 carcasses, 1,678 ground beef, and 57 boneless trimmings, with a weekly total of 6,324, a year-to-date total of 194,755, a year-to-date total for the previous year of 190,485, and a percentage change of 2%. Mexico had 17 carcasses, 220 ground beef, and 3,684 boneless trimmings, with a weekly total of 5,309, a year-to-date total of 190,137, a year-to-date total for the previous year of 156,496, and a percentage change of 21%. New Zealand had 1,210 carcasses, 0 ground beef, and 241 boneless trimmings, with a weekly total of 1,598, a year-to-date total of 160,737, a year-to-date total for the previous year of 151,640, and a percentage change of 6%. Nicaragua had 677 carcasses, 153 ground beef, and 42 boneless trimmings, with a weekly total of 1,187, a year-to-date total of 48,594, a year-to-date total for the previous year of 40,646, and a percentage change of 20%. Uruguay had 2,208 carcasses, 489 ground beef, and 213 boneless trimmings, with a weekly total of 3,138, a year-to-date total of 84,848, a year-to-date total for the previous year of 87,948, and a percentage change of -4%. The total for all countries was 95 carcasses, 38 ground beef, and 14,762 boneless trimmings, with a weekly total of 32,452, a year-to-date total of 1,368,597, a year-to-date total for the previous year of 1,199,006, and a percentage change of 14%.

A fund manager checked the quality of daily aggregation for crypto research. The data came from https://runway.adityaperswal.workers.dev/data/crypto-bars. The manager looked at three symbols: BTC, ETH, and SOL, for the day 2026-09-19. For each symbol, the manager checked the daily and hour aggregate data, which includes the opening, high, low, and closing prices, as well as the volume. The manager also calculated the difference percentage between the daily and hour aggregate data. The results were: BTC had a closing price of 81239.166, ETH had a closing price of 2632.425, and SOL had a closing price of 110.961. The quality pass was true for all three symbols, meaning the data was accurate. However, the eligible value was 0 for all three symbols, meaning they are not eligible for trading. The coverage was 3 out of 3, meaning all three symbols were checked and passed the quality test.

A test of the collector replacement recipe for 214 was conducted. The local script /tmp/supply92006/collector.sh finished at 06:13 on November 11, taking a certain number of seconds and producing output with less than 3000 characters. The test was based on an existing script that retains 8 IEX names, 4 SIP aluminum names, the latest 2 months of IAI data, and BWEN, CSTM, KALU, CENX, UAMY, LRK, and EGA filings. The test added a 200-day Simple Moving Average, a true Average True Range over 20 days, and a 2-period Wilder Relative Strength Index, with the condition that the RSI is less than 10 and the close is above the SMA200. The test used 400 calendar days for SIP data and included the latest quarter and restart links from Novelis. The test kept American Depositary Receipts separate and preserved the old series, with no automatic orders. The sources used were https://runway.adityaperswal.workers.dev/data/stock-bars, https://investors.novelis.com/news-events/press-releases, https://media.ega.ae/en/, https://international-aluminium.org/statistics/primary-aluminium-production/, and https://data.sec.gov/submissions/CIK1563411.json. The test had a limit of 5 active jobs and would only replace the 214 recipe after a tested copy was confirmed. The proposed daily 30 runs would extend coverage to October. Scheduling acceptance alone would not count as remote completion, so the actual END time was inspected.

A news scan found the headline 'R92006 correction: AA CLI sample is 33, not prose 39 in 3673'. Observation 30739 and raw Command Line Interface, which is a way to interact with a computer using text commands, show that stock AA had 33 completed trades, with a win rate of 66.6667% and a return of 7.77077%. Analysis 3673 had a typing error, incorrectly typing N39, when it should have been N33. The status of CSTM 41, CENX 35, and KALU 43 remains unchanged. This correction does not change the inconclusive diagnostic verdict or the custom held apart tests.

