Notes

The fund had 3 closed trades that were all liquidated by the operator, with no changes to the N3. The returns were -5.3923644805% for SNOW, +7.9399413074% for DELL, and -5.1304990470% for GTLB. The win rate was one third, with an average win of 7.9399413074% and an average loss of 5.2614317638%, resulting in a mean of -0.8609740734%. The total profit before overhead was -1.0460528968 USD. A calculation to estimate the optimal bet size, known as the Kelly criterion, which is a formula to determine the optimal size of a bet, resulted in 0. Due to the small sample size, it is not possible to estimate the effectiveness of the planned 4-6 week Post-Earnings Announcement Drift, or PEAD, strategy, which is a trading strategy based on the idea that a company's stock price will drift after an earnings announcement, or to determine slippage, which is the difference between the expected price of a trade and the actual price, or stop and target rates, which are the prices at which a trade will be automatically stopped or closed. query_trades fund=quant trade IDs9,7,1; python numpy recalculation Sep6 13:51UTC

On September 6 at 13:51 UTC, a review of quantitative trades showed 3 closed trades, 0 open trades, and 0 pending trades. Four monitors, numbered 46, 47, 48, and 42, are all set to run after the next scheduled run on September 7 at 00 UTC. The fund's cash balance is $245.69, and its equity is $850.99, which means it has a 25% capital limit of $212.75 and a 1% risk limit of $8.51. A check of the exchange calendar confirmed that September 7 is a Labor Day holiday, and the next regular trading day will be September 8. The next secondary headlines job is scheduled for 19 UTC, and another job is set for September 7 at 02:30 UTC, which previously showed all 19 countries as eligible with no issues. A job scheduled for 16:45 UTC is currently active. No monitored events require attention or rescheduling. Additionally, a social monitor, number 6, is not relevant as it belongs to a different, retired fund. This information comes from a trading eligibility and unattended interval audit. query_trades fund=quant; list_monitors fund=quant; list_jobs fund=quant; https://www.nyse.com/trade/hours-calendars

A review of the clinical watcher was done on September 6 at 13:51 UTC. The last scheduled job, job 35, ran at 12:45:35.965 UTC and finished at 12:45:40 UTC with no errors. The output from this job was parsed and found to be 3 valid lines of JSON data for the symbols INBX, TYRA, and IRD, with each line having 1 item and no error records. The total characters in the output were 1662, not including the final newline, which is 1 character less than the previous output of 1663 characters, which included a newline. All 3 existing notices were kept and not updated during this review. The job is set to run 15 more times, every 240 minutes, with the next run at 16:45:35.965 UTC, then at 20:45, and the next research run at 00 UTC. This review replaces a pending note, number 449, and observation 7296, and only includes the first matching keyword link, without assuming anything from the absence of news or checking the information independently. list_jobs fund=quant job35; query_research notes455; issuer URLs https://inhibrx.investorroom.com/news-releases https://tyrabio.investorroom.com/news-releases https://ir.opusgtx.com/press-releases

A news scan found the headline 'September 6 12:33 UTC local test: 1663 ASCII characters/bytes, 3 valid JSON lines, ordered INBX/TYRA/IRD, all 3 nonempty first matching items.' This is a test result, where ASCII refers to a character encoding standard. The test also showed that the length was less than 2000 characters. This test replaces a previous job, number 31, which had an issue with its output. A new job, number 35, is now active and runs every 240 minutes with a 60-second timeout. The scheduler has not yet completed its first run. The current test result does not guarantee the same length for future tests with longer titles, and the coverage remains limited to the first link only. local clinical_watch_test.jsonl; schedule_job35; list_jobs fund=quant

A news scan found the following information: Retrieved on September 6, 2026, at 12:33:11 UTC, the primary issuer-controlled listing, which may have promotional bias, had 3 matching clinical or results links. The first link was an existing webcast notice dated September 2, 2026. A compact scraper includes this issuer in its complete output. The web listing was inspected using the same source, but this was not an independent verification. No new clinical trial results were seen, but the first-page and keyword search cannot confirm that there is no important news. The existing monitor, which is dated, will be kept, and no new entry will be made. https://ir.opusgtx.com/press-releases

A news scan found the notice that the TYRA Clinical issuer listing was rechecked on September 6, 2026, at 12:33:11 UTC. This listing is controlled by the issuer, which means it may have promotional bias, or a tendency to present information in a favorable way. The scan found 3 matching links to clinical results, and the first one was a webcast notice that was already known, dated September 4, 2026. A tool that collects information, called a compact scraper, included this issuer in its complete output. The web listing was checked again using the same source, but this was not an independent check to confirm the information. No new clinical trial results were seen, but it is not possible to be sure that there is no important news just by looking at the first page and keywords. The existing monitor, which tracks updates, was kept as it was, and no new entry was made. https://tyrabio.investorroom.com/news-releases

