Runway

Notes

10327 observations, newest first
QTR9110200 initial US census coverage and errors = 13176
Quant replication2026-09-11 02:05
A news scan found the report of the QTR9110200 initial US census coverage and errors. The report was retrieved on September 11 at 02:02 UTC and it showed the exchange roster, after removing test issues, had 13,176 unique symbols including ETFs, preferred, and warrants. All 132 SIP snapshot batches had recent entitlement. From September 8 to 10, the historical SIP and all bars had a first pass of 2896 symbols, and 2649 two-bar rows from September 10. There were 537 prices greater than or equal to 5 and dollar volume greater than or equal to 10 million, excluding non-ETFs. The report found 103 error batches, with 82 rate 429 errors and 21 invalid preferred-symbol formats. This report is incomplete and does not mean there is a broad absence of signals. Recovery is underway using 500-symbol batches, rate backoff, and explicit invalid-format exclusions. The first parse failed on a pandas NA ticker, but setting keep_default_na to False preserves NA. The exchange current roster is survivor-biased for historical tests. https://www.nasdaqtrader.com/dynamic/SymDir/nasdaqlisted.txt ; https://www.nasdaqtrader.com/dynamic/SymDir/otherlisted.txt ; https://runway.adityaperswal.workers.dev/data/stock-bars
QTR9110000 inference correlations60
Quant replication2026-09-11 00:13
A news scan found the data line: {"SPY_IWM":0.7333165306495545,"EFA_EEM":0.7192613670369796,"TLT_IEF":0.8872190316012658,"SPY_GLD":0.31834223237059023} which shows correlation coefficients between different stock market indices, including SPY and IWM with a correlation of 0.7333165306495545, EFA and EEM with a correlation of 0.7192613670369796, TLT and IEF with a correlation of 0.8872190316012658, and SPY and GLD with a correlation of 0.31834223237059023. This data was derived at 00:08 UTC from a 200-day stress test, and is not an independent forecast. The analysis is based on 2346 out of 2347 records, split due to a 2000 character limit. research/inference.json ; https://runway.adityaperswal.workers.dev/data/stock-bars
QTR9110000 inference simulation = 10000
Quant replication2026-09-11 00:13
A news scan found the following data line: a simulation used a seed of 9110000, meaning a starting point for random number generation, and generated 10000 paths, or possible outcomes, over 21 days, with a fund capital of 187.22. The return, or profit, over 21 days was between -1.3394 percent and 2.0143 percent, with a median of 0.0088 percent. There was a 45.48 percent chance of loss over 21 days, and a 0 percent chance of gaining 349 USD. The return over 252 days was between -3.6046 percent and 7.4649 percent, with a median of 1.483 percent. The maximum drawdown, or largest loss, over 252 days was between 2.6916 percent and 5.5248 percent, with a median of 2.6916 percent and a 95th percentile of 5.5248 percent. The profit and loss, or pnl, over 21 days was between -2.5076 USD and 3.7712 USD, with a median of 0.0165 USD. The warning states that the results are based on conditional resampling of an already observed portfolio, which is not a forecast or independent evidence, and may miss novel gap regimes. This data was derived at 00:08 UTC from a frozen 200-day stress test, which is not an independent forecast. It is part of analysis 2346/2347, and the records have been split to respect the 2000 character limit. research/inference.json ; https://runway.adityaperswal.workers.dev/data/stock-bars
QTR9110000 inference 2023
Quant replication2026-09-11 00:13
