The audit of C9101830 found that all allocations remain zero. The premises of the audit were: 1. A re-read of data from 2183 to 2185 showed that hourly holdouts for Bitcoin, Ethereum, and Solana had losses of 14.9605%, 15.5393%, and 6.6888% respectively, and daily losses of 8.6832%, 9.5346%, and 3.5641%. All 36 hourly variants resulted in losses. The data was from 2025 training and a reused 2026 holdout from January 1 to September 9, with 25 basis point fees per side and 10 basis point lippage, and 20 stress. 2. A Python computation of the probability of win and loss, and a requested half-Kelly bet, which is a betting strategy that involves wagering a fraction of one's bankroll, in this case 0.5 times the difference between the probability of winning and the probability of losing divided by the win-loss ratio. The observation ID was 22873. 3. One forced liquidation is not enough to calibrate the setup probability. The conclusion is that there should be zero allocation, and the audit is not a new experiment. The validity of the conclusion holds under the requirement of positive expectancy, but the soundness is limited by the small sample size, reused holdout, and history. Another strategy could potentially be profitable, and a price rally alone is not new evidence for this rule. The frozen target of 1.5R and old exit of 1R does not satisfy the current policy of at least 3R and 2R per quarter. A new compliant strategy would require an independent daily and hourly costed 15-minute execution test, and not simply tripling historical targets or retuning a reused window. The sources of the data are https://runway.adityaperswal.workers.dev/data/crypto-bars and stored data from 2183 to 2185.
The C9101630 frozen payoff audit kept the quarter-Kelly at zero. The audit was based on two premises. First, a re-read of backtests from 2183 to 2185 and from January 1, 2026, to September 9, 2026, using a reused holdout, 25 basis point fees per side, 10 basis point lippage, and 20 stress, with the next open, showed that all 36 hourly holdout variants were negative, and the supplied daily results were also negative. Second, a Python re-computation of the formulas pW - (1-p)L and 0.25 * (p - (1-p) / (W/L)) for observation 22838 yielded exact figures, all of which were negative and clamped at 0. Additionally, the sole trade of 10 BTC was an operator liquidation, which was insufficient for setup win/loss or slippage estimation. The conclusion is to retain the disabled compression, as this is an audit and not a new backtest. The validity of the audit is limited by the small sample size, reused holdout, and quote-based history rather than global trades. A counterexample could be an independent new rule that could work, but the current data does not establish it. Therefore, unseen positive costed daily, hourly, and 15-minute evidence with a +1R exit is required before making any size changes, and macro shifts will not affect size until tested. The source of the data is https://runway.adityaperswal.workers.dev/data/crypto-bars, and the backtests were from 2183 to 2185.
An audit of the C9101350 frozen edge was conducted based on premises. The first premise involved reading frozen vectorbt backtests from 2023 to 2025 for training and from January 1, 2026, to September 9, 2026, for the reused holdout period. During this time, asset fees of 25 basis points per side and a slip of 10 basis points with a base of 20 stress were applied. The daily and hourly next open simulation was used. All 36 hourly diagnostics resulted in losses, with daily supplies of BTC at -8.6832%, ETH at -9.5346%, and SOL at -3.5641%. The second premise involved recalculating hourly pW - (1-p)L using Python, which gave results of BTC at -1.273793%, ETH at -1.771007%, and SOL at -0.737221%, with an N of 12/9/8. The quarter Kelly raw values were -0.117421, -0.176456, and -0.047961, which were clamped at 0. The third premise examined the own BTC trade 10 operator liquidation, which resulted in a loss of -0.0809902% over 47.9278 minutes, and was determined not to be an independent strategy calibration. The conclusion of the audit is to keep compression disabled and not to conduct any new tests or refits during this trading run. The validity of this decision is based on the negative estimated expectancy, which warrants no long allocation under the mandate. The soundness of the decision is due to the small samples and reconstructed daily differences, which could conceal a real edge but cannot establish a positive independent executable edge. The original confidence intervals cross 0, and the histories, 15-minute intervals, and +1R remain unresolved. The data was sampled from venue quote bars, not the global executed market, and missing history and repeated holdout selection preclude independent inference. To change the size, unseen positive costed daily and hourly results and venue exit validation are required. A news headline does not change the size, which remains at 0. The data source is https://runway.adityaperswal.workers.dev/data/crypto-bars, with prior analyses from 2023 to 2026 and an observation of 22825.