A news scan found the notice that INBX Clinical issuer listing was rechecked on September 6, 2026, at 12:33:11 UTC. This was a primary listing controlled by the issuer, which may have promotional bias, meaning it could be influenced by the issuer's own interests. Three matching clinical or results links were found, and the first one was a webcast notice that already existed, dated September 4, 2026. A compact scraper, a tool that extracts data, included this issuer in its complete output. The web listing was checked again using the same source, but this was not an independent verification, meaning it did not come from a separate source. No new clinical trial results were seen, but it is not possible to be certain that there is no important news, as only the first page and certain keywords were checked. The existing monitor, which tracks updates, was kept as it was, and no new entry was made. https://inhibrx.investorroom.com/news-releases

A news scan found the headline that the US stock market will be closed on Monday, September 7, 2026, for Labor Day. Verified on September 6 at 12:32 UTC from the exchange calendar, the primary exchange source, the market will be closed on September 7 and will resume its regular session on September 8, with ordinary core hours of 09:30-16:00 ET. No quant monitor is due before the next research run on September 7 at 00 UTC; 4-armed monitors 46, 47, 48, and 42 are in place. The query_trades returned 3 closed trades and no pending or open trades, which is consistent with the user's all-cash state. https://www.nyse.com/trade/hours-calendars

A news scan found the headline 'Closed quant trade audit; operator-liquidated sample'. This is a report about a set of trades made by a computer program. The program made 3 trades and won 1 out of 3 times, which is a 33.33333333333333% win rate. On average, the program made 7.939941307427634 when it won and lost 5.261431763773078 when it lost. The average result was a loss of 0.8609740733728405. The total loss was 1.0460528968119984. The quarter Kelly criterion, a measure of risk, was 0. The trades were: a loss of 5.392364480519829 on SNOW, a gain of 7.939941307427634 on DELL, and a loss of 5.130499047026326 on GTLB. Returns are the change in price from when the trade started to when it ended. Profit and loss, or P&L, is the quantity of the trade times the price change, before any extra fees. All 3 trades were stopped by the operator, none were stopped by the strategy, and none reached their target. There is no information about the difference between the requested price and the actual price, so it is not possible to estimate slippage. This sample is not a fair test of the program's performance. query_trades fund=quant retrieved 2026-09-06T12:32Z

A data scan found two sets of stock market data. The first set has all adjustments and the second set has raw data. Each set has 2 pages and includes 7 symbols: BIL, DBC, EEM, EFA, GLD, SPY, and TLT, with 2684 daily bars each. The data covers the period from January 4, 2016, to September 4, 2026, with aligned unique dates and positive valid open, high, low, and close prices, and no series has a zero range, which means there is always some price movement, and this movement is never zero percent. The source of the data is https://runway.adityaperswal.workers.dev/data/stock-bars. Note that the adjustment flags are not from independent providers, and there are gaps in the data due to historical revisions and dividend and cash timing, which refers to the timing of when dividends and cash are paid out. https://runway.adityaperswal.workers.dev/data/stock-bars

A technical analysis of DBC QTS90608 on September 4 found the following data: the close price was 31.9, the 50-day simple moving average, or sma50, was 29.3594, and the 200-day simple moving average, or sma200, was 27.2321. The average true range, or atr20, over 20 days was 0.4209, and the average daily range, or adr20, was 0.3217. The median dollar volume over 20 days was 17,293,815.51. The whole shares capital is 6. An illustrative stop was set at 31.4791 and an illustrative target at 32.5314. The risk-limited quantity is 6. From August 12, the excess percentage was 40.4371. The data includes USD prices, average true range, or ATR, average daily range, or ADR, and volume in USD, as well as excess momentum percentage. The ATR includes the completed bar on September 4, which differs from the previous run's lagged ATR. The illustrative barriers are not orders, and the monthly signal is held constant, with an actual quantity of 0. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline 'GLD QTS90608 September 4 technical feasibility'. The close price of GLD is 406.77. The simple moving average over 50 days, or sma50, is 388.88. The simple moving average over 200 days, or sma200, is 415.44. The average true range over 20 days, or atr20, which measures price volatility, is 7.71. The average daily range over 20 days, or adr20, is 5.53. The median dollar volume over 20 days is 4,357,485,495.70. The whole shares capital is 0. An illustrative stop, which is a theoretical price level, is 399.06. An illustrative target, which is another theoretical price level, is 418.34. The risk limited quantity is 0. The August 12 month excess percentage is 23.81. This data includes USD prices, average true range, or ATR, and average daily range, or ADR, as well as volume in USD and excess momentum percentage. The average true range, or ATR, includes the completed bar on September 4 and differs from the previous run's lagged ATR. The illustrative barriers are not orders and the monthly signal is held constant, with an actual quantity of 0. https://runway.adityaperswal.workers.dev/data/stock-bars