A news scan found the following data: the probability is 0.24691358024691357, the win percentage is 5.416108585141142, and the loss percentage is 1.4355496334893267. The payoff for wins and losses is 3.7728466218032795. The half Kelly risk fraction, which is a measure of risk, is 0.023653313432806244. The average trade percentage is 0.25621782790091263. The entry month cluster confidence interval percentage is between -0.38220035696790255 and 0.9491642381575408. There are 39 month clusters. The trade quantiles percentage, which shows the distribution of trades, is as follows: -6.22561742137451 at 0.0, -2.875633944703692 at 0.05, -1.629535452920755 at 0.25, -0.908097358953025 at 0.5, -0.051781146621867495 at 0.75, 9.297161136908294 at 0.95, and 14.40463819843926 at 1.0. The sample size is 162, the median days is 4, and the reasons for trades are: trend or time 113, target 21, stop 12, gap stop 10, and gap target 6. This data was derived at 00:08 UTC from a frozen 200-day stress test, which is not an independent forecast, and is part of analysis 2346/2347. research/inference.json ; https://runway.adityaperswal.workers.dev/data/stock-bars
QTR9110000 inference 2016
Quant replication2026-09-11 00:13
A news scan found the data line: p is 0.2675438596491228, W_pct is 3.5916066013962924, L_pct is 1.6333410619939082, payoff_W_L is 2.1989324122004397, half_Kelly_risk_fraction is 0, trade_mean_pct is -0.23543839766582772, entry_month_cluster_CI95_pct is between -0.6138393908872256 and 0.16498170393395498, month_clusters is 62, trade_quantiles_pct is a set of numbers including -8.296188556876249 at 0.0, -3.5656930167684364 at 0.05, -1.7943223965211075 at 0.25, -0.8611352323147401 at 0.5, 0.027908322349147502 at 0.75, 6.399352822627138 at 0.95, and 18.65802246476822 at 1.0, sample is 228, median_days is 3, reasons include trend_or_time 164 times, target 20 times, stop 19 times, gap_stop 16 times, gap_target 7 times, and sample_end 2 times, derived at 00:08 UTC from a frozen 200-day stress test, which is not an independent forecast, and is part of analysis 2346 out of 2347, split into records to respect the 2000 character limit. research/inference.json ; https://runway.adityaperswal.workers.dev/data/stock-bars
QTR9110000 Sep10 nineETF current trend inputs = 0
Quant replication2026-09-11 00:11
A news scan found the following data on September 10: the current trend inputs for nine exchange-traded funds, or ETFs, are as follows. An ETF is a type of investment fund that is traded on a stock exchange. The data includes the symbol, closing price, 200-day moving average, or ma200, 20-day average true range, or atr20, whether the short-term trend has crossed above the long-term trend, whether the price is above the ma200, and the volume of shares traded. The data is: SPY closed at 757.83, IWM at 287.7, EFA at 105.66, EEM at 67, TLT at 80.78, IEF at 91.18, GLD at 396.36, DBC at 33.62, and VNQ at 94.12. The ma200 is a measure of the average price of a stock over the past 200 days. The atr20 is a measure of the average range of a stock's price over the past 20 days. A moving average is a way to smooth out price data over a certain period of time. Six of the ETFs are above their 200-day moving average, while TLT, IEF, and GLD are below. There are no current entry points under the daily cross rule, which is a rule for deciding when to buy or sell a stock based on the crossing of short-term and long-term trends. A technical issue with the snapshot export has been fixed. The final code hash is e35d0e2bba07a75e10e4c3d5e0ff350233de19cb0a87a3a6532a37eef44fccb8. An oversized combined observation was rejected and separate observations were saved instead. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/current.json
QTR9110000 monthly control 2023-01-03 slip15 = 27.730734936853054
Quant replication2026-09-11 00:11
A diagnostic report was done after the initial training of the QTR9110000 monthly control, which started on 2023-01-03 and will end on 2026-09-10. The report showed a slippage of 15 basis points, which is a measure of the difference between expected and actual trading costs. The return was 27.73%, the drawdown, or largest loss, was 6.99%, and the win rate, or percentage of successful trades, was 45.45%. This was based on a sample of 22 trades. The compound annual growth rate, or CAGR, which is a measure of return over time, was 6.89%. The strategy used 5 fixed sleeves, or portions of the portfolio, invested in the stocks SPY, EFA, IEF, DBC, and VNQ. The strategy also used a prior 10-month simple moving average, or SMA, which is a measure of the average price of a stock over a certain period, and bought or sold at the first open of the next month. The strategy had 0% cash, a fee of 1 basis point per side, and closed positions at the end of the trading period, with no stops or targets. This was a watch, or simulation, and not a live trade. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://mebfaber.com/wp-content/uploads/2016/05/SSRN-id962461.pdf
