Quantopian 大市值科技股市值加权回测

Quantopian 大市值科技股市值加权回测

1.源码

from quantopian.pipeline import Pipelinefrom quantopian.algorithm import attach_pipeline, pipeline_outputfrom quantopian.pipeline.data.builtin import USEquityPricingfrom quantopian.pipeline.factors import SimpleMovingAveragefrom quantopian.pipeline.classifiers.morningstar import Sectorfrom quantopian.pipeline.data import morningstar as morningstarfrom quantopian.pipeline.filters.morningstar import Q500US, Q1500USfrom quantopian.pipeline.factors.morningstar import MarketCapfrom quantopian.pipeline.data import Fundamentals import quantopian.algorithm as algo def make_pipeline(): q500 = Q500US() mktcap = MarketCap() sector = morningstar.asset_classification.morningstar_sector_code.latest sectors_311 = sector.element_of([102, 311]) mkt_cap = morningstar.valuation.market_cap.latest growth = Fundamentals.revenue_growth.latest forward_pe_ratio = Fundamentals.forward_pe_ratio.latest current_pe = Fundamentals.pe_ratio.latest pipe = Pipeline() attach_pipeline(pipe, 'pipeline_tutorial') pipe.add(Sector(), 'Sector') pipe.add(mktcap, 'mktcap') pipe.add(growth, 'growth') pipe.add(forward_pe_ratio, 'forward_pe_ratio') pipe.add(current_pe, 'current_pe') # log.info(current_pe / forward_pe_ratio) pipe.set_screen(sectors_311 & q500) # pipe.set_screen(q500) return pipe def initialize(context): set_benchmark(symbol('QQQ')) algo.schedule_function( rebalance, algo.date_rules.month_start(), algo.time_rules.market_open(hours=1), ) algo.attach_pipeline(make_pipeline(), 'pipeline') #购买持仓数量 context.buy_stock_count = 10 #杠杆大小 context.leverage = 1.0 context.buy_stock_percent = float(context.leverage / context.buy_stock_count) cash = context.portfolio.cash context.purchase_value = cash / context.buy_stock_count context.stocks_sold = [] def before_trading_start(context, data): output = pipeline_output('pipeline_tutorial') context.my_short_universe = output.sort('mktcap', ascending=True).iloc[0:context.buy_stock_count] context.my_long_universe = output.sort('mktcap', ascending=False).iloc[0:context.buy_stock_count] def rebalance(context, data): """ Execute orders according to our schedule_function() timing. """ if True: # purchase_value = cash / left_buy_count for stock in context.my_long_universe.index: buy_stock_percent = context.my_long_universe["mktcap"][stock] / context.my_long_universe["mktcap"].sum() if stock != symbol('QQQ'): order_target_percent(stock, buy_stock_percent * context.leverage) for stock in context.my_short_universe.index: if stock != symbol('QQQ'): order_target_percent(stock, -context.buy_stock_percent * 0) for stock in context.portfolio.positions: if stock != symbol('QQQ'): if stock not in context.my_long_universe.index and stock not in context.my_long_universe.index: order_target_value(stock, 0) pass

2.回测结果：

比上​​《Quantopian 做多大市值科技和消费周期股，做空小市值和消费周期股回测》​​的要好些，特别是这2年。

上一篇的回测结果：

3.结论

我认为，市值加权是长期有效的策略，有点进化论的思想在里头。好的东西一定会变大。但是也需要定期重新寻找。所以如果你的组合跑不赢大盘，试着用点仓位以市值加权方式配置点指数前10大持仓。至少能跟得上指数。

版权声明：本文内容由网络用户投稿，版权归原作者所有，本站不拥有其著作权，亦不承担相应法律责任。如果您发现本站中有涉嫌抄袭或描述失实的内容，请联系我们jiasou666@gmail.com 处理，核实后本网站将在24小时内删除侵权内容。