请问如何优化因子提高策略收益

策略分享
标签: #<Tag:0x00007fb3d8208580>

(chenys101) #1

16年收益300%,17以后一直下降最终收益130%,试了很多因子都比这个差,不知道是哪里出问题了

克隆策略
In [49]:
# 本代码由可视化策略环境自动生成 2019年3月1日 14:18
# 本代码单元只能在可视化模式下编辑。您也可以拷贝代码,粘贴到新建的代码单元或者策略,然后修改。


class conf:
    start_date = '2012-12-16' #2016-04-20   '2016-10-27'  2017-03-16  2016-02-26
    split_date = '2015-12-31'
    start_trade = '2016-01-01'
    end_date='2019-03-05'

    # 特征 https://bigquant.com/docs/data_features.html,你可以通过表达式构造任何特征
    features1 = [
             #量价因子
        #过去i个交易日平均净主动买入额排名
        'rank_avg_mf_net_amount_2',
        'return_2',
        #营业收入同比增长率,升序百分比排名
        'rank_fs_operating_revenue_yoy_0',
        #过去i个交易日平均净主动买入额
        'avg_mf_net_amount_2/avg_mf_net_amount_20',
        #过去30天跌幅
        'close_0/shift(open_0,30)-1',
        #过去i个交易日的平均交易额,0表示今日
        'avg_amount_0/avg_amount_5',
        #市净率 (LF)
        'pb_lf_0',
        'ps_ttm_0',
        
    ]
    extra_fields = ['st_status_0',       
                    'price_limit_status_0',
                   ]
    
m1 = M.instruments.v2(
    start_date=conf.start_date,
    end_date=conf.split_date,
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m2 = M.advanced_auto_labeler.v2(
    instruments=m1.data,
    label_expr="""# #号开始的表示注释
# 0. 每行一个,顺序执行,从第二个开始,可以使用label字段
# 1. 可用数据字段见 https://bigquant.com/docs/data_history_data.html
#   添加benchmark_前缀,可使用对应的benchmark数据
# 2. 可用操作符和函数见 `表达式引擎 <https://bigquant.com/docs/big_expr.html>`_

# 计算收益:5日收盘价(作为卖出价格)除以明日开盘价(作为买入价格)
shift(close, -5) / shift(open, -1)

# 极值处理:用1%和99%分位的值做clip
clip(label, all_quantile(label, 0.01), all_quantile(label, 0.99))

# 将分数映射到分类,这里使用20个分类
all_wbins(label, 20)

# 过滤掉一字涨停的情况 (设置label为NaN,在后续处理和训练中会忽略NaN的label)
where(shift(high, -1) == shift(low, -1), NaN, label)
""",
    start_date='',
    end_date='',
    benchmark='000300.SHA',
    drop_na_label=True,
    cast_label_int=True
)
m33 = M.input_features.v1(
    features=conf.features1
)

m3 = M.input_features.v1(
    features=conf.features1+conf.extra_fields
)



m4 = M.general_feature_extractor.v6(
    instruments=m1.data,
    features=m3.data,
    start_date='',
    end_date='',
    before_start_days=60
)

m5 = M.derived_feature_extractor.v2(
    input_data=m4.data,
    features=m3.data,
    date_col='date',
    instrument_col='instrument'
)

m7 = M.join.v3(
    data1=m2.data,
    data2=m5.data,
    on='date,instrument',
    how='inner',
    sort=False
)

# Python 代码入口函数,input_1/2/3 对应三个输入端,data_1/2/3 对应三个输出端
def m15_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码

    df = input_1.read_df()
    ins= m1.data.read_pickle()['instruments']
    start= m1.data.read_pickle()['start_date']
    end= m1.data.read_pickle()['end_date']
    
    df1 = D.features(ins, start, end, fields=['mf_net_pct_main_0','market_cap_float_0'])#,'mf_net_amount_0'

    df_final=pd.merge(df,df1,on=['date','instrument'])
    df_final = df_final[df_final['mf_net_pct_main_0'] > 0.1]
    df_final = df_final[df_final['price_limit_status_0'] == 2]
    df_final = df_final[df_final['market_cap_float_0'] < 10000000000]

    
    print('-----------------------------------',len(df_final))
    data_1 = DataSource.write_df(df_final)  