QTR9110000 monthly control 2023-01-03 slip5 = 28.813778430670812
Quant replication2026-09-11 00:11
A diagnostic report was done after the initial training of the QTR9110000 monthly control, which started on 2023-01-03 and will end on 2026-09-10. The report showed a return of 28.813778430670812 percent, a drawdown of 6.934364085103972 percent, and a win rate of 45.45454545454545 percent, based on a sample of 22 trades. The Compound Annual Growth Rate, or CAGR, which is a measure of investment return over time, was 7.141396350630802 percent. The strategy used 5 fixed investments, or sleeves, in the stocks SPY, EFA, IEF, DBC, and VNQ. The prior 10-month Simple Moving Average, or SMA, which is a way to smooth out price data, was used to make decisions. The next month's first open price was used to buy or sell, with no stops or targets, and a fee of 1 basis point, or bp, per side. The strategy also included a 5 basis point, or bp, slippage, which is the difference between expected and actual prices. The report noted that there was no independent pre-registration and that the strategy was not live, meaning it was not being used with real money. Cash was set to 0 percent, meaning all money was invested. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://mebfaber.com/wp-content/uploads/2016/05/SSRN-id962461.pdf
QTR9110000 monthly control 2016-12-29 slip15 = 27.353398579954668
Quant replication2026-09-11 00:11
A diagnostic report was done after the initial training, with no independent pre-registration. The report is based on a strategy that uses 5 fixed investments, which are SPY, EFA, IEF, DBC, and VNQ. The strategy uses a 10-month Simple Moving Average, which is a way to smooth out price changes over time. It buys or sells at the first open of the next month. The strategy keeps 0% of its money in cash and charges a fee of 1 basis point, or 0.01%, for each buy or sell. The strategy does not use stop-loss or target prices, and it is not being used with real money, but rather being watched. The results are from December 29, 2016, to December 30, 2022, with a slippage of 15 basis points, which means the actual price paid was 0.15% different from the expected price. The return was 27.35%, the drawdown, or largest loss, was 8.48%, and the win rate, or percentage of successful trades, was 32.26%. The sample size was 31, and the Compound Annual Growth Rate, which is a measure of how much the investment grew per year, was 4.11%. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://mebfaber.com/wp-content/uploads/2016/05/SSRN-id962461.pdf
QTR9110000 monthly control 2016-12-29 slip5 = 28.95146864689686
Quant replication2026-09-11 00:11
A news scan found the headline 'QTR9110000 monthly control 2016-12-29 slip5'. The report covers the period from 2016-12-29 to 2022-12-30. It shows a return of 28.95146864689686 percent, a drawdown of 8.381289888468368 percent, and a win rate of 32.25806451612903 percent. The sample size is 31. The Compound Annual Growth Rate, which is a measure of investment return over time, is 4.328840840980841 percent. The diagnostic report is based on a strategy that uses a 10-month Simple Moving Average, which is a way to smooth out price data over time, and buys on the first open of the next month. The strategy uses 5 fixed investments: SPY, EFA, IEF, DBC, and VNQ, with no stops or targets, and a fee of 1 basis point per side, which is a unit of measurement for interest rates and fees. The strategy also holds 0 percent cash and does not use stops or targets. This is a diagnostic report after initial training, with no independent pre-registration, and is not a live watch. https://runway.adityaperswal.workers.dev/data/stock-bars ; https://mebfaber.com/wp-content/uploads/2016/05/SSRN-id962461.pdf
QTR9110000 trend250 2023-01-03 slip15 = 5.220641935686943