The TLT QTS90608 had a close price of 82.21 on September 4. The 50-day simple moving average, or sma50, was 83.41, and the 200-day simple moving average, or sma200, was 86.25. The average true range, or atr20, which measures price volatility, was 0.66, and the average daily range, or adr20, was 0.45. The median dollar volume over 20 days was 2,161,076,035.08. The whole shares capital is 2. An illustrative stop was set at 81.55 and an illustrative target at 83.20. The risk limited quantity is 2. On August 12, the excess percentage was -3.91. The prices are in USD, and the average true range includes the completed bar on September 4, which differs from the previous run's lagged average true range. The illustrative barriers are not orders, and the monthly signal was held constant, with an actual quantity of 0. https://runway.adityaperswal.workers.dev/data/stock-bars

A technical analysis of EEM on September 4 found the following: the closing price was 68.7, the 50-day simple moving average, or sma50, was 65.8026, and the 200-day simple moving average, or sma200, was 61.741350000000004. The average true range, or atr, over 20 days, which measures price volatility, was 0.923889999999998. The average daily range, or adr, over 20 days, was 0.6696399999999997. The median dollar volume over 20 days was 1085458828.825. The whole shares capital is 3. An illustrative stop, or potential stop-loss price, was 67.77611, and an illustrative target, or potential sell price, was 70.085835. The risk-limited quantity is 3. On August 12, the excess percentage, which measures how much the price moved compared to usual, was 32.150743809642734. The analysis includes USD prices, average true range, and average daily range, as well as volume in USD and excess momentum as a percentage. The average true range includes the completed bar on September 4, which differs from the previous run's lagged average true range. The illustrative barriers are not actual orders, and the monthly signal was held constant, with an actual quantity of 0. https://runway.adityaperswal.workers.dev/data/stock-bars

The technical feasibility of EFA QTS90608 on September 4 is as follows: the closing price was 108.35, the 50-day simple moving average, or sma50, was 105.873, and the 200-day simple moving average, or sma200, was 101.5884. The average true range, or atr20, which measures price volatility, was 0.7829, and the average daily range, or adr20, was 0.6326. The median dollar volume over 20 days was 901668846.015. The whole shares capital is 1. An illustrative stop, which is a theoretical price level to limit losses, was 107.5671, and an illustrative target, which is a theoretical price level to take profits, was 109.5243. The risk-limited quantity is 1. From August 12, the excess momentum percentage was 17.1339. The average true range includes the completed bar on September 4 and differs from the previous run's lagged average true range. The illustrative barriers are not actual orders, and the monthly signal was held constant, with an actual quantity of 0. https://runway.adityaperswal.workers.dev/data/stock-bars

A technical analysis of the SPY fund on September 4 found the following data: the closing price was 770.19, the 50-day simple moving average, or sma50, was 756.8572, and the 200-day simple moving average, or sma200, was 712.1529. The average true range, or atr20, over 20 days was 5.236859999999979, which is a measure of price volatility. The average daily range, or adr20, over 20 days was 4.4095599999999875. The median dollar volume over 20 days was 28041884838.015. The whole shares capital is 0. An illustrative stop, which is a theoretical price level, was 764.9531400000001, and an illustrative target, which is another theoretical price level, was 778.04529. The risk limited quantity is 0. On August 12, the excess percentage, which is a measure of how much the fund's return exceeded a benchmark, was 15.93406932713981. The average true range includes the completed bar on September 4 and differs from the previous run's lagged average true range. The illustrative barriers are not actual orders, and the monthly signal has been held constant, with an actual quantity of 0. https://runway.adityaperswal.workers.dev/data/stock-bars

A news scan found the headline 'Issuer Investor Agenda 2026' retrieved on September 6, which confirms that Inditex, ticker ITX.MC, will release its half-year results on September 9, with no specific hour given. This event is already covered by an existing catalyst monitor dated September 19. The impact of foreign share revenue and foreign exchange eligibility on this event is unverified and this information is provided for event awareness only and is not a proposed trade. https://www.inditex.com/itxcomweb/es/en/investors/finance

A 2025 Form 10K was published, which was retrieved on September 6. The document states that there was a change in the index methodology, effective November 10, 2025. It explicitly warns that older performance could differ under the new methodology. The sector cap is 40% and the commodity cap is 12.5% at the weight setting. For example, the actual weight of gold on December 31 was 13.09%, which can drift. The management fee is 0.85% annual, and the affiliate fee waiver is separate. The current Invesco solution page lists a gross fee of 0.89% and a net fee of 0.84%, which is not a contradiction, but rather a different fee concept and period. The futures roll selection seeks a favorable implied roll yield, and collateral income contributes to the returns, which are not the same as spot commodity returns. https://www.sec.gov/Archives/edgar/data/1328237/000119312526083563/dbc-20251231.htm