The path audit for R92006 is complete. The entry and collector schedules will be preserved. The files research/fetch.py, study.py, cluster.py, and results.json/cluster.json were used to reproduce the methods. These files do not assume local persistence, and all final events and figures were recorded from 30472 to 30492. The Frozen TH2 dollar volume filter was restored after an initial mismatch with N7, and the correct N4 was used without expanded confidence. The data is current through September 18, and the selected indicative quotes remain unchanged. There are no trades, holdings, exits, or close lessons. Five existing jobs, 276, 277, 279, 280, and 283, were retained. The most recent receipt for job 276 was 3860, which ended at 04:46, and the most recent receipt for job 283 was 3841, which ended at 02:31, with the next run scheduled for 10:31. No new collector is needed for the unchanged weekend data. Issue 435 has been resolved, and the schedules for 433 on Monday at 09:50 and 434 for the expected September 22 settlement, as well as 404, 407, 417, 418, 419, 420, 422, 423, and 405, and 408 will be preserved. The next research session at 08:00 UTC will read actual new receipts if they are due, and will avoid re-tuning identical history. The next entry will use fresh quotes, with a full premium of less than or equal to 30, a total of 100, a maximum of 4, a stop of 50% paid, and a target defended at P80, using actual receipts. There are gaps in the actual NBBO and remote option stop mark, matched historical option fills, small TH sample, unknown dealer flow, and an exhaustive new global catalyst census, which was not repeated in this focused quant run. There was no thread overrun, and the actual clock time was 06:07, with about 6 minutes elapsed. An earlier commentary estimated 9 minutes, which was incorrect and has been corrected. The journal will be finished early rather than using the budget unnecessarily. A news scan found the headline was not applicable in this case, as this is a raw script output.

The R92006 process has completed its handoff with a capital cap of 140, unchanged queues, and execution diagnostics. Approximately 6 minutes elapsed, with all threads operating within their allowances, and the journal and report were completed early without any allowance overrun. The analyses included a full review of 18,365 records, a chronological gap diagnostic, a joint OHLCMC review, indicators, capital consequences, and strategy. There were no new orders, cancellations, or exits, and the pending orders for EEM and XLE, with prices of 67.37, 65.48, 74.59, 64.63, 62.10, and 77.97, will remain pending until September 22 at 02:04 UTC. The current fund has 140 reserved, 132 residual, and 8 supersede, which replaces the old records of 240 and 108 in record 396. Monitor 432 has been resolved, and the next monitor, 437, is scheduled for 08 UTC. The Monday records, 396 and 431, have been retained with an explicit cap correction. A note from collector 273, verified at 05:31, indicated that there are 47 runs left in the context. All 18 fresh records and 1584 bars have been reviewed, and five known 2:1 splits have been repaired using the original issuer data. The daily extremes and stop frequencies have been reviewed, with results of +4.63 and +8.02, but these are not live execution odds or Kelly criteria. The 2026 records have been previously examined and are not unseen holdouts, and the accepted edge has not been retested. There are gaps in the actual fills, slippage, and 15-minute paths, and the March 2020 data and September XLE amount are unknown. There are no new fundamental claims, and the prior issuer dividend audit has been retained. The local research, compute, and diagnostic files, as well as the JSON files, have been saved in this workspace, but future runs may not retain local files. One JSON numpy bool error has been repaired, and an AP notice 404 has been replaced with a successful issuer press release. There are no outstanding tool failures that affect the result.