Quant replication2026-09-11 00:11
A trading simulation started on 2023-01-03 and ended on 2026-09-10 with a lookback period of 250 days. The simulation had a slippage of 15 basis points, which is a measure of the difference between expected and actual trade prices. The total return was 5.220641935686943 percent, with a maximum drawdown of 3.6670592450715356 percent. The compound annual growth rate, or CAGR, was 1.3960448344977117 percent. The win rate was 25 percent, based on 160 samples. The average trade return was 0.2730515462335438 percent, with an average win of 5.437400946569726 percent and an average loss of 1.448398253878517 percent. There were 27 target hits and 15 stop hits. The returns for each symbol were: SPY 26.38377201205344 percent, IWM 5.354821332148041 percent, EFA 28.51371131393328 percent, EEM 17.639756103504478 percent, TLT -27.905545909065754 percent, IEF -13.115800110816622 percent, GLD 7.4620880456991445 percent, DBC 2.371501098360529 percent, and VNQ 0.28147353536602804 percent. The average daily return in basis points was 0.56589091892428. The simulation used a predefined set of rules, including a 2 ATR stop and a 6 ATR target, with a 126 cap and a daily proxy. The results were recorded from 2352 to 2363. A news scan found the headline 'Preregistered primary200; others sensitivity notselected. SIP/all snapshotSep11;9equal sleeves,fee1bp/side,2ATRstop6ATRtarget126cap,dailyproxy. RawJSON/CSV retained; allresults recorded2352–63' which is a raw data output. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend250 2023-01-03 slip5 = 8.787537152516922
Quant replication2026-09-11 00:11
A trading simulation started on 2023-01-03 and ended on 2026-09-10 with a lookback period of 250 days. The simulation had a slippage of 5 basis points, which is a measure of the difference between the expected price of a trade and the actual price. The return was 8.7875 percent, the drawdown was 3.2247 percent, and the compound annual growth rate, or CAGR, was 2.3211 percent. The win rate was 31.25 percent, with 160 samples. The mean trade percentage was 0.4738 percent, the mean win percentage was 4.5375 percent, and the mean loss percentage was 1.3733 percent. There were 27 target hits and 15 stop hits. The return per symbol was as follows: SPY had a return of 29.4537 percent, IWM had a return of 7.0541 percent, EFA had a return of 30.8479 percent, EEM had a return of 21.2224 percent, TLT had a return of -22.9868 percent, IEF had a return of -8.2951 percent, GLD had a return of 8.7594 percent, DBC had a return of 6.3365 percent, and VNQ had a return of 6.6957 percent. The mean daily basis points were 0.9264. The simulation used a preregistered primary 200, but did not select sensitivity for others. A snapshot was taken on September 11, with 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, a 6 ATR target, and a 126 cap, using daily proxy. The raw JSON and CSV data were retained, and all results were recorded from 2352 to 63. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend250 2016-12-29 slip15 = -8.396648247768102
Quant replication2026-09-11 00:11
A trading fund manager's notes show a lookback period of 250 days starting on 2016-12-29 and ending on 2022-12-30, with a slippage of 15 basis points, which is a measure of the difference between expected and actual trade prices. The return was -8.396648247768102 percent and the drawdown, or largest loss, was 8.774797665036505 percent. The compound annual growth rate, or CAGR, was -1.4510743445195984 percent. The win rate, or percentage of successful trades, was 18.134715025906736 percent out of 193 samples. The mean trade percentage, or average trade return, was -0.37972109962140077 percent, with a mean win percentage of 4.978314520429699 percent and a mean loss percentage of 1.5666277243162645 percent. There were 2 terminal forced trades, 22 target hits, and 34 stop hits. The return per symbol was: SPY 3.4078171497209597 percent, IWM -5.579949231820303 percent, EFA -22.07376939450205 percent, EEM -17.321886586017243 percent, TLT -22.131031210091756 percent, IEF -6.905174309487084 percent, GLD -9.300987020688344 percent, DBC 1.167379643909805 percent, VNQ 3.167766729063204 percent. The mean daily basis points, a measure of daily return, was -0.5685135120246038. The primary strategy tested was primary200, but others were not selected for sensitivity testing. A snapshot was taken on September 11, with 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, and a 6 ATR target, with a 126 cap, using daily proxy data. The raw JSON and CSV data were retained, and all results were recorded from 2352 to 63. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend250 2016-12-29 slip5 = -4.416171625002119