A news scan found the schedule verified on August 6, 2025, for the Bank of Canada's interest rate decision, which was scheduled for September 2, not September 9, with the next decision on October 28. The current key rate page states that on September 2, the rate announcement was 2.25%, which is outside the next two-session catalyst window, a term that refers to a period of time when important events can affect the market. https://www.bankofcanada.ca/2025/08/bank-canada-publishes-2026-schedule-policy-interest-rate-announcements-other-major-publications/

A news scan found the headline is not available, but a data line shows that as of September 3, the TLT issuer's duration was 15.05 years and its SEC yield was 5.17%. The expense was 0.15%. On September 1, the ex-date was set, and on September 4, a distribution of 0.314686 was paid. The median 30-day bid-ask spread, which is the difference between the prices at which people are willing to buy and sell, was 0.01% as of September 4. The page fields were not updated at the same time, showing a September 3 close of 82.07 and a September 4 NAV, or net asset value, of 82.17, so there was no opportunity to make a trade based on the difference between the two prices. If the yield were to rise by 25 basis points, or 0.25%, in a parallel shift, the price would likely fall by approximately 3.7625%, not taking into account convexity, which refers to the change in duration, and carry, which refers to the cost of holding a security; this is a hypothetical scenario, not a forecast. https://www.ishares.com/us/products/239454/ishares-20-year-treasury-bond-etf?fundSearch=true&qt=TLT

A research paper by Moskowitz, Ooi, and Pedersen from 2012, published in the Journal of Financial Economics, is the primary source for this trend strategy. The paper, which is 23 pages long, studied 58 futures and forwards contracts from 1985 to 2009. It found that the excess return from the past 12 months could predict the return for the next month, with a long position for positive signs and a short position for negative signs. The study used a 40% individual ex-ante volatility scale, which resulted in a 12% aggregate sample annual volatility. Our adaptation uses 6 cash ETFs, does not include short positions or volatility scaling, and uses BIL as a comparator. We also use next open monthly cohorts. This is an adaptation of the original strategy, not an exact replication, and the authors are affiliated with AQR, which is the same lineage as our source, and not independent sources. https://pages.stern.nyu.edu/~lpederse/papers/TimeSeriesMomentum.pdf

A news scan found the headline about an FDA search on September 8-9. The FDA held a biocompatibility town hall on September 9 from 13-14 ET, which discussed ISO 10993-1 risk assessment, but did not make a product decision. The FDA and PDUFA searches for specific dates did not find a confirmed product decision on September 8 or 9, which is a gap in coverage and not proof that none was made. A secondary search found three issuer clinical readouts, but a future-dated search result from the AACR on September 8 was excluded as of September 6. https://www.fda.gov/medical-devices/medical-devices-news-and-events/town-hall-biocompatibility-risk-assessment-09092026

The issuer confirmed on September 2 that there will be an investor call on September 9 at 8:00 ET or 12:00 UTC to discuss the topline results of Phase 1 and 2, cohort 1. This call will be followed by a full analysis at medical meetings. The early cohort results do not indicate approval, with the date of observation being September 6. https://ir.opusgtx.com/press-releases/detail/542/opus-genetics-to-present-opgx-best1-cohort-1-clinical-data-in-investor-webcast-and-at-upcoming-leading-ophthalmology-conferences

A news scan found the headline that on September 9, there will be a webcast to discuss the initial results of the SURF302 Phase 2 trial of dabogratinib for IRNMIBC, which was confirmed by the issuer on September 4. The discussion will take place at 8 ET or 12 UTC. This is not an FDA decision, and the first press release will be released at the same time, although the information in it has not been confirmed yet. https://tyrabio.investorroom.com/2026-09-04-Tyra-Biosciences-to-Host-Conference-Call-on-Initial-Results-of-Oral-Dabogratinib-from-the-SURF302-Phase-2-Study-for-IR-NMIBC-on-September-9%2C-2026

The issuer confirmed that on September 8, there will be a webcast at 5:00 PM PT, 8:00 PM ET, or 12:00 UTC to discuss the Phase 2 clinical readout of INBRX-106, also known as HexAgon, for the treatment of first-line head and neck squamous cell carcinoma, which is a type of cancer. The webcast will focus on the primary endpoint, meaning the main goal of the study. This is not related to a PDUFA, which stands for Prescription Drug User Fee Act, or approval. It is possible that the results will be published before the call, and this information is for research purposes only, with no estimated probability of success. https://inhibrx.investorroom.com/2026-09-04-Inhibrx-To-Host-Webcast-Presentation-to-Provide-Primary-Endpoint-Results-for-INBRX-106-Phase-2-First-Line-HNSCC-Study