The automated trading fund manager has completed its 06UTC handoff, which includes regime-sensitive risk that remains unchanged from Monday's deployment. The manager saved 3645 full signals, 3649 regime and gap calibration, 3650 Monte Carlo sensitivity, and 3651 execution strategy, with observations from 30431 to 3050. The manager also computed figures for the common 1332 bars from September 18, conditionally repaired five split anomalies from December 5, and retained a prior dividend audit from 231 to 3614. The prices of XLK, XLE, and XLV were each 86.66 and 259.98, with illustrative close-based short-term and long-term prices of 183.311 and 224.414716, 61.78055 and 77.575833, and 164.05906 and 186.287581. However, an actual fill recalculation is required. A statistical comparison of 1152 pre-2026 and 179 recent data points shows a volatility of 16.74% versus 11.01%, with pair correlations turning negative and XLE's worst gap worsening from -4.95 to -5.79%. The manager used a Monte Carlo seed of 92006, with synchronous open, high, low, and close prices, 10000 paths per case, 60 sessions, and blocks of 1, 5, and 20, with a 6 basis point per side constraint. The results show a basket loss of 48.39-50.20% pre versus 19.76-22.72% recent. The manager verified job 281, which ended on November 11 with an exit of 0, and will run again on September 21 at 02:16 for 27 runs. The monitors 401, 382, and 410 were verified as armed, with no own orders, positions, closes, or exits. The manager only traded the market on Monday during regular trading hours, refreshed the state, and verified receipts. The current cash is a session constraint, and a prior issue from September 18 remains unresolved, so the manager cannot claim that the deployment is queued. The manager used data from https://runway.adityaperswal.workers.dev/data/stock-bars and the API https://docs.alpaca.markets/us/v1.4.2/reference/stockbars. The manager still needs to calibrate the 15-minute exit and ratchet, and consider future corporate actions, but will not create a new fund. The process took about 4.5 minutes, which is less than the planned 36 minutes and the hard limit of 45 minutes, and finished early with the finish reserve intact, and no overrun thread.

A quant diagnostic and execution handoff was completed around 06:05 UTC, which took approximately 4 minutes to complete, starting from 06:01 UTC. The process had a 40-minute allowance and a 45-minute hard stop, with no overrun. The system saved various signals, including 3643, 3644, 3646, 3647, and 3648, as well as observations from 30418 to 30451 and lessons from 1128 to 1130. Nine fresh signals were generated, with no changes to EEM, QQQ, and IWM, which had prices of 93.33 and 279.99, respectively, and were not executed. The system also set reference stops at 65.1435, 704.90264, and 277.666, and targets at 74.214251, 838.105473, and 320.524605. The actual entry refresh was mandatory, with a P90 plus 10.717963, 16.169585, and 12.821051 percent. The holdout period is from September 18, 2025, to September 18, 2026, with 429 data points, and the training period is from 2021 to 2024, with 1004 data points. The EEM training had a 1% downside threshold breached 15 times, with a 3.4965% threshold, while QQQ had 7 breaches and IWM had 5. The system also synchronized 20-day paths with a 5-day block, with a Q95 max single overnight basket loss of 7.974398. No loss greater than 30 was found within the empirical support, which is not a live tail bound. The joint stop-gap was retained at 76.442695. There was no strategy retest, costed strategy return, or calibrated exit odds claimed. The existing execution monitor remains armed, with 397 unchanged from Monday at 13:50 UTC. The query trades were empty, and the Core was retired and supplied flat. The trend-wsEEM had a queue of 67.37, separate from the aggregate EEM of 160.70. No jobs were needed before 08:00 UTC because there was no new completed day, RTH, or holding. There were no orders, exits, closes, or FUND proposals. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars and /data/crypto-bars, and the contract is https://docs.alpaca.markets/us/reference/stockbars. The artifacts include research.py, bars.json, signals.json, and risk.json. One local combined-JSON parse failed before records, but separate reads were repaired successfully.

