Notes

A review of the automated trading fund's activity from September 2 to September 4 was conducted using Python and numpy to directly audit all three closed lots. The results showed that MIRM had a return of -0.4393213%, a profit and loss of -0.2855149, and an R value of -0.0698270. IONS had a return of -2.8424202%, a profit and loss of -1.8472889, and an R value of -0.4110577. RARE had a return of -45.8600619%, a profit and loss of -29.8044542, and an R value of -6.4403478. The average loss for these equal trades was 16.3806011%, with no wins out of three trades. There were two operator liquidations and one stop. The RARE exit was below the 24 stop by 10.01024 per share, which was $25.1766833 worse than the theoretical stop. If the observed 45.8601% loss is applied to the 49.14 fund, it would result in a loss of 22.5356344 USD, compared to the account's 1% loss of 8.5099. This scenario is not expected to continue into the future. The fund's current cash is 49.14, with an equity of 850.99, and there are no open or queued trades. There is a shortfall of 349.01 in fund cash, which is 710.2361% of the fund's cash, and this is not a target return or permission to borrow cash from other funds. query_trades fund=catalyst and supplied venue account context, 2026-09-06T08:00:43Z

A news scan found the headline 'Sep 6 06UTC actual fund edge audit and cash state'. The audit of closed trades on September 6 shows two trades: one for FRO, which went from 44.49 to 46.18, with a quantity of 2.584625758, a return of 3.80516296%, and a profit and loss of 4.37555688 USD, and another for STNG, which went from 79.76 to 80.36, with a quantity of 1.00288365, a return of 0.75225677%, and a profit and loss of 0.60173019 USD. The total profit and loss is 4.97728707 USD, which rounds to 4.98 USD. The average win is 2.27870986%, with 2 wins and 0 losses, so the average loss is undefined. Both trades were liquidated by the operator on September 4, and were not the result of a stop or target being reached. The trades had common war exposure and were not independent. There is no reference to the quote at the time of the order, so it is not possible to estimate the execution slippage. The user's context is a flat or cash position of 278.67 USD in the fund, with an account balance of 850.99 USD, a capital of 212.7475 USD, and a nominal 1% risk of 8.5099 USD. There are no held positions, pending orders, or due supply to monitor, and the social monitor shows 6, which is not ours. The Kelly criterion, a measure of bet size, is undefined based on the actual fund record, with a size of 0. query_trades fund=supply IDs 6,3

A review of provider documentation on September 6, 2026, verified that extended-hours trades can update the minute high, low, open, and close prices, which are also known as OHLC, but daily bars only update the volume. This means that daily endpoints do not show the full overnight stop path. A rule was noted at the time of research retrieval, but it is not certain if the specification has always been the same historically. https://docs.alpaca.markets/us/docs/market-data-faq

A news scan found the headline about the May 18, 2022 issuer prospectus, which is a document that describes the details of an investment. The prospectus says the annual management fee is 0.55%, which is a fee paid to the company that manages the investment. This fee does not include the costs of buying and selling things in the portfolio, which is a collection of investments, or the costs of using derivatives, which are financial contracts. The prospectus also talks about the historical objective of the NSPY and NIWM investments, which had an exposure, or a level of investment, that was close-to-open, meaning it was based on the prices at the start and end of the trading day. Just because the prospectus exists, it does not mean that the investment is currently eligible, or available, for purchase. https://www.sec.gov/Archives/edgar/data/1199046/000158064222002817/unifiedseriestrust_497.htm

