Notes

A news scan found the final audited results for LPG N911 as of 2026-09-10 04:00:00+00:00, with a close price of 54.41. The prior 20-day average was 55.2 and the 50-day simple moving average, which is a calculation to find the average price of a stock over a period, was 45.5232. The average daily range, which is the difference between the high and low prices, over 20 days was 1.6903, and the realized volatility, which is a measure of how much the price of a stock changes, over 20 days was 36.0464. The signal, which is an indicator to buy or sell, was false. The simulation results showed different scenarios with various parameters, including the probability of loss, median return, and drawdown, which is the decrease in value from a peak. The audit results included information on trades, exit reasons, and hypothetical stop and target prices. The base mean percentage return was 3.1465 and the one-share equity percentage return was 6.3937. The hypothetical stop price was 51.02945 and the hypothetical target price was 64.55165. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=LPG&timeframe=1Day&feed=sip&adjustment=all

A news scan found the following data: the LPG N911 audited LPG 40 lookback sweep resulted in the following statistics for different training and holdout sets with various parameters, such as 0.002 and 0.005, which refer to the risk management settings. The results are as follows: for 40_train_0.002, the return was 35.41308918032699, the drawdown was 34.46040681428914, the win rate was 43.47826086956522, the sample size was 23, the mean percentage change was 1.8474491962704767, the average win percentage was 12.457092377850358, the average loss percentage was 6.313814789560201, the half Kelly criterion, a betting strategy, was 0.07415250446225233, and the mean confidence interval was between -2.2336421657925714 and 6.265553720701358. For 40_train_0.005, the return was 17.95795321012563, the drawdown was 36.77789718299185, the win rate was 43.47826086956522, the sample size was 23, the mean percentage change was 1.2381861974593074, the average win percentage was 11.784361289537316, the average loss percentage was 6.874256181062237, the half Kelly criterion was 0.052535142424673636, and the mean confidence interval was between -2.8184916129508 and 5.629861119255484. For 40_holdout_0.002, the return was 45.740373422395045, the drawdown was 18.098671685819955, the win rate was 58.333333333333336, the sample size was 12, the mean percentage change was 3.7091516734554317, the average win percentage was 11.376833168871604, the average loss percentage was 7.025602420127211, the half Kelly criterion was 0.16301336313887949, and the mean confidence interval was between -2.1276906311228663 and 9.624300745540019. For 40_holdout_0.005, the return was 35.61579329814571, the drawdown was 20.04092850430521, the win rate was 50, the sample size was 12, the mean percentage change was 3.0887517591584572, the average win percentage was 12.552290872382974, the average loss percentage was 6.374787354066058, the half Kelly criterion was 0.12303538017726312, and the mean confidence interval was between -2.7131738923849507 and 8.968515738238617. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=LPG&timeframe=1Day&feed=sip&adjustment=all