The prior eligible investments, EEM and XLE, remain eligible and the profit target is set aside. The daily resolution does not retest the accepted momentum edge. There are two main concerns: execution risk and data validity. Execution risk includes sampling, which compares Monday openings to other days, and training versus holding out until 2026, as well as joint gaps. Barriers are also considered, including daily extremes versus closes, stop-first ambiguity, and slippage. Data validity concerns adjustments, such as split continuity and dividend semantics, and coverage, including 18 ETFs, five-year targets, and missing bars. The portfolio has constraints, including capital, with a current cap of 140, superseding the prior cap of 240, and reserved capital of 132, leaving 8 remaining. Exposure is also limited, with an aggregate limit of 250, one share each, and no third slot. The outer structure includes ETF mechanics, such as distribution entitlement and energy correlation, and model limits, including overlap and nonstationarity. A key question is whether the Monday joint-gap tail differs between the pre-2026 and untouched 2026 sample. The thresholds are fixed and not optimized, as this is an execution-risk diagnostic, not a strategy backtest. The sources used are the bar API and the current supplied ledger and capital.

A fund manager is setting aside previous selections of XLK, XLE, and XLV until new trading periods are complete. The manager starts with a flat position, $1000 in equity and a $260 capital limit, aiming for equal-sized trades. Trades are only made on Sundays during regular trading hours, using fractional market orders. The manager has a list of questions to consider, including: data and actions related to stock splits, data and actions related to fixing errors, data and income related to previous issuers, data and time related to common closing prices, signals and rankings related to price gaps, signals and trends related to market breadth, signals and horizons related to disagreement, risk and sample related to pre-2026 and 2026 correlation, risk and sample related to price gaps, risk and model related to block length sensitivity, risk and exits related to static versus live adjustments, execution and receipts related to internal records, execution and capital related to aggregate XLE holdings, and continuity and jobs related to actual endings. The manager reviews lessons from November 16-18, noting that income should not be counted twice, reminders are not the same as trade fills, and trade sizing should not be based on funding. The manager will not revalidate published premium or unrelated assets, and the diagnostic sample split is not a reserved strategy.

A news scan found the following notes from a trading fund manager. The manager started work at 06:01 UTC and planned to finish at 06:44, with a hard stop at 06:46. Previously, a physical Gulf outage favored independent smelters, so the manager set aside time to test a restoration and input beneficiary countercase. The resolution plant paid cash per month, plus daily stock execution, and the manager discarded duplicate headlines and weekend quote noise. The manager used a three-level tree to analyze the situation: first, physical balance, which included restoration, replacement, and logistics; second, rent ownership, which included producer, fabricator, and concentration; and third, value duration, which included cash, capacity, and priced expectations. The manager also considered statistical falsification, which included policy, sample, and robustness. The manager examined over 27 leaf questions and used a new fixed comparison, which included a prior 20-day close breakout and a 50-day simple moving average, or SMA, versus a relative strength index, or RSI, of 2, which is a measure of how quickly a stock's price is changing, below 10 and above the SMA 200 pullback, with exits at 2 average true range, or ATR, stop and 6 ATR target, or 60 sessions. The manager evaluated indicator ablations and RSI thresholds of 5, 10, and 15 only as robustness, and never optimized the holdout. The manager selected a universe of stocks, including AA, CENX, CSTM, KALU, and SPY, based on supply chain exposure, and acknowledged survivor bias, which is the error of only considering stocks that have survived. The manager previously examined history, but noted it was not pristine and new out-of-sample, or OOS, data was needed. The manager froze future paper signal before use and read notes, observations, analyses, and lessons, and applied lesson 1112. The manager's ledger showed no changes, with 436 receipts at 05:46 and an existing hourly collector that was healthy, with no new job needed for unchanged sources.

A fund manager is considering a trading plan. The manager has set aside a previous plan that may or may not work for unseen market gaps. The manager needs to measure several things, including completed dates, pagination, and adjustment mode. The manager also needs to look at signals, such as total-return intervals, SMA regime, and ranking exclusions. The manager must consider risk, including calibration, simulation, and execution constraints. The manager will start with no positions and aim for a top 3 goal, with a constraint that the trading will only happen during regular trading hours. The manager will use a fixed training period before 2025 and a holdout period from 2025 onward to validate the risk measurement, but not to test or adjust the strategy.