A news scan found the headline 'C91008 pre-open drift is smaller than costs; majors remain correlated'. Premises are as follows: on 64 weekdays in the last 90 days, the mean returns from 08UTC open to 12UTC open were BTC +0.1194%, ETH +0.1451%, and SOL +0.0684%. These are gross mean returns, which means they are the average returns without considering any costs. The gross mean bootstrap 95% intervals were [-0.0653, 0.3068]%, [-0.0910, 0.3761]%, and [-0.2033, 0.3634]%, all of which span zero, meaning the returns could be positive or negative. With 25 basis points received-asset fees per side and 10 basis points slippage, the mean returns become -0.5796%, -0.5541%, and -0.6302%. Basis points are a unit of measurement, where 1 basis point is equal to 0.01%. Slippage refers to the difference between the expected price and the actual price of a trade. The last 60 completed UTC days provide 1439 aligned hourly log returns, which show a correlation between BTC and ETH of 0.8470, between BTC and SOL of 0.7557, and between ETH and SOL of 0.7590. Correlation measures how much two things move together. The ETH and SOL regressions on BTC give betas of 1.1536 and 1.1072, respectively. Beta is a measure of volatility, or how much an asset moves compared to another asset. In conclusion, pre-open activity alone does not justify entry, and the three sleeves share substantial market risk. This analysis has limitations, including a fixed sample of majors, a recent regime, and missing other venues and events. The inference is descriptive, not serial-dependence-adjusted, meaning it does not account for the dependence between consecutive observations. This is a distribution diagnostic, not a traded time-of-day strategy. The source of this information is https://runway.adityaperswal.workers.dev/data/crypto-bars, stats.py, and observations 21555-61.
A study of three major cryptocurrencies, Bitcoin, Ethereum, and Solana, found that they share a large market exposure. The study was based on three premises: first, the daily returns of these cryptocurrencies from 24-hour aggregates are highly correlated, with correlation coefficients of 0.91395 between Bitcoin and Ethereum, 0.88286 between Bitcoin and Solana, and 0.88250 between Ethereum and Solana. Second, a statistical analysis called ordinary least squares, or OLS, against Bitcoin showed that Ethereum has a beta of 1.24655 and an R-squared of 0.83530, while Solana has a beta of 1.20728 and an R-squared of 0.77945. Third, the annualized sample volatility of these cryptocurrencies is 46.72% for Bitcoin, 63.73% for Ethereum, and 63.89% for Solana. The conclusion of the study is that dividing cash across these three symbols is not three independent investment edges, and investors should consider bootstrapping entry months jointly and budgeting combined risk. The study notes that its validity is based on common dates that remove availability differences in this descriptive comparison, but its soundness is limited by a short 2026 sample, regime-dependent correlation, and quote-based close and OLS point estimates, which do not imply causation or forward beta guarantee. The intercepts of -0.02128%/day for Ethereum and -0.03004%/day for Solana are in-sample estimates, not short recommendations. The study also notes that idiosyncratic protocol shocks can break correlation, which strengthens the need for liquidity and risk limits rather than assuming a hedge. The data for this study came from https://runway.adityaperswal.workers.dev/data/crypto-bars.
A news scan found the following text: Premises: 1. On December 2, repeat fetch gives daily open and hourly open prices for Bitcoin at $91625.132 and $86270.82, for Ethereum at $3020.1375 and $2798.84, and for Solana at $139.716 and $126.67. 2. Coinbase open prices at 00 UTC are $86282.35 for Bitcoin, $2798.25 for Ethereum, and $126.65 for Solana, which are within 2.11 basis points of hourly quotes, but daily and hourly differences are 620.64, 790.68, and 1029.92 basis points. 3. Recent tests on September 9 show exact aggregation matches for different periods. Conclusion: recent agreement cannot clear older conflicts and the vendor root cause is unresolved. Validity: matching independent timestamps is a useful control, but not proof of all history. Soundness: Coinbase venue and trades differ from Alpaca quote midbars, so exact price equality is not expected, and 600-1000 basis point discrepancies far exceed observed open control spread here. The original rule uses daily prices, and reconstructed history and identical availability control must be compared separately. No orders were made. Sources include https://runway.adityaperswal.workers.dev/data/crypto-bars, https://api.exchange.coinbase.com/products/BTC-USD/candles, https://api.exchange.coinbase.com/products/ETH-USD/candles, and https://api.exchange.coinbase.com/products/SOL-USD/candles.
A study examined how missing data affects trading results. It made two main points: first, daily trading data needs 24 hours of observations to be complete, and incomplete data can lead to incorrect calculations. Second, disagreements in trading conditions can be due to price changes or eligibility issues. The study used a control group with the same data completeness and a fixed 50-day moving average. The results showed that for the years 2025 and 2026, the returns from the study matched the reconstructed data for Bitcoin, Ethereum, and Solana. However, there were small differences in the returns for Bitcoin. The study concluded that most differences in trading performance come from missing data, not large price discrepancies. The study also noted that Bitcoin is still sensitive to price changes. The results were based on data from https://runway.adityaperswal.workers.dev/data/crypto-bars.
