加载中...
AI 原生量化投研平台AI-native quantitative research platform
正在加载…Loading…
页面仍在加载，可以继续等待，或重新加载。The page is still loading. You can keep waiting or reload.
部分页面资源未能加载，请重新加载。Some page resources could not load. Please reload.
设备当前处于离线状态，恢复连接后可以继续等待或重新加载。Your device is offline. Once connected, you can keep waiting or reload.
因子逻辑
对20日日收益相对均值的偏离求和，按偏度公式的系数和分母做标准化并乘-20，记法上是收益偏度的变形；原公式存在书写歧义，值大致与收益偏度方向相反。
因子公式
-20*（20-1 ）^1.5*SUM(CLOSE/DELAY(CLOSE,1)-1-MEAN(CLOSE/DELAY(CLOSE,1)-1,20),20)/((20-1)*(20-2)(SUM((CLOSE/DELAY(CLOSE,1),20)^2,20))^1.5)BigQuant SQL
SELECT date, instrument, -20*(19)^1.5*m_sum(close/m_lag(close, 1)-1-m_avg(close/m_lag(close, 1)-1, 20),20)/(19*18*(m_sum(close/m_lag(close, 1), 20)^2)^1.5) AS _factor_temp, 1 * _factor_temp AS factorFROM cn_stock_bar1dQUALIFY COLUMNS(*) IS NOT NULLORDER BY date, instrumentbq examples new strategies/multi_factor.py bqf_10397_strategy.py --name 'Alpha191 #166 选股' \ --set 'TABLES=["cn_stock_factors_alpha_191"]' --set 'FACTORS=["alpha_166"]' --set 'HOLD=30' \ --set 'EVERY=20' --set 'BENCHMARK="000300.SH"' --set 'START="2021-01-01"' \ --set 'END="2026-09-18"' \ && python bqf_10397_strategy.pySELECT date, instrument, alpha_166 FROM cn_stock_factors_alpha_191带入当前因子，在 Cowork 中继续研究。
from bigquant import factor
values = factor.values(["BQF_10397"], "2021-01-01", "2026-09-18")report = factor.analyze( values, start_date="2021-01-01", end_date="2026-09-18", universe=("all", "hs300"), neutralize=(None, "size+industry"), depth="deep",)report.render()