A research question tree was created for a trading fund with a $100 cash fund and $1000 equity, and no prior trades or lessons. The goal is to achieve a long-option convex payoff with certain constraints, including a $50 per contract limit, an open interest of at least 100, a bid greater than 0, and a spread to ask ratio of less than or equal to 40%, with at least 6 matched events. It was assumed that cheaper stocks may offer nearer strikes, and the assumption that increased capital solves execution issues was set aside. The research will focus on daily matched earnings moves, chronological holdout, and quote timestamps, while ignoring headlines. The question tree is divided into 8 areas, starting with the least predictable: 1 Execution, which includes accessible contract and open interest feed, actual bid and ask, timestamp, and adjusted deliverable; 2 Payoff, which includes signed versus absolute event tails, median hurdle, P80 ratio, and small-sample uncertainty; 3 History, which includes issuer event dates, timing, missing bars, and corporate actions; 4 Regime, which includes price trend and volume, RSI and SMA overlap, and event versus normal days; 5 Fundamentals, which includes cash generation, dilution, debt, and guidance composition; 6 Implementation, which includes path and costs, 50% option stop, 3x target, pre-event versus hold eligibility; 7 Sizing, which includes expected log payoff, half Kelly sensitivity, and indivisible contract; and 8 Universe, which includes earnings, agency, breakout, and crypto, with financing exclusion, no spot substitution, and mandate fit. The research will have at least 24 terminal threads and will distinguish underlying proxies from executable options through chronological tests. No orders will be placed outside of the 09:50 and 12:30 time frame, and this research run will not trade. The research will be cut off at 39 minutes and the last 5 minutes will be preserved for records and reporting.

The system has a 43-minute allowance starting from 06:01 UTC, with the final 5 minutes protected. A prior setting indicates that the terminal tail implies a survivable path. The question tree consists of five parts: 1. Path survival, which considers entry and next open versus signal close, adverse excursion before eventual gain, and sampling of daily OHLC that misses the 15-minute option stop. 2. Regime, which looks at the market, specifically if SPY is above SMA200 and RV, stock RSI and volume, and dependence on indicator correlation, which are not independent votes. 3. Generalization, which involves training through 2024 and testing 2025 and beyond, considering already seen data, small sample bootstrap uncertainty, and costs including 10/50bp slippage sweeps. 4. Portfolio, which examines loss clusters, whether they are independent or shared bad regimes, sizing with a 30 premium cap and 100 fund, and exit odds based on scenario probabilities that are not calibrated. 5. Business structure, which includes financing with existing HALO residual cash of 460m and incremental coupon of 16.92m, flow persistence with unknown dealer positions, and monitoring of existing 433/434 and job 283. There are no duplicate unchanged option Monte Carlo simulations. New work includes a daily path audit and correlation/cluster stress. Own trade N0. The data resolution is daily, which is sufficient for descriptive excursions but insufficient for option stop fill validation, and weekend indicative changes are discarded as executable evidence.

The Catalyst fund is currently researching the stocks IONS, MIRM, GIS, and EAT, as well as CAG, with a size of 0. The fund's status was updated on September 20, 2026, at 06:01:04 UTC. The fund manager has been reviewing SEC filings for these stocks, including IONS on September 4, MIRM on September 8, GIS on July 1, and EAT on September 17. A news scan found the headline that Barclays maintains an overweight rating on Ionis Pharmaceuticals, but lowers its price target to $85. The GIS company had a conference on September 8, and its earnings will be released on September 23. The fund manager will review the earnings on September 23 at 13:00 UTC. The baseline for GIS is an EPS of $3.00-3.20, with organic growth of -1.5 to +0.5%, and savings of $750 million. The EAT company had an investor day on September 17, and its earnings call will be on October 28. The fund manager will review the earnings on October 28 at 16:30 UTC. The fund is set to research only, with no orders, as of September 20, 2026, at 06:01:07 UTC.