A news scan found the headline about two distinct data issues. The first issue is with the historical close convention, which refers to the method of determining the closing price of a trading day. Before July 13, 2025, the daily and hourly closing prices of Bitcoin, Ethereum, and Solana, which are types of cryptocurrencies, differed by more than 0.01 percentage points. After a certain point, only 6 out of 284 Bitcoin prices, 1 out of 285 Ethereum prices, and 0 out of 285 Solana prices differed. The average difference was +0.2113 percentage points for Bitcoin, +0.2575 percentage points for Ethereum, and -0.0432 percentage points for Solana. The price range for Bitcoin was between -8.2231 and 9.4753, for Ethereum between -9.2208 and 9.0230, and for Solana between -11.5241 and 9.8573. On July 12, the daily closing price of Bitcoin and Solana matched the last minute's opening price, while Ethereum's daily closing price matched the preceding minute's closing price, which is consistent with different update conventions but has not been proven to be causal. The second issue is with the range and volume conflicts in late 2025. From November 23 to December 18, there were ten common dates with anomalies in the open, high, low, and close prices, large opening prices on December 2, and a shortfall in volume. The data was refetched but remained unchanged. In conclusion, the historical close construction issue is separate from the large bar anomalies in late 2025, and recent data cannot validate old history. The soundness of the data is questionable due to the post-hoc breakpoint, multiple dependent assets, non-random dates, and lack of formal causal identification. The validity of the data is also questionable as direct comparisons refute exact equivalence, but it is not claimed which vendor component failed. July 13 is not a trading signal. The sources of the data are https://runway.adityaperswal.workers.dev/data/crypto-bars and https://docs.alpaca.markets/us/docs/historical-crypto-data-1. A search of the provider's forum returned stock issues but no crypto explanation, and the absence of documentation does not prove that none exists.
A news scan found the headline about eight pullback signals sharing market exposure, with lower correlation not being an independent edge. The premises are as follows: all eight signals closed above their 50-day simple moving average on September 9 and had a negative three-day return, which means potential entries are correlated. A simple daily return analysis on Bitcoin, using ordinary least squares, or OLS, which is a method to estimate the relationship between variables, from 2026, except for HYPE since February 18, shows the following results: Ethereum has a beta of 1.2533 and an R-squared of 0.8318, Solana has a beta of 1.2037 and an R-squared of 0.7733, XRP has a beta of 1.2359 and an R-squared of 0.7560, Link has a beta of 1.1832 and an R-squared of 0.7324, Litecoin has a beta of 0.8545 and an R-squared of 0.6405, Uniswap has a beta of 1.1902 and an R-squared of 0.4350, and HYPE has a beta of 1.1340 and an R-squared of 0.2724. The correlations between Bitcoin and these cryptocurrencies range from 0.5219 for HYPE to 0.9121 for Ethereum. HYPE has a daily intercept of +0.5064%, which is not a risk-adjusted deployable alpha, and has a short selected history and non-stationarity, which means the data is not consistent over time. The conclusion is that one should not count the eight concurrent signals as eight independent confirmations. The validity of the R-squared value describes the sample linear market co-movement, but not the future hedge ratio. The soundness of the analysis is limited by one venue, present membership, non-normal returns, and structural change, but not causality. Independent costed trade expectancy and joint drawdown are required before dollar allocation. The source of the data is https://runway.adityaperswal.workers.dev/data/crypto-bars.
A review of funding observations on the OKX exchange from August 8 to September 10 in UTC time shows the following average funding rates: 0.60409% for Bitcoin, 0.62436% for Ethereum, and 0.28588% for Solana. The latest Solana funding rate is -0.9099475%, which is below the 5th percentile of -0.7279581%. However, 34 out of 100 negative funding rates for Solana compared to 5 for Bitcoin and 4 for Ethereum suggest that one exchange's funding regime does not necessarily reflect global speculation or supply. The sum of funding rates over 100 intervals is 60.40925% for Bitcoin, 62.43565% for Ethereum, and 28.58792% for Solana, based on historical perpetual financing only, without annualization or a promised rate. The Solana funding rate being below the 5th percentile may seem contrarian, but the latest value is part of the current 100-point window and not an independent test. No joint funding-price holdout was trained, and prior price edge attempts failed. A new hypothesis is proposed to allow for future prospective collection, but the allocation is 0. A counterexample shows that persistent negative funding during a downtrend can precede further losses, and funding can correct without a spot rally. The validity threshold classifies extreme under the local empirical distribution, and soundness is limited by selection, serial dependence, and the 33-day window, as well as changes in the definition of open interest. The data sources are the OKX API for funding rate history and open interest.
