Notes

A news scan found the headline 'SOL/USD C91104 largest native versus aggregated l discrepancies'. Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. This data compares the native and aggregate values of SOL over full 24-hour UTC days. The largest 3 absolute relative differences are listed below, without claiming any cause. The differences are as follows: on December 2, 2025, the native value was 138.327 and the aggregate value was 124.3, with a difference of 11.284794851166534 percent; on November 23, 2025, the native value was 130.545 and the aggregate value was 127.4, with a difference of 2.4686028257456583 percent; on December 8, 2025, the native value was 132.47 and the aggregate value was 131.545, with a difference of 0.7031814208065867 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline about SOL/USD C91104 largest native versus aggregated h discrepancies. Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. The data compares native and aggregate values on full 24-hour UTC days, showing the largest 3 absolute relative differences. The differences are as follows: on November 23, 2025, the native value was 133.445 and the aggregate value was 142.4051776645, a difference of -6.29203081759413 percent; on December 8, 2025, the native value was 136.565 and the aggregate value was 139.3, a difference of -1.9633883704235577 percent; on December 14, 2025, the native value was 133.1842088665 and the aggregate value was 133.8, a difference of -0.46023253624813965 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline about the largest native versus aggregated open discrepancies for SOL/USD from 2024 to 2026, retrieved on September 11, 2024, at 04 UTC. The comparison is based on full 24-hour UTC days and shows the largest 3 absolute relative differences, without claiming any cause. The discrepancies are as follows: on December 2, 2025, the native price was 139.716 and the aggregate price was 126.67, with a difference of 10.299202652561789 percent; on November 23, 2025, the native price was 131.95 and the aggregate price was 127.5445, with a difference of 3.454088573007841 percent; and on December 3, 2025, the native price was 142.18 and the aggregate price was 138.72, with a difference of 2.4942329873125724 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the data line: Observed Sep 11 04UTC; it shows an audit of SOL/USD prices for the years 2024, 2025, and 2026. The audit checked full 24-hour days only, with years separated and no imputation. The results are as follows: for 2024, there were 127 matched days, with 126 days having some difference in open, high, low, or close prices, no difference in open prices, and 126 differences in close prices. For 2025, there were 363 matched days, with 203 days having some difference in open, high, low, or close prices, 11 differences in open prices, and 193 differences in close prices. For 2026, there were 252 matched days, with no differences in open, high, low, or close prices. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline 'ETH/USD C91104 largest native versus aggregated c discrepancies'. Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. This data compares the native and aggregate values of ETH over 24-hour periods. The three largest absolute relative differences are listed: on March 31, 2025, the native value was 1823.594 and the aggregate value was 1821.9125, a difference of 0.09229312604202011 percent; on June 1, 2025, the native value was 2537.955 and the aggregate value was 2540.245, a difference of -0.09014878486129918 percent; and on February 24, 2025, the native value was 2515.6035 and the aggregate value was 2513.4, a difference of 0.08767008832657464 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data: Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. The data compares native and aggregate values for ETH/USD over full 24-hour days. The three largest absolute relative differences are listed. On December 2, 2025, the native value was 2975.05 and the aggregate value was 2783.3485, a difference of 6.887441511546255 percent. On December 3, 2025, the native value was 3033.2 and the aggregate value was 2986.247, a difference of 1.572307983900867 percent. On December 8, 2025, the native value was 3078.7395 and the aggregate value was 3042.72, a difference of 1.183792790661009 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data: Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. This data compares native and aggregate values for ETH/USD over full 24-hour periods. The three largest absolute relative differences are listed below. A list of discrepancies was found, including: on December 8, 2025, the native value was 3152.7225 and the aggregate value was 3180.15, with a difference of -0.8624593179567097 percent; on December 14, 2025, the native value was 3104.4 and the aggregate value was 3130.1675, with a difference of -0.8231987585328837 percent; and on November 23, 2025, the native value was 2844.349 and the aggregate value was 2855.7525, with a difference of -0.39931681754633574 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data: Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. The data compares native and aggregate values for ETH/USD over full 24-hour UTC days. The largest 3 absolute