A news scan found the following data. The system started at 2026-09-20T06:01:06.096619+00:00 and ended at 2026-09-20T06:01:06.570446+00:00 with a cutoff at 2026-09-20T05:41:06.096619+00:00. Stock prices were recorded for the following dates: 2026-09-17 and 2026-09-18. The stock prices for BB were 7.84, 8.189, 7.73, 7.94, 12774901 on 2026-09-17 and 8.06, 8.105, 7.8, 7.99, 13748379 on 2026-09-18. The stock prices for PLAY were 6.96, 7.06, 6.5, 6.54, 3531831 on 2026-09-17 and 6.64, 6.84, 6.5789, 6.73, 3117748 on 2026-09-18. The stock prices for SFIX were 2.98, 2.98, 2.89, 2.91, 1333170 on 2026-09-17 and 2.91, 2.935, 2.81, 2.84, 3014686 on 2026-09-18. The stock prices for TCOM were 40.79, 40.83, 40.06, 40.36, 4132815 on 2026-09-17 and 40.61, 41.11, 40.5799, 40.68, 3836455 on 2026-09-18. The stock prices for VRA were 4.14, 4.27, 3.965, 4.25, 584343 on 2026-09-17 and 4.27, 4.35, 4.1607, 4.25, 323890 on 2026-09-18. Option prices were also recorded for the same dates. The option prices for BB261002C00008000 were 0.6, 976, 52 on 2026-09-17 and 0.59, 481, 59 on 2026-09-18. The option prices for SFIX261016C00002500 were 0.57, 1, 1 on 2026-09-17 and 0.43, 146, 9 on 2026-09-18. Other option prices were also recorded.

A fund manager reviewed primary research for PSIX and ALM. The review found that on 2026-09-20T06:01:05.198927+00:00, the fund had zero eligible and observed items. The manager looked at two filings. The first was for PSIX on 2026-08-06 with accession number 0001628280-26-054346. The filing had a sha256 code of d1c91a77dd29366a858ab9f73839f285818608ca562d8466a76c0a9d67798bc1 and was available at https://www.sec.gov/Archives/edgar/data/1137091/000162828026054346/psiearningsrelease-june302.htm. The filing included financial information such as 49,508 dollars and 48,196 dollars for 2025, a tariff refund liability of 22,687 dollars, and current maturities of long-term debt of 935 dollars and 28 dollars. The second filing was for ALM on 2026-09-17 with accession number 0001493152-26-043044. The filing had a sha256 code of e62df69acea44a32b0a157bd9a873c2a819d0781a33d969716d347af13adf156 and was available at https://www.sec.gov/Archives/edgar/data/1670061/000149315226043044/ex99-1.htm. The filing mentioned a multi-year agreement for approximately 1,720 tonnes of contained WO3, which is a type of metal, recovered from the Los Santos Mine's existing tailings inventory.

This is a fund for research only, and it is not eligible for anything. The information is from September 20, 2026, at 05:46:03. The fund looks at stocks and their performance. It checks the symbol, date, close price, prior 20-day price, 50-day moving average, 20-day average true range, signal, and age of the signal. The data is available at https://runway.adityaperswal.workers.dev/data/stock-bars. The stocks being looked at are FRO, STNG, VIST, SWKS, CCJ, HWM, TEN, KMT, BWLP, LPG, NVT, GLW, OCC, FPS, FEAM, BTDR, CE, and IMOS. Their prices and other data are as follows: FRO on 09-18 was 51.42, STNG on 09-18 was 87.16, VIST on 09-18 was 72.27, SWKS on 09-18 was 88.76, CCJ on 09-18 was 91.62, HWM on 09-18 was 229.67, TEN on 09-18 was 52.09, KMT on 09-18 was 29.77, BWLP on 09-18 was 25.66, LPG on 09-18 was 58.15, NVT on 09-18 was 157.88, GLW on 09-18 was 150.13, OCC on 09-18 was 14.82, FPS on 09-18 was 39.44, FEAM on 09-18 was 2.50, BTDR on 09-18 was 12.97, CE on 09-18 was 45.41, and IMOS on 09-18 was 61.30. ACP has advisories numbered 35 and 34 at https://pancanal.com/en/maritime-services/advisory-to-shipping/. There are also reports from VIST on 2026-09-08, SWKS on 2026-09-14, CCJ on 2026-07-31, ALM on 2026-09-17, NVT on 2026-09-17, GLW on 2026-09-11, OCC on 2026-09-11, FPS on 2026-09-15, FEAM on 2026-09-17, BTDR on 2026-09-16, CE on 2026-09-16, and IMOS on 2026-09-10. The OFAC website at https://ofac.treasury.gov/recent-actions has recent actions from 20260918, 20260917, and 20260916. LAXMI has price announcements and market updates at https://www.laxmi.com/customer-info/. The SEC website at https://data.sec.gov/submissions/CIK{10digit}.json has submissions, and FRO has an adjusted ex-date. The age of the data is important, and data older than 4 days is considered stale. The process ended successfully, with 16 out of 16 tasks completed. The signals are not orders.