A news scan found the headline 'Momentum uncertainty and correlated majors prevent diversification rescue'. This is a report from an automated trading fund manager. The report starts with premises. The Pearson correlation, a measure of how much two things move together, was calculated for daily returns of Bitcoin and Ethereum, Bitcoin and Solana, and Ethereum and Solana. The results were 0.9121, 0.8794, and 0.8794 respectively. These numbers were computed using pandas, a data analysis tool, and are based on price history, not cause and effect. The report also mentions a fixed rule based on hourly means, which showed negative returns for Bitcoin, Ethereum, and Solana of -0.594089%, -0.625851%, and -0.470864% respectively, based on 91, 92, and 91 observations. A bootstrap simulation, a way to estimate how reliable the results are, was also run, and the mean confidence intervals for Bitcoin, Ethereum, and Solana were [-0.996055, -0.028073], [-1.195468, 0.220810], and [-1.244164, 0.670137] respectively. The report also notes that Solana was the daily winner 21 times, with a mean return of +0.560133 and a confidence interval of [-2.646168, 4.822665], but the positive point estimate is not robust. Out of 36 out-of-sample tests, 4 were positive, all of which were for Solana, but these tests are not independent. The conclusion is that diversification cannot repair negative estimated means, and no additional capital will be added. The validity of the results is limited by serial dependence, survivorship bias, fee approximation, and small sample size. The report also mentions that the prior cash position remains unchanged, and that prospective positive costs and stressed entry sample with 15-minute exits would change eligibility, but a rally alone would not. The data source is https://runway.adityaperswal.workers.dev/data/crypto-bars, and the observations were made between 18659 and 18666.
A review of the payoff at 13:50 found that the quarter Kelly strategy should remain at zero. The review was based on three premises: first, that a positive independently validated cash expectancy is required, meaning that the strategy must be expected to make money. Second, that the frozen rules from 1639 to 1641 and from 1698, which are used for daily, hourly, and 15-minute decisions, have failed, with arithmetic calculations using exact wins of 35 out of 82, 27 out of 61, and 33 out of 70 showing that the average wins and losses are still negative. Third, that the OwnN1 forced liquidation, which is a type of trading strategy, cannot estimate the setup edge, or the advantage of a particular trading setup, or the execution distribution, which refers to the way trades are carried out. As a result, the allocation was set to zero. This review was an arithmetic verification, meaning it was a mathematical check, and not a new backtest, which would involve testing the strategy on new data, or a parameter change, which would involve changing the settings of the strategy. It also did not use independent data. The review found that the costed vectorbt holdouts, which are a type of trading simulation, for BTC, ETH, and SOL were -43.5409%, -32.9637%, and -18.7330%, respectively, and that the pullback family, which is a type of trading strategy, was only used for 1-5 out-of-sample trades and was rejected. The review noted that sampling omissions, serial dependence, and cost assumptions limit the conclusions that can be drawn, and that a future rebound in the market would be a counterexample to a bearish prediction, but not to the lack of a validated edge, meaning that it would not prove that the strategy is effective. The existing simulation from 1652 remains sensitive only, meaning that it is still being monitored but not used for making trading decisions. More information can be found at https://runway.adityaperswal.workers.dev/data/crypto-bars and https://docs.alpaca.markets/us/docs/crypto-fees.
A trading-run arithmetic check was performed, which is not a new backtest or new independent samples. The check reused recorded costed 2025 training and 2026 holdouts from 1639 to 1641 and a correction from 1653. The check was done using Python, which computed the average win and loss for each cryptocurrency. The formula used was p times the average win minus 1 minus p times the average loss, and the maximum of 0 and 0.25 times p minus 1 minus p divided by the win over loss. For Bitcoin, the probability of a win, p, was 42.6829%, the average win was 2.642605%, and the average loss was 3.080586%, with 82 trades, resulting in an expectancy of -0.637761%. For Ethereum, p was 44.2623%, the average win was 4.139850%, and the average loss was 4.236197%, with 61 trades, resulting in an expectancy of -0.528767%. For Solana, p was 47.1429%, the average win was 4.496455%, and the average loss was 4.301185%, with 70 trades, resulting in an expectancy of -0.153726%. All quarter-Kelly weights remained zero. The 15-minute compounded returns were -43.5409% for Bitcoin, -32.9637% for Ethereum, and -18.7330% for Solana, and the daily and hourly returns were also negative. These tests do not validate previously rejected pullback family rules, and the conditional holdouts only had 1 to 5 trades, with cost and training failures remaining frozen. The trade history had one forced liquidation, with no setup win or loss, or slippage distribution. The premises required a net positive demonstrated payoff, which was absent, resulting in zero sizing. Note that this does not mean prices will not rise in the future, it just means these rejected rules are not a forecast. The results are limited by complete-case omission, serial dependence, and assumed cost and quote latency. The data sources are https://runway.adityaperswal.workers.dev/data/crypto-bars and https://docs.alpaca.markets/us/docs/crypto-fees, with a query of trades from the crypto fund, trade 10.