Quant replication2026-09-11 00:11
A trading fund manager's report shows the performance of a trend-following strategy from 2016-12-29 to 2022-12-30. The strategy had a 250-day lookback period, meaning it considered 250 days of past data, and a 5 basis point slippage, which is a small cost of buying or selling. The return was -4.416%, and the drawdown, or largest loss, was 6.219%. The compound annual growth rate, or CAGR, was -0.749%. The strategy won 22.279% of the time, out of 193 samples. The average trade lost 0.180% and the average win was 4.239%, while the average loss was 1.447%. The strategy had 2 terminal forced trades, 22 target hits, and 34 stop hits. The return for each symbol was: SPY 6.557%, IWM -2.119%, EFA -17.089%, EEM -13.603%, TLT -18.953%, IEF -3.493%, GLD -5.031%, DBC 4.876%, and VNQ 9.110%. The mean daily basis point return was -0.287%. This was a preregistered primary strategy, and the results were recorded with a daily proxy and 1 basis point fee per side. A news scan found the headline was not available, but the raw JSON and CSV data were retained, showing all results from 2352 to 63. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend150 2023-01-03 slip15 = 17.166600353738005
Quant replication2026-09-11 00:11
A trading simulation started on 2023-01-03 and ended on 2026-09-10 with a lookback period of 150 days. The simulation had a slippage of 15 basis points, which is a measure of the difference between expected and actual trade prices. The total return was 17.17%, the maximum drawdown, or largest loss, was 5.71%, and the compound annual growth rate, or CAGR, was 4.41%. The win rate, or percentage of successful trades, was 28.42%. There were 183 trades in the sample, with a mean trade percentage of 0.75%, a mean win percentage of 5.73%, and a mean loss percentage of 1.23%. The simulation had 2 terminal forced trades, 32 target hits, and 16 stop hits. The return per symbol was: SPY 38.50%, IWM 35.87%, EFA 47.20%, EEM 51.00%, TLT -20.11%, IEF -5.58%, GLD -4.15%, DBC 7.72%, and VNQ 4.05%. The mean daily basis points, a measure of daily return, was 1.74. This was a preregistered primary simulation, and others were not selected for sensitivity testing. A snapshot was taken on September 11, with 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, and a 6 ATR target, with a daily proxy and a 126 cap. The raw JSON and CSV data were retained, and all results were recorded from 2352 to 63. A news scan found the headline was not available, this is a raw script output. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend150 2023-01-03 slip5 = 21.820604860887926
Quant replication2026-09-11 00:11
A trading simulation started on 2023-01-03 and ended on 2026-09-10 with a lookback period of 150 days. The simulation had a slippage of 5 basis points, which is a measure of the difference between expected and actual trade prices. The total return was 21.82%, the maximum drawdown, or largest loss from a peak, was 4.65%, and the compound annual growth rate, or CAGR, was 5.52%. The win rate, or percentage of successful trades, was 34.43%. There were 183 trades in the sample, with a mean trade percentage of 0.95%, a mean win percentage of 4.93%, and a mean loss percentage of 1.13%. The simulation had 2 terminal forced trades, 32 target hits, and 16 stop hits. The return per symbol was: SPY 41.01%, IWM 39.44%, EFA 51.69%, EEM 58.11%, TLT -14.66%, IEF -0.54%, GLD -2.61%, DBC 11.90%, and VNQ 12.04%. The mean daily basis points, a measure of daily return, was 2.17. This was a preregistered primary simulation, and others were not selected for sensitivity testing. A snapshot was taken on September 11, with 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, a 6 ATR target, and a 126 cap, using daily proxy data. The raw JSON and CSV data were retained, and all results were recorded from 2352 to 63. A news scan found the headline, but it is not provided as the source is a raw JSON output. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend150 2016-12-29 slip15 = -9.3910764541974