relative differences are listed, without claiming a cause. The discrepancies are as follows: on December 2, 2025, the native value was 3020.1375 and the aggregate value was 2798.84, a difference of 7.906757799659836 percent; on November 23, 2025, the native value was 2824.571 and the aggregate value was 2767.86, a difference of 2.048911433381728 percent; on December 3, 2025, the native value was 3053.4795 and the aggregate value was 2995.8575, a difference of 1.9233892132719976 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data line: Observed Sep 11 04 UTC, which is a record of observations made on September 11 at 04 UTC. The data shows an audit of ETH/USD yearly matched OHLC, which stands for Open, High, Low, Close, a set of numbers that describe the price of a stock or currency over a certain period of time. The audit covers full 24-hour days only, with years separated and no imputation, meaning no missing values were filled in. The audit results are as follows: for the symbol ETH in the year 2024, there were 365 matched days, with 365 days showing a difference in OHLC, 0 days showing a difference in open price, and 365 days showing a difference in close price. In 2025, there were 364 matched days, with 203 days showing a difference in OHLC, 10 days showing a difference in open price, and 194 days showing a difference in close price. In 2026, there were 252 matched days, with 0 days showing a difference in OHLC, 0 days showing a difference in open price, and 0 days showing a difference in close price. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data: Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. This data compares native and aggregate values for BTC/USD over full 24-hour days. The three largest absolute relative differences are listed, without claiming any cause. The differences are as follows: on January 15, 2025, the native value was 100416.199 and the aggregate value was 100511.3465, with a difference of -0.09466344180356634 percent; on April 4, 2024, the native value was 68592.7 and the aggregate value was 68529.9035, with a difference of 0.09163372016129934 percent; on November 19, 2024, the native value was 92345.65 and the aggregate value was 92424.43, with a difference of -0.08523720405957391 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A historical review from 2024 to 2026, retrieved on September 11, 2024, at 04 UTC, compared native and aggregate data for BTC/USD on full 24-hour UTC days. The largest three absolute relative differences were found on the following days: the difference on December 2, 2025, was 3.986148185381033 percent, with a native value of 89602.3682237864 and an aggregate value of 86167.6; the difference on November 23, 2025, was 2.4489796834810784 percent, with a native value of 86694.1 and an aggregate value of 84621.731; and the difference on December 8, 2025, was 0.16528944742564278 percent, with a native value of 89692.422 and an aggregate value of 89544.4145320195. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data: Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. This data compares the native and aggregate values for BTC/USD over full 24-hour UTC days. The largest 3 absolute relative differences are listed, without claiming any cause for these differences. The discrepancies are as follows: on November 23, 2025, the native value was 88127.942 and the aggregate value was 90165.3625169692, a difference of -2.2596487831851753 percent; on December 8, 2025, the native value was 91558.1005726285 and the aggregate value was 92306.181, a difference of -0.8104337318120658 percent; on December 14, 2025, the native value was 90240.3049507385 and the aggregate value was 90535.8714532015, a difference of -0.3264634202099437 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data: Historical data from 2024 to 2026 was retrieved on September 11 at 04 UTC. This data compares the native and aggregate prices of Bitcoin (a type of digital currency) over full 24-hour periods. The largest three absolute relative differences are listed below, without claiming any cause for these differences. The data shows the following discrepancies: on December 2, 2025, the native price was 91625.132 and the aggregate price was 86270.82, a difference of 6.2064 percent; on November 23, 2025, the native price was 87421.966 and the aggregate price was 84698.415, a difference of 3.2156 percent; on December 14, 2025, the native price was 88793.16 and the aggregate price was 90323.263, a difference of -1.6940 percent. https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the following data line: Observed on September 11 at 04 UTC, the results of a yearly audit of BTC/USD prices for the years 2024, 2025, and 2026 are as follows: the audit considered full 24-hour days only, with years separated and no missing data filled in. The audit results are: [{"symbol":"BTC","year":2024,"matched_days":365,"ohlc_any_difference":361,"open_difference":0,"close_difference":361},{"symbol":"BTC","year":2025,"matched_days":364,"ohlc_any_difference":203,"open_difference":12,"close_difference":193},{"symbol":"BTC","year":2026,"matched_days":252,"ohlc_any_difference":5,"open_difference":0,"close_difference":5}] https://runway.adityaperswal.workers.dev/data/crypto-bars