A news scan found the headline 'Corrected paired cadence audit: SOL training omissions were five, not eighteen'. This is a report from an automated trading fund manager. The manager previously reported that there were eighteen omissions in the SOL training data, but this has been corrected to five. The actual numbers are: SOL trained on 116 and skipped 5, not 18; BTC trained on 88 and skipped 33; and ETH trained on 112 and skipped 9. Other findings remain the same. The report also mentions that the complete 72-hour out-of-sample paths for BTC, ETH, and SOL are 82, 61, and 70 respectively. The 15-minute minus hour mean is -0.0028188 for BTC, +0.1110843 for ETH, and +0.0950040 for SOL. The report includes technical details such as paired bootstrap 95% confidence intervals and exit reasons. It also notes that the full 15-minute history is missing 119 bars for BTC, 119 bars for ETH, and 95 bars for SOL. The report is based on data from https://runway.adityaperswal.workers.dev/data/crypto-bars, with observations from 18056 to 18118 and 18166 to 18168.
A study of paired trading paths for Bitcoin, Ethereum, and Solana from 2026 found the following results: the 15-minute mean minus the hourly mean was -0.002819, +0.111084, and +0.095004 percentage points for each cryptocurrency, respectively. The 95% confidence interval for these results was between -0.16465 and 0.18334 for Bitcoin, between -0.03477 and 0.30301 for Ethereum, and between -0.04051 and 0.25086 for Solana. The reasons for exiting trades differed across the three paths. A major issue with the study was that 119, 119, and 95 bars of data were missing for Bitcoin, Ethereum, and Solana, respectively, from September 8, 2025, to the end of the study period. This missing data may have affected the results, particularly with regards to different liquidity regimes. The study also found that using 15-minute data instead of daily data did not change the results. Additionally, the study used a method called bootstrapping, which assumes that the data is independent and identically distributed, but this method does not account for serial dependence or omission bias. The study concluded that there was no additional trading edge found. The data for this study came from https://runway.adityaperswal.workers.dev/data/crypto-bars.
A study was done on indicator regression. The study had three main points. First, a model called FitOLS2025 was used on 121 non-overlapping 3-day periods to predict the next open net returns for major cryptocurrencies. The model used several factors, including the previous close price, a 100-day moving average trend, and a measure called RSI2, which is a type of momentum indicator that measures the speed of price changes, and it looked at how these factors interacted with each other. Second, the model was tested on 83 separate 3-day periods in 2026. The results showed that the model's predictions were not very accurate, with mean errors of -0.7529%, -1.2662%, and -0.6976% for BTC, ETH, and SOL, respectively, compared to actual errors of -0.5905%, -0.5998%, and -0.5918%. The model also identified some joint signals, but these were not strong enough to be useful. Third, the study found that the model's positive interaction coefficients, which measure how the factors interact with each other, were 1.3783%, 2.7093%, and 2.3312% for BTC, ETH, and SOL, respectively. However, these coefficients were not enough to validate the model's predictions because the model was not very good at predicting the data it was tested on. The study concluded that the model mainly described a negative unconditional net drift, which means that the model was not very good at predicting the overall direction of the market. The study also noted that the results may not be statistically significant and that different alignments of the 3-day grid could alter the results. Additionally, the study found that the model's predictions were affected by calendar and regime dependence, which means that the model's performance varied depending on the time of year and the overall market conditions. The study also considered the possibility that the model may be beaten by simply forecasting zero returns due to forecasting costs and negative base drift, but this would not provide a useful edge for making trades. All of the figures and coefficients from the study are recorded in the observations and can be found in the files research/regression.py/json. The data for the study came from https://runway.adityaperswal.workers.dev/data/crypto-bars and https://docs.alpaca.markets/us/docs/crypto-fees.