A news scan found the data line: LPG N911 audited LPG 20 lookback sweep with the following results. The results are for four different scenarios: 20_train_0.002, 20_train_0.005, 20_holdout_0.002, and 20_holdout_0.005. For 20_train_0.002, the return was 48.98218086253797, the drawdown was 31.115229826980396, the win rate was 50, the sample size was 28, the mean percentage was 1.9445669994768977, the average win percentage was 10.802684841082453, the average loss percentage was 6.913550842128656, the half Kelly was 0.09000387533670048, and the mean confidence interval was between -1.7719483382544146 and 5.897158410084646. For 20_train_0.005, the return was 25.94238282900291, the drawdown was 33.94851160410507, the win rate was 50, the sample size was 28, the mean percentage was 1.3347230309459528, the average win percentage was 10.139850606397669, the average loss percentage was 7.4704045445057625, the half Kelly was 0.06581571478498013, and the mean confidence interval was between -2.3595596902823157 and 5.263669591192631. For 20_holdout_0.002, the return was 47.449725642820454, the drawdown was 23.338751184893624, the win rate was 53.333333333333336, the sample size was 15, the mean percentage was 3.1464916759042922, the average win percentage was 12.073742388540126, the average loss percentage was 7.05608056710809, the half Kelly was 0.13030308145761024, and the mean confidence interval was between -2.03238332127213 and 8.47262017320489. For 20_holdout_0.005, the return was 34.758744854335895, the drawdown was 26.49188530632448, the win rate was 46.666666666666664, the sample size was 15, the mean percentage was 2.529457657563713, the average win percentage was 13.080890197868406, the average loss percentage was 6.703045815202893, the half Kelly was 0.09668522628436635, and the mean confidence interval was between -2.6184367216772775 and 7.8237246497807025. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=LPG&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline 'LPG N911 audited LPG 10 lookback sweep'. The results are as follows: there are four sets of data, each representing a different test of the LPG system. The first test, 10_train_0.002, had a return of 84.55523844841954, a drawdown of 33.07113324979668, and a win rate of 50, based on 32 samples. The average win percentage was 11.856249719555434 and the average loss percentage was 6.9267704190080215. The half Kelly criterion, a measure of optimal bet size, was 0.10394263399363207. The mean confidence interval, a measure of the range of possible outcomes, was between -1.0474643049770571 and 6.188136132381562. The second test, 10_train_0.005, had a return of 52.31432572453944, a drawdown of 35.62989732728781, and a win rate of 50, based on 32 samples. The average win percentage was 11.187112940208358 and the average loss percentage was 7.483545040376069. The half Kelly criterion was 0.08276415728585895. The mean confidence interval was between -1.6394096027069722 and 5.5529066517344186. The third test, 10_holdout_0.002, had a return of 61.820645632941904, a drawdown of 27.277803341230435, and a win rate of 58.82352941176471, based on 17 samples. The average win percentage was 10.900188338225798 and the average loss percentage was 7.2943698496142435. The half Kelly criterion was 0.15634185995657995. The mean confidence interval was between -1.5190204936264435 and 8.521028031618954. The fourth test, 10_holdout_0.005, had a return of 46.12863130691691, a drawdown of 28.575108967951113, and a win rate of 52.94117647058824, based on 17 samples. The average win percentage was 11.411944811523842 and the average loss percentage was 6.910301648057741. The half Kelly criterion was 0.12222768465671097. The mean confidence interval was between -2.1081448887233365 and 7.871842926892385. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=LPG&timeframe=1Day&feed=sip&adjustment=all

A news scan found the headline about LPG N911 measured changes in Panama from July to August. The changes include an increase of 37.50985028% in canal hours, which is equivalent to an increase of 9.52 hours, a decrease of 2.46445498% in transits, and a decrease of 6.49717514% in arrivals. The non-transit difference of means changed from 14.51 to 23.63 hours. This data is about a canal, which is a man-made waterway, and transits, which refer to the number of ships passing through the canal. The data also mentions that the N2 monthly aggregates did not match the queues or causal estimate, which is a statistical prediction. The data was saved in a file called physical_stats.json. https://pancanal.com/wp-content/uploads/2026/09/adv-34-2026-monthly-canal-operations-summary-august-2026.pdf ; https://pancanal.com/wp-content/uploads/2026/08/ADV-26-2026-Monthly-Canal-Operations-Summary-July-2026.pdf

A news scan found the headline with the following details: at 00:00-01, there were zero open, pending, or closing items, and two closed operator liquidation wins. The fund's cash was 278.67, equity was 850.99, and capital was 212.7475. A job, number 173, started at 00:00:19 and ended at 21/21, with three remaining, and the next one is at 01:00:16, which covers the next period at 02 UTC. There were 12, 18, 36, and 73 active items. There was an issue with Bombardier and a CENTCOM error, and a collector missed some amendments in an 8K form, which required manual reading of all BWEN and GD forms. No action was required for orders or stops and targets. query_trades fund=supply; list_jobs fund=supply

A report from September 10, 2026, was recovered on September 11. The report shows that electrical inventory increased by 2.3 percent and sales increased by 0.1 percent, on a seasonally adjusted basis, excluding manufacturer branches. A survey of approximately 4200 firms had a response rate of 54.8 percent, with 61.5 percent of respondents reporting sales and 62.1 percent reporting inventories. The data is nominal, meaning it cannot be used to determine if there is a real shortage or the cash position of individual issuers. https://www.census.gov/wholesale/pdf/mwts/currentwhl.pdf

A news scan found the text of a May 27 FY26 10-K report. The report says that 99% of the revenue of Helios Pool comes from this pool, and no individual charterer makes up more than 10% of the revenue. Using the canal saves about 15 days compared to going around the Cape. However, the time spent waiting to use the canal is typically not paid, and the rising tolls to use the canal may not be passed on to the charterers. New dual-fuel Panamax ships that are chartered in can use the old locks. The concentration risk is related to the pool, not because 99% of the revenue comes from one ultimate charterer. https://www.sec.gov/Archives/edgar/data/1596993/000159699326000025/lpg-20260331x10k.htm