A news scan found the following data line: the key is all/SPY/later/lagged_negative/10, which means this is a set of metrics for the SPY symbol with a specific rule, where lagged_negative refers to a trading strategy based on negative lagged indicators, which are indicators that use past data to predict future market movements. The adjustment is all, the symbol is SPY, the split is later, the rule is lagged_negative, and the cost is 10 basis points, which is a unit of measurement for costs, where 1 basis point equals 0.01%. The number of trades is 652, the return percentage is -62.51826604233749%, the drawdown percentage is 62.9763913022138%, the win rate percentage is 40.644171779141104%, the average win percentage is 0.42090283915557786%, and the average loss percentage is 0.5375564840492398%. The mean percentage is -0.14799863029268048%, and the quarter Kelly fraction is 0, which is a measure used to determine the optimal bet size. The first trade was on 2021-01-05 and the last exit was on 2026-08-27. The mean block confidence interval percentage is between -0.19624990153619945% and -0.10356644927065436%, the endpoint below stop percentage is 2.7607361963190185%, and the endpoint above target percentage is 0.6134969325153374%. These metrics are diagnostic endpoints and have not been validated for venue stop and target fills, and the full reproduction can be found in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A data scan found the following metrics for the SPY fund: the adjustment is all, the symbol is SPY, the split is later, and the rule is unconditional, which means it applies without conditions. The cost is 10 basis points, which is a unit of measurement for costs, and the number of trades is 1420. The return percentage is -90.82795864792173, the drawdown percentage is 90.82795864792239, the win rate percentage is 37.11267605633803, the average win percentage is 0.38098952495020744, and the average loss percentage is 0.48901175066470415. The mean percentage is -0.16613099555973346, and the quarter Kelly fraction, a measure of risk, is 0. The first trade was on 2021-01-04 and the last exit was on 2026-08-31. The mean block confidence interval percentage is between -0.19778698530549468 and -0.13505907702353667, the endpoint below stop percentage is 2.2535211267605635, and the endpoint above target percentage is 0.49295774647887325. These metrics are diagnostic and the endpoints are not validated venue stop or target fills. The full data is available in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the following metrics for IWM: the return percentage was -63.958240504263664, the drawdown percentage was 66.03737394062739, and the win rate percentage was 42.616033755274266. The average win percentage was 0.5492654547012883 and the average loss percentage was 0.6516928737041563. The mean percentage was -0.13989206708411456. These metrics are based on data from January 4, 2021, to August 27, 2026. The cost was 10 basis points, which is a measure of the cost of a trade, expressed as a fraction of a percentage point. The quarter Kelly fraction, a measure used to determine the optimal bet size, was 0. The mean block confidence interval percentage was between -0.19686863094923798 and -0.08214481131464536. The endpoint below stop percentage was 1.4064697609001406 and the endpoint above target percentage was 0.42194092827004215. These metrics are diagnostic endpoints and have not been validated as actual stop or target fills. The full data can be found in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A data scan found the following metrics for IWM: the adjustment is all, the symbol is IWM, the split is later, and the rule is unconditional, which means a condition that is always true. The cost is 10 basis points, which is a unit of measurement for costs, equal to 0.01%. The number of items is 1420. The return percentage is -88.04402474255946%, the drawdown percentage is 88.70301134323483%, the win rate percentage is 42.816901408450704%, the average win percentage is 0.5477205899520025%, and the average loss percentage is 0.6652157031873501%. The mean percentage is -0.14587396640655695%. The quarter Kelly fraction, a measure of investment size, is 0. The first date is 2021-01-04 and the last exit date is 2026-08-31. The mean block confidence interval percentage is between -0.18956859085943498% and -0.10204244025267112%. The endpoint below stop percentage is 1.4084507042253522% and the endpoint above target percentage is 0.49295774647887325%. These metrics are diagnostic endpoints and have not been validated against actual venue stop and target fills. The full reproduction is available in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the following data line: the key is all/EFA/later/lagged_negative/10, which means this is a set of metrics for the EFA symbol, using a later split and a lagged negative rule, with a cost of 10 basis points. The results are based on 631 events, with a return of -78.64969250341667 percent, a drawdown of 78.98879264913583 percent, and a win rate of 38.51030110935024 percent. The average win was 0.4976719590999809 percent and the average loss was 0.7035957047173002 percent. The mean return was -0.24098391025200805 percent. The quarter Kelly fraction is 0, which is a measure of the optimal bet size. The first event was on 2021-01-04 and the last exit was on 2026-08-31. The mean block confidence interval is between -0.3096025984574118 and -0.17506289187011287 percent. The endpoint below stop percentage is 8.240887480190175 percent and the endpoint above target percentage is 1.4263074484944533 percent. These metrics are diagnostic endpoints and have not been validated for actual stop and target fills on a trading venue. The full reproduction of these metrics can be found in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A data line found the following metrics for EFA: the adjustment is all, the symbol is EFA, the split is later, and the rule is unconditional. The cost is 10 basis points. There were 1420 events. The return percentage was -94.21296658592856, the drawdown percentage was 94.31557132915188, and the win rate percentage was 38.80281690140845. The average win percentage was 0.5093843100593881 and the average loss percentage was 0.6453991653057801. The mean percentage was -0.19731064775211277. The quarter Kelly fraction was 0. The first event was on 2021-01-04 and the last exit was on 2026-08-31. The mean block confidence interval percentages were -0.2382358245093248, -0.19718379453554197, and -0.15793349253333994. The endpoint below stop percentage was 7.042253521126761 and the endpoint above target percentage was 1.056338028169014. These metrics are in units of percent, except for counts and costs, which are in basis points. They are diagnostic endpoints and not validated venue stop or target fills. The full reproduction is in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline 'EEM Overnight later lagged_negative final metrics'. The given data is a set of metrics for the symbol EEM, which is a stock ticker. The adjustment is all, the split is later, and the rule is lagged negative, meaning a strategy that waits for a negative trend to emerge before