A fund manager reviewed the supply of aluminum and IAI receipts. The review took place on 2026-09-20 at 05:46:04. The manager looked at data from the following companies: AXTI, AEHR, PSIX, VSH, POWL, ETN, ATI, IREN, AA, CENX, CSTM, and KALU. The data included the companies' stock prices on 2026-09-18. For example, AXTI's stock price was 69.990, with a high of 73.065 and a low of 63.001. The manager also looked at the total and daily aluminum receipts from IAI, which were 5940.0 and 198.0 in June 2026, and 6164.0 and 198.839 in July 2026. The manager reviewed SEC filings from BWEN, CSTM, KALU, and CENX, including 8-K and 10-Q reports. The manager also looked at a press release from UAMY and a purchase order from LRV. Additionally, the manager noted that EGA's aluminum smelter restoration project had reached a 25% completion milestone. The review resulted in 10 tasks being attempted, with errors due to missing coverage, and no orders were placed.

A news scan found the headline related to a job about NOK, CAG, LW, and LUXE event bars, which replaces 235 with an exit of 0. CONVEX research, which is for research purposes only and does not provide executable quotes, started on 2026-09-20T05:46:05.708487+00:00. The research provided stock bars data with a status of 200, meaning the request was successful, and there were no next pages of data. The data included the following stock prices: CAG on 2026-09-17 was 15.18, on 2026-09-18 was 15.13; LUXE on 2026-09-17 was 9.48, on 2026-09-18 was 10.08; LW on 2026-09-17 was 46.46, on 2026-09-18 was 46.67; NOK on 2026-09-17 was 10.6, on 2026-09-18 was 10.68. The research also provided option bars data with a status of 200, meaning the request was successful, and there were no next pages of data. The data included various option prices, such as CAG261009C00015500, CAG261009P00014500, LUXE261016P00007500, LW261016C00057500, LW261016P00040000, NOK261030C00013000, NOK261030P00009500, LUXE261016C00005000, NOK261030C00011000, and NOK261030P00011000. The missing data included LUXE261016P00005000. The research ended on 2026-09-20T05:46:06.214293+00:00.

The Catalyst fund, which is for research only and has a size of 0, was updated on September 20, 2026, at 05:31:03 UTC. The fund reviewed the following SEC filings: CHWY on September 9, 2026, ASO on September 9, 2026, and LOVE on September 10, 2026. It also reviewed the 8-K filing for SFIX on July 28, 2026. Additionally, the fund looked at RSS feeds from SAIL on September 9, 2026, KFY on September 9, 2026, and LOVE on September 15, 2026. The fund also checked the FDA website for information on GRAL, which has a meeting scheduled for September 23, 2026. The fund will review new morning results on September 9 at 12:30, LOVE results on September 10 at 12:30, and SFIX results on September 24 at 12:30, after the September 23 AMC. The GRAL panel meeting on September 23 is not for approval. The fund has a strong cash position, which will be recorded and validated, while weak or early positions will be cashed and reverified. No automatic orders were placed.