A news scan found the headline about a query on September 11. The query was about trades from a fund called fundconvex. The results showed that there were no live closed trades, so there was no observed win rate or slippage. A test was done using 12 events to see what would have happened if certain trades had been made. This test was called a counterfactual intrinsic-only test. It showed that for call trades, there were 3 wins and 9 losses, with a mean net result of -0.251973R, an average win of +1.384390R, and an average loss of -0.797427R. For put trades, there were 2 wins and 10 losses, with a mean result of -0.739913R, an average win of +0.330379R, and an average loss of -0.953971R. The term 1R refers to the full premium. The results are not the actual profits or losses from option trades. The time left until the trades expire could improve the results, but the costs of making the trades could make them worse. The chronological holdout was isolated before the computations were done, but the candidate was selected today, so it is not the same as a genuinely prospective trading series. A counterfactual is a what-if scenario, and intrinsic-only refers to a type of trade that is based on the underlying value of an asset. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://runway.adityaperswal.workers.dev/data/option-bars ; Convex query_trades

A news scan found the headline 'TXNM REV9110400 due248 primary and collector audit'. The audit report for September 11, 04:08, shows a direct 403 error, and the web open recovered the latest acquisition index from August 18 to September 9, but actual responses from September 9 are not visible in the index, which does not prove non-filing or approval. The FTK homepage 200 still links to the Lake Street presentation, but paid collections and restart have not been revalidated. The AVAV node 21391200 shows that the September 11 real-time hours open is still in the future and frozen at 229, unchanged. The last actual END recovered note 2151 on September 11 at 02:30:28, with three panels, had a prior time of 01:45, which is a 45-minute cadence. Job 203's last note 2167 was truncated during TXNM, with no END, and had a 30-minute receipt gap with no full coverage at 03:15. A compact replacement test was conducted at 04:11:37, which included START and END for all three panels, with TXNM 403, FTK 200 correct Lake Street link, and AVAV 200, but the first test had an invalid AVAV 404 and generic compliance matches that were corrected before scheduling. No orders were placed. https://www.txnmenergy.com/investors/rates-and-filings/pnm-nmprc-filings.aspx

A test was done on the stock symbol KALU, which is a measure of how much stress a position can take. The test used 50,000 paths, which are like different possible outcomes, and 20 roundtrips, which are like back-and-forth trades. It used 0.25, or 25%, of the total equity, and the initial notional, or initial investment, was 212.7475. The results showed that the return on investment could be as low as -17.225869134387388% or as high as 21.700392079364647%. There is a 53.396% chance of losing money, and the biggest possible loss, or drawdown, could be 19.88527913234894%. The test was trained on data from 2020 to 2024, and used a fixed rule, with a seed of 91104. This means that the test used a fixed fraction of the account equity for each trade. It also assumes that the returns on each trade are independent and do not change over time, which may not be realistic. The actual allocation was 0, which means that no money was actually invested, and the results are based on a zero-sized distribution, which is not the same as a risk-free investment. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

A news scan found the headline is not available, but a data line was found for KALU. The symbol is KALU and the label is point_quarter_probe. There were 50,000 paths and 20 roundtrips. The equity fraction is 0.00044063114064938486 and the initial notional is 0.37497269438122. The terminal return percentage quantiles are -0.032538057725157166, -0.0005875514963782358, and 0.03653816677399166. The probability of loss percentage is 51.11 and the drawdown 95 percentage is 0.03849808764110452. The training target frequency percentage is 6.451612903225806 and the training stop frequency percentage is 58.06451612903226. This data is based on 50,000 independent and identically distributed paths of 20 roundtrips from 2020 to 2024 using a fixed rule training with seed 91104, which assumes a fixed fraction of account equity per trade. This method assumes stationary independent trade returns, which may underestimate unseen gaps and regime persistence. The actual allocation is 0, and a zero-sized distribution is mechanically zero, not risk-free alpha. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

A test was done on the stock symbol CSTM, which is called a position cap stress test. This test had 50,000 paths and 20 roundtrips. It used 0.25 or a quarter of the equity. The test started with 212.7475 dollars. The results showed that the return could be as low as -20.732574897046543 percent, as high as 11.988630337550346 percent, and had a middle value of -6.561491360041982 percent. There was a 73.342 percent chance of losing money and a 22.50457021412355 percent chance of a big loss. The test was trained on data from 2020 to 2024 and used a fixed rule. The test assumed that each trade was independent and did not consider big gaps in the market. The actual allocation was 0, meaning no money was invested, so the results do not show any risk-free return. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

A report on the stock symbol CSTM, labeled as point_quarter_probe, shows the following results: it used 50,000 paths and 20 roundtrips, with an equity fraction and initial notional of 0. The terminal return percentage quantiles were 0, the probability of loss percentage was 0, and the drawdown 95 percentage was 0. The training target frequency percentage was 6.67 and the training stop frequency percentage was 50. This data is based on 50,000 independent and identical paths of 20 roundtrips from 2020 to 2024, using a fixed rule for training with a seed of 91104. The results assume that trade returns are stationary and independent, which may underestimate unseen gaps and regime persistence. The actual allocation is 0, meaning the distribution is zero-sized, which is not the same as being risk-free. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