acting. The cost is 10 basis points, which is a unit of measurement for costs, equal to 0.01%. The metrics include: a return percentage of -79.80553832682853%, a drawdown percentage of 82.28519604465909%, a win rate percentage of 40.73506891271057%, an average win percentage of 0.7002782764076565%, and an average loss percentage of 0.8849350946967326%. The mean percentage is -0.23919733556385744%, and the quarter Kelly fraction is 0. The first date is 2021-01-04, and the last exit date is 2026-08-28. Other metrics include a mean block confidence interval percentage, endpoint below stop percentage of 8.728943338437979%, and endpoint above target percentage of 2.1439509954058193%. The data is part of the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A data scan found the following metrics for EEM: the adjustment is all, the symbol is EEM, the split is later, and the rule is unconditional, which means a condition that is always true. The cost is 10 basis points, which is a unit of measurement for costs, equal to 0.01%. The number of trades is 1420, the return percentage is -94.2317785750696, the drawdown percentage, which is the maximum loss from a peak, is 94.76020886613621, the win rate percentage is 41.12676056338028, the average win percentage is 0.6833226812314862, and the average loss percentage is 0.8093470915405144. The mean percentage, which is the average return, is -0.19546036809062126, and the quarter Kelly fraction, a measure of risk, is 0. The first trade was on 2021-01-04 and the last exit was on 2026-08-31. The mean block confidence interval percentage, which is a measure of uncertainty, is between -0.2472967394604112 and -0.1451539235243315. The endpoint below stop percentage, which is the percentage of trades that ended below a stop, is 6.971830985915493, and the endpoint above target percentage, which is the percentage of trades that ended above a target, is 1.4788732394366197. These metrics are diagnostic and the endpoints are not validated venue stop or target fills. The full reproduction is in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline 'QQQ Overnight later lagged_negative final metrics'. The QQQ fund had a return percentage of -60.46964239786383, a drawdown percentage of 61.989078815168305, and a win rate percentage of 42.70516717325228. The average win percentage was 0.6159441284507898 and the average loss percentage was 0.6978791257083992. The mean percentage was -0.13680870865865433. The quarter Kelly fraction, a measure of risk, was 0. The first trade was on 2021-01-05 and the last exit was on 2026-08-27. The cost was 10 basis points, which is a unit of measurement for costs, equal to 0.01 percent. The mean block confidence interval percentage was between -0.1973419171175598 and -0.07865702865288977. The endpoint below stop percentage was 1.6717325227963524 and the endpoint above target percentage was 0.7598784194528876. These endpoints are for diagnostic purposes only and do not reflect actual stop or target fills. The full data is available in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A data scan found the following metrics for QQQ: the adjustment is all, the symbol is QQQ, the split is later, and the rule is unconditional, which means a condition that is always true. The cost is 10 basis points, which is a unit of measurement for costs, equal to 0.01%. The number of trades is 1420, the return percentage is -89.71%, the drawdown percentage is 89.93%, the win rate percentage is 40.49%, the average win percentage is 0.55%, and the average loss percentage is 0.64%. The mean percentage is -0.16%, the quarter Kelly fraction is 0, the first trade was on 2021-01-04, and the last exit was on 2026-08-31. The mean block confidence interval percentage is between -0.20% and -0.12%, the endpoint below stop percentage is 1.83%, and the endpoint above target percentage is 0.49%. These metrics are diagnostic endpoints and have not been validated for venue stop or target fills. The full reproduction can be found in the QON90606 archive. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline 'SPY QON90606 archive part0390'. This is a Base64 encoded JSON file map, which is a type of encoded data, and it needs to be concatenated with other ordered parts to be decoded. The scripts and every computed result or ledger are included, and a manifest follows. The payload is a long string of encoded characters. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline is not available, but a Base64 encoded JSON file map was found. The file contains a payload with a large amount of encoded data. The data appears to be a collection of scripts and computed results, followed by a manifest. The payload is encoded in Base64 and needs to be concatenated with other ordered parts to be decoded. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline or data line that appears to be a Base64 encoded string. The text includes a mention of an XZ JSON file map and a payload, but the content is not readable without decoding. The provided text is a long string of characters that does not contain any understandable information without further processing. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline 'SOL/USD Verified 06:15 funding collector open-interest'. A sandbox job executed on 2026-09-06 at 06:15:31.193Z. The job read data at 06:16 from a direct exchange. The forecast did not realize funding, which means the funding was not actually collected. Open interest, or the total number of outstanding contracts, was unsigned and had no direction, meaning it did not indicate a specific trend. The funding rule was not changed because it failed. The data included the symbol SOL, with an endpoint of open-interest, and was marked as eligible 0. The timestamp was 1788675334076, with a local age of 0.074 seconds and an exchange age of 0.147 seconds. The data was marked as fresh. The open interest was 3048824.86999999975 contracts, 3048824.86999999975 coins, and 323724224.696599973455 USD. https://www.okx.com/api/v5/public/open-interest

A news scan found the headline 'SOL/USD Verified 06:15 funding collector funding-rate'. This is a data line from a computer system. It says that on September 6, 2026, at 06:15:31, a job was done. The system looked at data from a direct exchange at 06:16. The forecast for funding was not realized, which means it did not happen as predicted. The system uses something called 'open interest' or OI, which is a measure of the number of contracts that have not been settled, and it was unsigned and had no direction. A funding rule failed and did not change. The system provided more details in a format that includes the symbol 'SOL', an endpoint for funding rates, and other technical information like timestamps and ages of data in seconds. https://www.okx.com/api/v5/public/funding-rate

A news scan found the headline is not available, but a Base64 encoded JSON file map was found, which includes a payload. The payload is a long string of characters that appears to be a coded message or data. It is recommended to concatenate the ordered parts of the JSON file map to decode the message. The manifest follows the payload. https://runway.adityaperswal.workers.dev/data/stock-bars