Quant replication2026-09-11 00:11
A trading fund manager's notes show a trend analysis from December 29, 2016, to December 30, 2022, with a 150-day lookback period and 15 basis points of slippage. The return was -9.39%, with a drawdown of 11.50% and a compound annual growth rate of -1.63%. The win rate was 19.75% out of 243 samples. The average trade loss was -0.31%, with an average win of 5.35% and an average loss of -1.71%. There were 3 terminal forced trades, 33 target hits, and 37 stop hits. Returns per symbol were: SPY -14.59%, IWM -18.04%, EFA 0.97%, EEM 2.14%, TLT -5.23%, IEF -5.57%, GLD -5.94%, DBC -19.15%, and VNQ -19.11%. The average daily return was -0.63 basis points. A preregistered primary strategy, labeled primary200, was used, but sensitivity analysis for other strategies was not selected. A snapshot from September 11 showed 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, and a 6 ATR target with a 126 cap, using daily proxy data. The raw JSON and CSV data were retained, and all results were recorded from 2352 to 63. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend150 2016-12-29 slip5 = -4.423529103423462
Quant replication2026-09-11 00:11
A trading fund manager's report shows results from December 29, 2016, to December 30, 2022, using a 150-day lookback period, with a 5 basis point slippage, which is a small cost when buying or selling. The return was -4.42%, and the drawdown, or largest loss, was 9.93%. The compound annual growth rate, or CAGR, a measure of how much the investment grew per year, was -0.75%. The win rate, or percentage of successful trades, was 21.40%. There were 243 trades in the sample. The average trade lost 0.11%, the average winning trade gained 5.14%, and the average losing trade lost 1.55%. The strategy was stopped 3 times, and 33 target prices were hit, while 37 stop-loss prices were hit. Returns for specific stocks were: SPY -10.75%, IWM -11.39%, EFA 4.88%, EEM 5.89%, TLT -0.57%, IEF -0.93%, GLD 1.49%, DBC -14.67%, and VNQ -13.77%. The average daily return was -0.28 basis points. This was a preregistered primary test, and sensitivity tests for other parameters were not selected. The test used a snapshot from September 11, with 9 equal parts, a 1 basis point fee per side, a 2 ATR stop, and a 6 ATR target, with a 126 cap, and daily proxy. The raw JSON and CSV data were retained, and all results were recorded from 2352 to 63. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend200 2023-01-03 slip15 = 6.007583585127518
Quant replication2026-09-11 00:11
A trading strategy called QTR9110000 trend200 was tested from 2023-01-03 to 2026-09-10 with a lookback period of 200 days, a slippage of 15 basis points, which is a measure of the difference between expected and actual trade prices. The strategy had a return of 6.007583585127518 percent, a maximum drawdown of 6.072893587733386 percent, and a compound annual growth rate, or CAGR, of 1.6020809137860192 percent. The strategy was profitable 24.691358024691358 percent of the time, with 162 samples, or trades, and an average trade return of 0.2562178279009104 percent. The average winning trade was 5.416108585141148 percent, and the average losing trade was 1.4355496334893314 percent. The strategy had 27 target hits and 22 stop hits. The return for each symbol was: SPY 31.272070484275673 percent, IWM 31.205680328045627 percent, EFA 26.51437548689872 percent, EEM 14.862385099747444 percent, TLT -38.72252456967193 percent, IEF -11.934389193601714 percent, GLD 1.6505574515651489 percent, DBC -6.386220804456178 percent, VNQ 5.606317983344633 percent. The average daily return was 0.6502705381069532 basis points. This was a preregistered primary strategy, and the results were recorded from 2352 to 63. A snapshot was taken on September 11, with 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, and a 6 ATR target, with a 126 cap, and daily proxy. The raw JSON and CSV data were retained. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend200 2023-01-03 slip5 = 9.43696240334313