A recovery model used a technique called Monte Carlo, which is a way of simulating different outcomes, to test the sensitivity of the EGA R91104. The model used 50,000 paths, which is a way of simulating different possible outcomes, and a seed of 91104, which is a starting point for the simulation. The model assumed that the rate of recovery would follow a triangular pattern, with rates of 3, 5.5, and 8 cells per day. The model also assumed that there would be a 7-day stabilization period, and that 315 cells were already stable as of August 26. The model estimated that the days after August 26 would be 138.35, 178.92, and 256.53 for different quantiles, which are measures of the spread of the data. The model also estimated that the missing cast proxy, which is a measure of the amount of material produced, would be 237,423.11, 304,144.57, and 421,742.72 for different quantiles through March 31. The model finished by March 31 in 81.662% of the simulations, but this is a model sensitivity and not a real-world probability. The model ignored scrap, shipments, and heterogeneous cells, and did not infer any trading probability. https://media.ega.ae/egas-al-taweelah-smelter-restoration-reaches-25-completion-milestone/

A news scan found the headline 'KALU R91104 costed L40 holdout slippage0.002'. The trading results for the symbol KALU are as follows: the lookback period was 40 days, and the results are for the holdout period, which means the period used to test the trading strategy. The slippage, or the difference between the expected and actual trade price, was 0.002. The return percentage was 5.499603791798744%, and the drawdown percentage, which is the maximum loss from a peak to a trough, was 23.73443674070127%. There were 16 trades, with a win percentage of 37.5%. The mean percentage return per trade was 0.8491566740459443%. The average win percentage was 13.23224997714553%, and the average loss percentage was 6.580699307813807%. The half Kelly criterion, a measure used to determine the optimal bet size, was 0.032086632111416796. There were 3 target exits, 9 stop exits, and 4 time exits. The ambiguous exits, which are trades that do not fit into any category, were 0. The trading strategy used was a long-only strategy, meaning only buy trades were made, and it was implemented using the vectorbt system. The accounting rules allowed for two trades per day, with the possibility of entering and exiting a trade on the same day, using the original open and close prices except for the exit price. The fees were 10 basis points per side, plus the listed slippage. The daily touch or gap model was not used, and the venue 15-minute model was also not used. There were zero ambiguous bars. No parameters were selected from the holdout period. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

A news scan found the headline 'KALU R91104 costed L40 holdout slippage0.001'. The trading results for the symbol KALU are as follows: the lookback period was 40, the period was holdout, and the slippage was 0.001. The return percentage was 8.930195851262615, the drawdown percentage was 23.198818871819448, and there were 16 trades. The win percentage was 37.5, the mean percentage was 1.051057291813047, the average win percentage was 13.458941622128103, and the average loss percentage was 6.3936733063759865. The half Kelly criterion, a measure of optimal bet size, was 0.03904680328225005. There were 3 targets, 9 stops, and 4 time exits, with 0 ambiguous exits. The trading strategy used was a long-only strategy, with two accounting points per day allowing for entry and exit on the same day, using original open and close marks except for the exit price. The fees were 10 basis points per side, plus the listed slippage. The daily touch or gap model did not use 15-minute venue data, and there were zero ambiguous bars. No parameters were selected from the holdout period. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

A news scan found the headline 'KALU R91104 costed L40 train slippage0.002'. The trading results for the symbol KALU are as follows: the lookback period is 40, the period type is train, and the slippage is 0.002. The return percentage is -20.170352764852886, the drawdown percentage is 60.66153052675468, and there are 23 trades. The win percentage is 39.130434782608695, the mean percentage is -0.5154549774926237, the average win percentage is 10.046954676920679, and the average loss percentage is 7.305575469615462. The half Kelly criterion is 0, there are 2 targets, 13 stops, and 8 time exits, with 0 ambiguous exits. The trading strategy used is a long-only strategy with vectorbt, allowing for two accounting points per day to permit entry and exit on the same day, using original open and close marks except for the exit price. The fees are 10 basis points per side plus the listed slippage. The daily touch or gap model does not use venue 15-minute data and there are zero ambiguous bars. No parameter was selected from the holdout set. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

A news scan found the headline 'KALU R91104 costed L40 train slippage0.001'. The trading results for the symbol KALU are as follows: the lookback period was 40 days, the period type was train, and the slippage, which is the difference between expected and actual trade prices, was 0.001. The return percentage was -16.41241017260007, and the drawdown percentage, which is the maximum loss from a peak to a trough, was 59.544485118869986. There were 23 trades, with a win percentage of 39.130434782608695. The mean percentage return was -0.3162863204532431, the average win percentage was 10.267269341568516, and the average loss percentage was 7.120000674610089. The half Kelly criterion, a betting strategy, was 0. There were 2 targets, 13 stops, and 8 time exits, with 0 ambiguous exits. The trading strategy used was a long-only strategy, meaning only buy trades were made, and was run using the vectorbt platform. The strategy allowed for two accounting points per day, permitting entry and exit on the same day, using the original open and close prices except for the exit price. Fees were 10 basis points per side, plus the listed slippage. The daily touch or gap model did not use venue 15-minute data, and there were zero ambiguous bars. No parameters were selected from the holdout set. https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all