On August 5, the issuer reported results for the June quarter of 2026. The results showed a time charter equivalent, or TCE, of 75,926 dollars per day, revenue of 187.9 million dollars, net income of 138.3 million dollars, and adjusted net income of 107.2 million dollars. The issuer also reported a gain of 30.1 million dollars from the sale of Cobra. The adjusted earnings before interest, taxes, depreciation, and amortization, or EBITDA, was 165.4 million dollars, which includes a gain from the sale of an asset. The TCE already accounts for voyage expenses, so fuel costs should not be subtracted again. These figures are historical and do not reflect the fixture on September 11. https://www.dorianlpg.com/newsroom/news-and-press-releases/news-details/2026/Dorian-LPG-Ltd--Announces-First-Quarter-Fiscal-Year-2027-Financial-Results/default.aspx

A news scan found the headline 'XLE N911 EIA export mix unchanged'. This refers to a report from the Energy Information Administration, or EIA, which is a government agency that tracks energy data. The report, released on September 10, is about the week of September 4. It shows that the total export mix was 11.868 million barrels per day, with 3.417 million barrels per day of crude and 8.451 million barrels per day of products. The changes from the previous week were -0.028, -1.066, and +1.038. The data is encoded in a format called CSV cp1252, which requires special decoding. It is also noted that not all product exports are transported by clean tankers, which are a type of ship that carries oil products. https://ir.eia.gov/wpsr/table7.csv

A news scan found the headline that TSM's August revenue was unchanged. On September 11, the issuer's browser read the data. There were 403 direct requests that were recovered through the web. In August, the NTD was 5,148,060, which is an increase of 53.3% from the same time last year, although this number has not been officially audited. It is worth noting that this revenue observation does not take into account the paid capacity or margins. https://investor.tsmc.com/english/monthly-revenue/2026

A news scan found the headline that the public replay for AVAV N911 AVAV is still inaccessible. The issuer had an archived conference on September 10 at 12:50 EDT, and the conference link was checked on the web and through direct requests on September 11 at 00:02, but it still shows a 404 error, which means the page cannot be found. There is no commentary available and no assertion from management, as they have not provided any information. https://event.summitcast.com/view/e82pfcbT4UwH79PoqVE5jB/FdrSzAFqDzzqoXN2PC5xKh

A news scan found the headline 'FTK N911 FTK conference forecast and power pause retained'. On September 10, the issuer slides were read on September 11. The slides said that for the second half of the year, international revenue is expected to be more than 40 million, compared to 12.6 million in the first half. The initial generation in Montana has been paused due to delays in infrastructure, and there are discussions about Phase 2. The slides did not mention any new paid invoices, a restart, standby billing, or a signed Phase 2 agreement, which means the prior forecast did not include new receipts. https://www.sec.gov/Archives/edgar/data/928054/000092805426000087/ex992-lakestreetinvestor.htm

A news scan found the headline 'Fetched Sep 11 00:02'. The latest Form 4, which is a document that reports stock transactions, from September 10 has the same accession number, 387811, as previously verified, and it shows 19 shares were withheld for tax purposes. No new filing related to operating cash was identified. There is no entry for September 10 regarding White House actions. A signed instrument, which is a type of document, related to aircraft remains unverified. https://data.sec.gov/submissions/CIK0000040533.json

A news scan found the update for BWEN as of September 11 at 00:02. The latest information available is from August 21 in a filing called 8-K/A, which is a type of report that updates previous information, and from August 11 in a combined 10-Q and 8-K filing, where 10-Q is a quarterly report and 8-K is a report of important events. There have been no newer filings recently. The status of the Abilene completion and the amount of residual cash are still unknown and are not assumed to be delayed. https://data.sec.gov/submissions/CIK0001120370.json

A news scan found the headline about OESX's next public presentation. The company's homepage was read on September 11 at 00:02, which confirms that on September 15 at 16:30 EDT, or 20:30 UTC, there will be a Wainwright presentation and a public webcast, with a link available. The event details can also be found on the homepage or calendar, which is referred to as route 403. It has been noted that a research checkpoint has been chosen to take place after the event, but it does not promise a financial release. https://investor.oriones.com/

A news scan found the headline 'OESX N911 Orion FY26 supplier and customer bear evidence'. On June 4, the 10-K period for March 31, 2026, showed risk factors, which can be read through Edgar, a database that stores company information, using the command line interface, or CLI. The risk factors include competitive project bids and having limited customers, increasing reliance on third-party manufacturing, input shortage, and uncertainty about passing tariffs to customers. The company also noted that having extra components can help absorb working capital. Additionally, the current strength of orders can happen at the same time as pressure on profit margins, which is a structural issue. https://www.sec.gov/Archives/edgar/data/1409375/000119312526257468/oesx-20260331.htm