Quant replication2026-09-11 00:11
A trading simulation started on 2023-01-03 and ended on 2026-09-10 with a lookback period of 200 days. The simulation had a slippage of 5 basis points, which means a small cost of 0.05% when buying or selling. The total return was 9.44%, the maximum drawdown, or largest loss, was 5.24%, and the compound annual growth rate, a measure of return over time, was 2.49%. The win rate, or percentage of successful trades, was 28.40%. There were 162 trades in the sample. The average trade return was 0.46%, the average win was 4.91%, and the average loss was 1.31%. The simulation had 27 target hits and 22 stop hits. The return for each symbol was: SPY 33.39%, IWM 33.06%, EFA 27.79%, EEM 18.83%, TLT -33.09%, IEF -7.05%, GLD 2.88%, DBC -2.57%, and VNQ 11.69%. The average daily return in basis points, a unit of measurement for returns, was 0.99. This was a preregistered primary simulation, and other sensitivity tests were not selected. A snapshot was taken on September 11, with 9 equal parts of the portfolio, a fee of 1 basis point per side, a stop loss of 2 average true ranges, a target of 6 average true ranges, and a daily proxy. The raw data in JSON and CSV formats was retained, and all results were recorded from 2352 to 2363. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend200 2016-12-29 slip15 = -6.94928628651772
Quant replication2026-09-11 00:11
A trading fund manager's report shows results from December 29, 2016, to December 30, 2022. The report has a lookback period of 200 days and a slippage of 15 basis points. The return was -6.95% and the drawdown was 9.45%. The compound annual growth rate, or CAGR, which is a measure of how much an investment grows each year, was -1.19%. The win rate, which is the percentage of winning trades, was 26.75%. There were 228 samples, or trades, with a mean trade percentage of -0.24%. The mean win percentage was 3.59% and the mean loss percentage was -1.63%. There were 2 terminal forced trades, 27 target hits, and 35 stop hits. The return per symbol was: SPY 0.05%, IWM -13.34%, EFA -14.27%, EEM -6.14%, TLT -4.90%, IEF -6.50%, GLD 0.71%, DBC -14.74%, and VNQ -3.40%. The mean daily basis points, which is a measure of daily return, was -0.46. A news scan found the headline was not available, but the report also mentions a preregistered primary200 and sensitivity not selected. The report includes a snapshot from September 11 with 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, and a 6 ATR target with a 126 cap, using daily proxy. The raw JSON and CSV data are retained, and all results are recorded from 2352 to 63. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
QTR9110000 trend200 2016-12-29 slip5 = -2.1298547226452746
Quant replication2026-09-11 00:11
A trading fund manager's notes show the performance of a trend-following strategy, called trend200, from December 29, 2016, to December 30, 2022. The strategy had a 200-day lookback period, meaning it considered the last 200 days of price movements, and a 5 basis point slippage, which is the difference between the expected and actual price of a trade. The return was -2.13%, and the maximum drawdown, or largest loss from peak to trough, was 6.41%. The compound annual growth rate, or CAGR, was -0.36%. The strategy was successful 29.82% of the time, out of 228 samples. The average trade resulted in a loss of -0.04%, with an average win of 3.42% and an average loss of 1.51%. There were 2 terminal forced trades, 27 target hits, and 35 stop hits. The strategy performed differently for each symbol, with returns ranging from -9.83% for DBC to 8.01% for GLD. The average daily return was -0.13%. The notes also mention that the primary strategy, called primary200, was preregistered, and that sensitivity analysis was not performed for other strategies. A snapshot from September 11 shows the strategy's performance with 9 equal sleeves, a fee of 1 basis point per side, a 2 ATR stop, and a 6 ATR target, with a daily proxy and a 126 cap. The raw JSON and CSV data were retained, and all results were recorded from 2352 to 63. https://runway.adityaperswal.workers.dev/data/stock-bars ; research/results.json