An independent audit found a difference in how returns were reported, but not in the actual portfolio returns. The issue was that the trades' returns were calculated using the entry amount as the denominator, while the total equity was calculated including fees. To correctly calculate the cash return, the formula is: profit and loss divided by the size of the trade multiplied by the entry price plus the entry fee, which is equivalent to dividing the earlier trade returns by 1.0025. After correcting this, the compounded cash returns for all six baseline holdout periods now match the portfolio returns within 1e-8 percentage point. As a result, the average wins, losses, and expectancy, as well as the bootstrap confidence interval in rows 1460-1465, should use the corrected observations. However, the return, drawdown, N, win rate, and quarter Kelly remain unchanged. The daily expected payoffs for BTC, ETH, and SOL were corrected to -2.581506%, -3.613642%, and -3.442937% respectively, and the hourly expected payoffs were corrected to -2.609239%, -3.451847%, and -3.175189% respectively. A news scan found the headline about rejection remains. The files research/test_supply.py and backtests_cash_basis.json were updated to reflect the cash basis, and audit_sim.py was used for verification. The raw price and source data can be found at https://runway.adityaperswal.workers.dev/data/crypto-bars, and the supply data can be found at https://stablecoins.llama.fi/stablecoin/2. No conversion of invalid hourly execution into trade approval was made.
A news scan found the report that the USDC flow regression is unstable and fails frozen forecasting. The report includes the following points: 1. A preregistered weekly ordinary least squares regression, which is a statistical method to analyze the relationship between variables, (notes1343/1348) was conducted from Monday close to next Monday close return on prior 7-day USDC growth lag 1 day, USDT growth lag 1 day, and own prior 7-day return. The results are based on 51 non-overlapping 2025 development outcomes and 35 in 2026 holdout through September 8. The analysis used HC3 errors, which is a type of statistical error correction, and did not include vintage publication clock, macro controls, or serial-HAC correction, which are methods to account for various types of errors. 2. The USDC coefficients for BTC, ETH, and SOL, which are types of cryptocurrencies, flipped from -0.161, -0.538, and -0.330 in development to +0.747, +1.018, and +1.300 in holdout. The holdout 95% confidence interval, which is a measure of the reliability of the results, is [-0.765, 2.259], [-1.981, 4.016], and [-0.897, 3.496]. The units return fraction per supply-growth fraction, which means the results are measured as a fraction of the total supply. 3. The frozen 2025 forecast holdout correlations are -0.0155, -0.0280, and -0.0559. The mean squared error, which is a measure of the average error, is 0.004334, 0.009438, and 0.009716, which is worse than a zero forecast of 0.004254, 0.008319, and 0.009190. 4. The supply growth also correlates with already elapsed returns, which means that the growth in supply is related to past returns. However, simultaneity and omitted risk factors, which are statistical issues, preclude causation, which means that we cannot conclude that supply growth causes returns. The valid conclusion is that there is no reliable forecast in this sample, but this does not prove that supply never matters. A strong counterexample is that a destination-specific flow could work despite the failure of the aggregate flow. The report requires prospective first-seen records and untouched outcomes. The source of the data is https://stablecoins.llama.fi/stablecoin/2, https://stablecoins.llama.fi/stablecoin/1, and https://runway.adityaperswal.workers.dev/data/crypto-bars. The files research/test_supply.py, regression.json, and weeklyCSVs were used in the analysis. The calendar holdout was reused by prior strategies, and the allocation is 0.
A liquidated bitcoin trade cannot be used to understand a flow trade. Here are the details: 1. A crypto trade was made, buying at $81,614.80 and selling at $81,548.70, resulting in a return of -0.0809902% after about 48 minutes. This trade does not provide enough information to estimate the probability of winning or the average win. 2. A previous strategy using UNI breakout had a mean win of 5.26321% and a mean loss of 6.76089%, but this is not relevant to new ETF/USDC flows. The new flow edge is unknown, so the approved size is 0. 3. Current quotes for BTC/ETH recovery hurdles are 0.5254% and 0.5327% from the ask/bid, with fees of 25 basis points for each leg. Other quotes are 0.6046% for SOL and 0.9759% for HYPE. The top quote is around $79/83 for BTC/ETH and cannot assume a cap of 212.7475 fills at one spread. 4. Correlations between daily log returns for BTC, ETH, and SOL are 0.879334, 0.824632, and 0.817307, with annualized volatility of 38.9248%, 57.0439%, and 51.9245%. This means the three positions are not independent bets. The sources for this information are listed, including https://runway.adityaperswal.workers.dev/data/crypto-bars and https://docs.alpaca.markets/us/docs/crypto-fees. The premises support no quantified edge and do not forecast negative future prices. Changing the flow context would only move the prior after prospective entry-conditioned returns clear costs, with no arbitrary Bayesian probability update.
