The QPAIR91004 robust row5 uses a Frozenbase60/2, which was trained from 2017 to 2019 and held out from 2020 to 2025, with a separate period in 2026. The methods used include bootstrap, regression, and simulation, which are found in the archived code, but there is no independent prospective evidence. The results are as follows: the era is holdout, the mode is long, the bp is 6, the months are 72, the block months are 3, the replicates are 10000, and the ci pct is between -0.5781541926096271 and 0.36928911705407547.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 robust row4
Quant replication2026-09-10 04:18
The QPAIR91004 robust row4 uses a dataset divided into training from 2017 to 2019, holdout from 2020 to 2025, and a separate set for 2026. It employs bootstrap, regression, and simulation methods found in archived code, but lacks independent prospective evidence. The configuration includes a training era, swing mode, 16 basis points, 36 months, and blocks of 3 months, with 10,000 replicates, and a confidence interval percentage of -0.4589240526102851 to -0.23863926749952258.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 robust row3
Quant replication2026-09-10 04:18
The QPAIR91004 robust row3 uses a Frozenbase60/2 method, which was trained from 2017 to 2019, held out from 2020 to 2025, and has a separate period in 2026. This method uses bootstrap, regression, and simulation techniques, which are archived in code, but it does not have independent prospective evidence. The results show an era of training, in swing mode, with 6 basis points, over 36 months, in blocks of 3 months, with 10,000 replicates, and a confidence interval percentage of -0.25569032051377794 to -0.03914910342951025.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 robust row2
Quant replication2026-09-10 04:18
The QPAIR91004 robust row2 uses a specific setup for its analysis, which includes Frozenbase60 divided by 2, with training data from 2017 to 2019, holdout data from 2020 to 2025, and a separate dataset for 2026. The methods used include bootstrap, regression, and simulation, which are found in the archived code, but there is no independent prospective evidence to support the results. The analysis has the following parameters: the era is set to train, the mode is set to difference, the number of basis points is 6, the number of months is 36, the block months is 3, and the number of replicates is 10000, with a confidence interval percentage of -0.02106834482306793 to 0.04549155402183688.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 robust row1
Quant replication2026-09-10 04:18
The QPAIR91004 robust row1 uses a dataset divided into Frozenbase60/2, with training data from 2017 to 2019, holdout data from 2020 to 2025, and a separate dataset for 2026. The methods used include bootstrap, regression, and simulation, which are found in the archived code, but there is no independent prospective evidence. The parameters used are: era is set to train, mode is set to long, bp is 6, months is 36, block months is 3, replicates is 10000, and the confidence interval percentage is between -0.3516100841644501 and 0.27512676994006013.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 inference row15 = 0.013711851529411798
Quant replication2026-09-10 04:18
The QPAIR91004 inference row 15 uses a specific data setup: Frozenbase60 divided by 2, trained on data from 2017 to 2019, held out from 2020 to 2025, and separate data for 2026. It employs bootstrap, regression, and simulation methods found in archived code, but this is not independent prospective evidence. The results for the 2026 era, in long minus partner mode, with a sample size of 85 and 9 clusters, show a mean trade percentage of 0.0137, a cluster confidence interval percentage between -0.1597 and 0.1746, and a trade return correlation of 0.9809.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 inference row14 = -0.34255364211764705
Quant replication2026-09-10 04:18
The trading model was trained on data from 2017 to 2019 and held out from 2020 to 2025, with a separate test in 2026. The model used bootstrap, regression, and simulation methods, which are archived in the code, but this is not independent prospective evidence. The results for 2026 show a mode of swing, with 16 basis points, and 9 clusters. The mean trade percentage is -0.34255364211764705, and the cluster confidence interval percentage is between -0.6230454682891414 and -0.045668734305555556. The partner alpha annual percentage is -3.9392585894466707, with a partner beta of 0.14011696448521246, and a partner alpha t of -2.364083502292945. The spy alpha annual percentage is -4.2735354805875705, with a spy beta of 0.04141893142583724, and a spy alpha t of -2.299066069105291.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 inference row13 = -0.14303908482352948
Quant replication2026-09-10 04:18
The system used a frozen base from 60 percent of the data, trained on 2017 to 2019, held out 2020 to 2025, and had a separate set for 2026. It used bootstrap, regression, and simulation methods found in archived code, but this is not independent prospective evidence. The results for 2026 include a mode of swing, 6 basis points, and 9 clusters. The mean trade percentage was -0.14303908482352948 percent. The cluster confidence interval percentage was between -0.4171793766310903 percent and 0.15614330273958316 percent. The partner alpha annual percentage was -2.158442797620927 percent and the partner beta was 0.14141396189263025. The partner alpha t-value was -1.394034923480547. The spy alpha annual percentage was -2.2126160918552107 percent and the spy beta was 0.042520290731805604. The spy alpha t-value was -1.2756098728276464.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all
QPAIR91004 inference row12 = 0.29037093
Quant replication2026-09-10 04:18
The QPAIR91004 inference row 12 uses a Frozenbase60/2 model, trained from 2017 to 2019 and held out from 2020 to 2025, with a separate dataset for 2026. The methods used include bootstrap, regression, and simulation, which are archived in the code, but there is no independent prospective evidence. The results for 2026 show a mode of long, with 16 basis points, divided into 9 clusters, and a mean trade percentage of 0.29037093. The cluster confidence interval percentage ranges from -0.19493983450980398 to 0.9019849711696117. The partner alpha annual percentage is 0.5585778231939095, with a partner beta of 0.8490021389973837, and a partner alpha t of 0.611899412056341. In comparison, the spy alpha annual percentage is -0.5406745877313621, with a spy beta of 0.19957072406236348, and a spy alpha t of -0.14918005933152123.
https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all