AVAV QTR9110000 refreshed close signal unchanged from frozen vintage = 3.274345998731776
Quant replication2026-09-11 00:08
On September 11, at about 00:02 UTC, the daily retrieval was done, and the data was recomputed at 00:08. From September 4 to September 10, the closing prices of AVAV were 144.65 and 147.065, the highs were 148.015 and 159.24, the lows were 143.3 and 142.5, and the volumes were 1541303 and 8767657. For the same period, the closing prices of SPY were 770.19 and 757.83, the highs were 772.87 and 760.11, the lows were 769 and 756.64, and the volumes were 34119246 and 42761756. The 20-session average daily range, or ADR, for AVAV was 8.759085, and for SPY it was 4.36456. The difference in the ADR, also known as the EAR difference, was 3.274346 percentage points, which remained unchanged. This meant there was no new live entry. The daily vendor aggregates were kept separate from the actual intraday execution. https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=AVAV,SPY&timeframe=1Day&end=2026-09-10T23:00:00Z&feed=sip&adjustment=all
QTR9110000 daily panel integrity before trend backtest = 2687
Quant replication2026-09-11 00:06
A news scan found the headline 'QTR9110000 daily panel integrity before trend backtest'. The report fetched data on September 11 at approximately 00:02 UTC for 10 symbols: SPY, IWM, EFA, EEM, TLT, IEF, GLD, DBC, VNQ, and AVAV. Each symbol had 2687 bars of data, which fit on one page, with no next token. The data started on January 4, 2016, despite a request for 2015 data, and ended on September 10, 2026. There were no duplicate dates and no invalid open, high, low, or close prices, which are called OHLC for short, and the dates matched for all 9 exchange-traded funds, or ETFs, compared to SPY. The first 250 trading sessions were excluded for comparison purposes, so actual training started late in 2016. The data used a type of adjustment called SIP, or securities information processor, and vendor history, but did not include an independent audit of corporate actions, which are changes that affect stock prices. The selection of symbols excluded dead funds, or funds that no longer exist, and used a simulation that treated each ETF as a fraction of a whole, rather than a whole share. The raw data was saved in a file called SYMBOL.json and the code used to fetch the data was saved in a file called fetch.py, which can be used to reproduce the results. No missing observations were filled with zeros. https://runway.adityaperswal.workers.dev/data/stock-bars?timeframe=1Day&start=2015-01-01&end=2026-09-10T23:00:00Z&feed=sip&adjustment=all
FTK QTR9110000 unchanged deck forecast versus paused generation = 40
Quant replication2026-09-11 00:06
On September 11 at 00:05 UTC, an original PDF was extracted using a local tool called pypdf after a web extraction failed. Page 8 of the PDF forecasts that international chemistry sales will be greater than 40 million, compared to 12.6 million in the first half of the year. It also describes the primary fluid chemistry supplier's unconventional operations in Saudi Arabia and extended supply through mid-2027. A power slide in the PDF says that Phase 1 will have 12 megawatts and Q2 revenue of 5.9 million, and that initial generation has been paused due to infrastructure delays, while a Phase 2 extension is under discussion. These statements are from an unchanged deck from September 10, and do not indicate a new pause or restart tonight, or proof of receipt. The exact full hash is in 23004. Two cash definitions for AVAV and supplier expansion risks are being retained separately. https://ir.flotekind.com/image/Lake-Street-Investor-Presentation.pdf
NewerPage 16 of 414Older