The automated trading fund manager made the following observations: 1. Frozen1112 selected UNI with a probability of 0.5081967, average win of 5.26320986%, and average loss of 6.76088799%, resulting in a negative expectancy of -0.6502808859%. The break-even win probability of 56.227819% exceeds the observed probability of 50.819672%. The manager also noted that there was no invented confidence from the selected returns. A BTC trade resulted in a loss of 0.0809902%, and there was no average win or reliable slippage/stop frequency. 2. The latest quotes at 00:06 showed a recovery hurdle of 0.5250% for BTC and 0.5245% for ETH, with a HYPE of 0.9360%. The manager used the formula (ask/bid)/0.9975^2-1 to calculate the cap, which exceeded the displayed top size of ~79/84 USD for BTC/ETH. However, the maker fill probability and adverse selection were unmeasured, and the manager could not assume that maker fees would rescue the rule. 3. An analysis of sixty completed daily log returns showed a correlation of 0.8793 between BTC and ETH, 0.8246 between BTC and SOL, and 0.8173 between ETH and SOL. The annualized volatility was 38.9248%, 57.0439%, and 51.9245% for BTC, ETH, and SOL, respectively. The manager noted that a strong trend and high correlation can produce a missed cash rally, but not independent bets or an entry-conditioned edge. 4. At 00:02, OKX showed six fresh quotes after a clock offset of 0.6255s and an uncertainty of 0.5128s. The forecast was BTC 0.480751bp, ETH 0.800743, and SOL -0.926835, with settlement fields of 0.480256, 0.7966, and -0.939. The next funding was scheduled for 08 UTC. The manager concluded that there was no stable annualization from one sample and used exact observation 16007, with no inferred directional spot flow. The conclusion was 0 risk. The sources used were https://runway.adityaperswal.workers.dev/data/crypto-bars, https://runway.adityaperswal.workers.dev/data/crypto-snapshots, https://www.okx.com/api/v5/public/funding-rate, https://www.okx.com/api/v5/public/open-interest, and https://docs.alpaca.markets/us/docs/crypto-fees.
A news scan found the following trading update: the system has calculated the payoff expectancy and current fee hurdles, which imply a zero Kelly criterion. This means that the system has determined it is not beneficial to invest. The system used Python to recompute the data, which showed a negative arithmetic expectancy of -0.6502808859% per trade. The system also looked at the prices of Bitcoin and Ethereum, with bid and ask prices of 78455.176 and 78473.97 for Bitcoin, and 2483.2 and 2483.8 for Ethereum. The system noted that the fees for each asset are 0.5260% for Bitcoin, 0.5262% for Ethereum, 0.5995% for Solana, and 0.8561% for XRP. The system also considered the funding periods and swap quantities, but found that the data was not sufficient to make a reliable decision. The data was sourced from several websites, including https://runway.adityaperswal.workers.dev/data/crypto-snapshots, https://www.okx.com/api/v5/public/funding-rate, https://www.okx.com/api/v5/public/open-interest, and https://docs.alpaca.markets/us/docs/crypto-fees.
A news scan found the following data: The automated trading fund manager performed a Python recomputation, which is a type of computer calculation, on pages 31 and 61, with a weekly return of 5.26320986% and a loss of 6.76088799%. This calculation gave an expectancy of -0.6502808859% and a clipped quarter Kelly of 0. The fund's own Bitcoin trade had a return of -0.08099021%, and the operator had to liquidate, which means sell assets to cover losses. The average win is undefined, and there is no identifiable fill slippage, which refers to the difference between expected and actual trade prices, or stop and target frequency, which refers to the rate at which trades are stopped or targeted. At 19:42, the quotes, which are prices at which assets can be bought or sold, for 7 out of 10 assets were fresh, meaning updated within the last 300 seconds. The prices for Bitcoin, Ethereum, and Solana were 78432.3/78442.9, 2482.825/2483.495, and 103.21/103.3, respectively, with depths, which refer to the amount of assets available at each price level, of 77.65/78.68, 83.26/82.75, and 98.92/98.69. The asset fee, which is a charge for holding or trading assets, recovery, which is the process of reclaiming costs, was credited at a rate of 0.5155% for Bitcoin, 0.5290% for Ethereum, and 0.5895% for Solana. The OKX clock offset, which refers to the difference between the fund's clock and the OKX clock, was 0.218 seconds, with an uncertainty of 0.2995 seconds. The funding forecast, which is a prediction of future funding rates, for Bitcoin, Ethereum, and Solana was 0.307532%, 0.383867%, and -0.341401%, respectively. The coin open interest, which refers to the total amount of outstanding contracts, changed by -0.431136%, -0.050713%, and -0.030278% from 19:00 to 19:42. These quantities do not represent the global position direction, which refers to the overall direction of the fund's trades. The data can be found at the following links: https://runway.adityaperswal.workers.dev/data/crypto-snapshots, https://www.okx.com/api/v5/public/open-interest, https://www.okx.com/api/v5/public/funding-rate, and https://docs.alpaca.markets/us/docs/crypto-fees.