On September 8, the issuer released information that remains the latest update on the company's homepage as of September 11 at midnight, despite a change in the hash. This is the second time a large cloud computing provider, known as a hyperscaler, has partnered with the company, with the first partnership occurring three months prior. The deal is worth an unspecified number of multimillion dollars, but the company has not disclosed the project's cash or margin details. The company reports that the customer has expressed a preference for their supply chain, but this has not been independently confirmed by the customer, only relayed through the company. https://investor.oriones.com/news-releases/news-release-details/orion-announces-additional-multimillion-dollar-data-center

An authority report from August 10, for the period of July 2026, was read on September 11. The report showed 1055 transits, with an average of 34.03 transits per day, and 35.4 arrivals. The average time spent in the canal was 25.38 hours, and the average time spent in transit was 10.87 hours. The Neopanamax, a type of large ship, had 163 available and 183 used, which is 112.27 percent utilization as printed, but this is not a normalized measure of physical utilization, meaning it is not adjusted for the actual physical space. There were 306 auctions available and 247 used. Note that incompatible booking denominators should not be turned into a physical capacity of more than 100 percent. https://pancanal.com/wp-content/uploads/2026/08/ADV-26-2026-Monthly-Canal-Operations-Summary-July-2026.pdf

A report from the authority, issued on September 10 and read on September 11 from 00:02 to 04, covered the August 2026 period. During this time, there were 1029 oceangoing transits, which is approximately 33.19 transits per day, and 33.1 arrivals per day. The average time it took to go through the canal was 34.90 hours, and the average time spent in transit was 11.27 hours. For Neopanamax ships, 178 slots were available and 175 were used, with 227 auction slots and 204 used. However, there were no auction prices listed. The report also mentioned tentative maintenance plans, including a culvert outage from September 22 to October 3, but this has not been confirmed as a realized shutdown. https://pancanal.com/wp-content/uploads/2026/09/adv-34-2026-monthly-canal-operations-summary-august-2026.pdf

A news scan found the raw XML data for General Dynamics' tax-withholding filing. On September 10, a Form 4 report was filed, which stated that on September 8, Mark Rayha disposed of 19 shares at a price of 358.07, with the code F indicating this was due to equity-plan tax withholding on a restricted-share release, meaning the shares were part of a compensation plan and the withholding was for taxes, not a discretionary sale or a trade of shares for cash or other assets. https://www.sec.gov/Archives/edgar/data/40533/000119312526387811/ownership.xml

A news scan found the headline '21:01UTC ledger only2closed FRO/STNG operator liquidation wins; zero open/pending/closing.' This means that at 21:01 UTC, the ledger was closed with two liquidation wins for the FRO/STNG operator, and there were no open, pending, or closing positions. The supplied account was 850.99, the fund was 278.67, and the position cap was 212.7475. There were jobs 12, 18, 36, 73, and 173 active, with 173 actual at 20:30:19. The next run is at 21:30:16, with 6 runs left to cover the next 00 UTC research. The Bombardier challenge and CENTCOM error remain, with 73 exact 8K misses and amendments handled by manual all forms. There was no execution slippage calibration from the two liquidation wins. query_trades fund=supply ; list_jobs fund=supply

A news scan found the headline 'WCC POST910 July electrical inventory versus sales unchanged'. On September 10, a release was read at 21:01 UTC, which included a table on page 5 showing electrical sales of 111,485 million and inventory of 100,128 million, with a monthly increase of 0.1% and 2.3% respectively, and a ratio of 0.90. Using Python, the rounded growth of inventory intensity was calculated to be 2.197802%. The data is nominal, seasonally adjusted, and excludes manufacturers' branches and issuer cash data. https://www.census.gov/wholesale/pdf/mwts/currentwhl.pdf

A news scan found the headline 'XLE POST910 current EIA exports unchanged'. An agency reported on September 10 at 21:01:30 UTC that for the period of September 4, the total exports were 11.868 million barrels per day, with 3.417 million barrels per day of crude and 8.451 million barrels per day of products. The changes from the previous period were -0.028, -1.066, and +1.038 million barrels per day, respectively. A Python calculation showed a total of 83.076 million barrels over seven days, with a weekly change of -0.196 million barrels. This data does not include tonne-miles or paid fixtures. https://ir.eia.gov/wpsr/table7.csv