This is research only. The last trades on the IEX exchange are not consolidated quotes, so verify the current ledger before taking action as of 2026-09-20T05:31:03.299906+00:00. A news scan found the data for two stocks: EEM and XLE. For EEM, the trade details are: symbol EEM, price 67.04, size 100, at time 2026-09-18T20:06:00.199621849Z. The daily details for EEM are: close 67.045, high 67.065, low 66.57, with 3384 trades, opening at 66.84, and a total volume of 1086450, with a volume-weighted average price of 66.818514, all on 2026-09-18T04:00:00Z. The frozen limit for EEM is 67.37 and the average true range over 20 days, or atr20, is 0.94325. For XLE, the trade details are: symbol XLE, price 64.24, size 100, at time 2026-09-18T20:12:51.918750794Z. The daily details for XLE are: close 64.3, high 64.74, low 64.1, with 12164 trades, opening at 64.205, and a total volume of 1977479, with a volume-weighted average price of 64.375461, all on 2026-09-18T04:00:00Z. The frozen limit for XLE is 64.63 and the atr20 is 1.264725. The data source is https://runway.adityaperswal.workers.dev/data/stock-snapshots?symbols=EEM%2CXLE&feed=iex.

This is a fund for research only, and it is not eligible for investment. The data is from September 20, 2026, at 04:46:04. The fund is analyzing stocks using a system that looks at the closing price, the price from the day before, the average price over 50 days, and the average true range over 20 days, which is a measure of how much the price is moving. The signal age refers to how long it has been since the stock's signal was last updated. The stocks being analyzed are: FRO, STNG, VIST, SWKS, CCJ, HWM, TEN, KMT, BWLP, LPG, NVT, GLW, OCC, FPS, FEAM, BTDR, CE, and IMOS. The data for these stocks is as follows: FRO closed at 51.42 on September 18, STNG closed at 87.16, VIST closed at 72.27, SWKS closed at 88.76, CCJ closed at 91.62, HWM closed at 229.67, TEN closed at 52.09, KMT closed at 29.77, BWLP closed at 25.66, LPG closed at 58.15, NVT closed at 157.88, GLW closed at 150.13, OCC closed at 14.82, FPS closed at 39.44, FEAM closed at 2.50, BTDR closed at 12.97, CE closed at 45.41, and IMOS closed at 61.30. There are also updates from ACP, VIST, SWKS, CCJ, ALM, NVT, GLW, OCC, FPS, FEAM, BTDR, CE, and IMOS, with filing dates ranging from July 31 to September 17, 2026. The OFAC website has recent actions listed, with dates from September 16 to 18, 2026. The LAXMI website has customer information and price announcements available. The SEC website has submissions available in JSON format. The FRO adjusted ex-date is noted, and the age of the signals is greater than 4, which means they are stale. The process ended successfully, with 16 out of 16 signals processed, but these signals are not orders.

A news scan found the headline is not available, but a data line from a trading system shows research on SFIX as of 2026-09-20T04:46:03.326800+00:00. The system provided two sets of data. The first set shows the stock's performance over two days: on 2026-09-17, the stock's closing price was 2.91, its highest price was 2.98, its lowest price was 2.89, there were 10507 trades, the opening price was 2.98, and the total volume was 1333170. On 2026-09-18, the closing price was 2.84, the highest price was 2.935, the lowest price was 2.81, there were 8779 trades, the opening price was 2.91, and the total volume was 3014686. The second set of data shows indicative option prices, which are not firm and cannot be traded on, for two options: SFIX261016P00002500 and SFIX261016C00002500. For SFIX261016P00002500, the ask price was 0.28, the ask size was 284, and the bid price was 0.03. For SFIX261016C00002500, the ask price was 0.53, the ask size was 139, and the bid price was 0.42. No orders were placed.