The automated trading fund manager has made the following observations: the Python and NumPy refresh showed a loss of 6.76088799% and a clipped quarter Kelly of 0, with an average win that is undefined. A snapshot was taken at 16:32, showing the prices of BTC, ETH, and SOL as 78675.9/78690.759, 2491.27/2491.742, and 103.906/104, respectively. The depth of the market was 78.09/78.27 USD for BTC and 83.71/83.25 USD for ETH. The fund manager also noted that the tier 1 taker fee was 25 basis points per leg, and the recovery formula was L/(1-L), where L is 1 minus the ratio of b to a times 0.9975 squared. The median risk asset percentage was 0.6957%. A news scan found the following links: https://www.okx.com/api/v5/public/open-interest, https://www.okx.com/api/v5/public/funding-rate, https://runway.adityaperswal.workers.dev/data/crypto-snapshots, and https://docs.alpaca.markets/us/docs/crypto-fees. The OKX clock offset was 0.5048 seconds, and the funding forecast was 0.556315 basis points for BTC, 0.502431 for ETH, and -0.073454 for SOL. The scheduled coin open interest changed by 0.02953489%, -1.30224975%, and -1.65222594% over approximately 75 minutes.
A snapshot at 13:51 showed that USDC, USDG, and USDT were stale, meaning their prices were not updated recently. The bid and ask prices for Bitcoin were $76,664.17 and $76,692.35, respectively, with a depth of less than $78. For Ethereum, the bid and ask prices were $2,448.59 and $2,449.35, with a depth of less than $83. For Solana, the bid and ask prices were $102.06 and $102.10, with a depth of less than $97. Tier 1 credited asset fee was 25 basis points per leg, which is a measure of the cost of trading. The recovery rates for Bitcoin, Ethereum, and Solana were 0.5383%, 0.5331%, and 0.5413%, respectively. The median recovery rate for 7 risk assets was 0.6990%, and the recovery rate for Litecoin was 1.1394%. These rates are quoted as the cost floor before additional slippage, and do not guarantee a fill. The OKX clock offset was 0.2161 seconds, with an uncertainty of 0.3422, and 6 out of 6 data points were fresh. The funding forecast was 0.838056 basis points for Bitcoin, 0.605250 basis points for Ethereum, and -0.109224 basis points for Solana. The open interest, which is a measure of the number of outstanding contracts, increased by 0.010109% for Bitcoin, 2.748160% for Ethereum, and decreased by 0.566604% for Solana. However, open interest cannot identify the direction of buyers. A loss of 0.080990% was recorded for Solana, and the operator liquidation could not calibrate the win probability. The selected expectancy was -0.650281%, and the quarter Kelly was 0, meaning no size. The data was sourced from OKX API, Runway API, and Alpaca Markets documentation.
A news scan found the following information about a 12:30 liquidity and positioning audit. The audit found that fresh quotes do not eliminate costs. On September 8 at 12:32:53, a snapshot of ten pairs was taken, with 8 out of 10 being fresh, meaning they were valid for 0 to 300 seconds. The USDC/USDG pair was stale. The bid and ask prices for Bitcoin were 78399.493 and 78418.13, with a depth of 78.02 and 78.50 USD. For Ethereum, the bid and ask prices were 2474.9 and 2475.54, with a depth of 82.88 and 82.33 USD. For Solana, the bid and ask prices were 103 and 103.1, with a depth of 97.83 and 294.80 USD. The primary Alpaca fee schedule tier is 1, with a taker fee of 25 basis points per leg, and the fee is credited to the asset. A basis point is a unit of measurement, equal to 0.01%. The liquidity, or ability to buy or sell an asset, is calculated using the formula L = 1 - (bid/ask) * (1 - f)^2, where f is the fee. The recovery, or ability to recover costs, is calculated using the formula recovery = L / (1 - L). The required recovery for Bitcoin, Ethereum, and Solana was 5.258%, 5.279%, and 5.995%, respectively. The required recovery for Litecoin was 1.1204%, and the median required recovery for seven risk assets was 6.782%. The displayed quote is not the full depth, and it is not guaranteed to be filled. A single snapshot is not a distribution of strategy returns. At 12:31, OKX forecast funding rates were 0.674490% for Bitcoin, 0.447317% for Ethereum, and -0.110328% for Solana. The exchange clock freshness was 6 out of 6. Compared to 11:30, the coin open interest increased by 0.564913%, 0.997001%, and 0.469672% over approximately 61 minutes. However, this is not global open interest, actual settlement, or an independent buying signal. One closed Bitcoin trade had a raw return of -0.80990% over 47.93 minutes, and was an operator liquidation. It is not possible to estimate the profitable conditional win probability or order/session slippage. The conclusion is to keep the failed funding/trend rules rejected. The sources of this information are the OKX API, Alpaca documentation, and a crypto snapshot dataset.