    return Outputs(data_1=data_1, data_2=None, data_3=None)

m15 = M.cached.v3(
    input_1=m7.data,
    run=m15_run_bigquant_run
)

m13 = M.dropnan.v1(
    input_data=m15.data_1
)

m6 = M.stock_ranker_train.v5(
    training_ds=m13.data,
    features=m33.data,
    learning_algorithm='排序',
    number_of_leaves=30,
    minimum_docs_per_leaf=1000,
    number_of_trees=20,
    learning_rate=0.1,
    max_bins=1023,
    feature_fraction=1,
    m_lazy_run=False
)

m9 = M.instruments.v2(
    start_date=T.live_run_param('trading_date', conf.start_trade),
    end_date=T.live_run_param('trading_date', conf.end_date),
    market='CN_STOCK_A',
    instrument_list='',
    max_count=0
)

m10 = M.general_feature_extractor.v6(
    instruments=m9.data,
    features=m3.data,
    start_date='',
    end_date='',
    before_start_days=60
)

m11 = M.derived_feature_extractor.v2(
    input_data=m10.data,
    features=m3.data,
    date_col='date',
    instrument_col='instrument'
)

# Python 代码入口函数,input_1/2/3 对应三个输入端,data_1/2/3 对应三个输出端
def m16_run_bigquant_run(input_1, input_2, input_3):
    # 示例代码如下。在这里编写您的代码

    df = input_1.read_df()
    ins= m9.data.read_pickle()['instruments']
    start= m9.data.read_pickle()['start_date']
    end= m9.data.read_pickle()['end_date']
    df1 = D.features(ins, start, end, fields=['mf_net_pct_main_0','market_cap_float_0'])#,'mf_net_amount_0'

    
    df_final=pd.merge(df,df1,on=['date','instrument'])
    df_final = df_final[df_final['mf_net_pct_main_0'] > 0.1]
    df_final = df_final[df_final['price_limit_status_0'] == 2]
    df_final = df_final[df_final['market_cap_float_0'] < 10000000000]
    df_final = df_final[df_final['st_status_0'] == 0]
    print('-----------------------------------',len(df_final))
    data_1 = DataSource.write_df(df_final)


    return Outputs(data_1=data_1, data_2=None, data_3=None)
m16 = M.cached.v3(
    input_1=m11.data,
    run=m16_run_bigquant_run
)

m14 = M.dropnan.v1(
    input_data=m16.data_1
)

m8 = M.stock_ranker_predict.v5(
    model=m6.model,
    data=m14.data,
    m_lazy_run=False
)

# 回测引擎:每日数据处理函数,每天执行一次
def m12_handle_data_bigquant_run(context, data):
    # 按日期过滤得到今日的预测数据
    ranker_prediction = context.ranker_prediction[
        context.ranker_prediction.date == data.current_dt.strftime('%Y-%m-%d')]

    # 1. 资金分配
    # 平均持仓时间是hold_days,每日都将买入股票,每日预期使用 1/hold_days 的资金
    # 实际操作中,会存在一定的买入误差,所以在前hold_days天,等量使用资金;之后,尽量使用剩余资金(这里设置最多用等量的1.5倍)
    is_staging = context.trading_day_index < context.options['hold_days'] # 是否在建仓期间(前 hold_days 天)
    cash_avg = context.portfolio.portfolio_value / context.options['hold_days']
    cash_for_buy = min(context.portfolio.cash, (1 if is_staging else 1.5) * cash_avg)
    cash_for_sell = cash_avg - (context.portfolio.cash - cash_for_buy)
    positions = {e.symbol: p.amount * p.last_sale_price
                 for e, p in context.perf_tracker.position_tracker.positions.items()}

    # 2. 生成卖出订单:hold_days天之后才开始卖出;对持仓的股票,按机器学习算法预测的排序末位淘汰
    if not is_staging and cash_for_sell > 0:
        equities = {e.symbol: e for e, p in context.perf_tracker.position_tracker.positions.items()}
        instruments = []
        for x in equities :
            if not context.has_unfinished_sell_order(equities[x]):
                instruments.append(x)
                
        for instrument in instruments:
            context.order_target(context.symbol(instrument), 0)
            cash_for_sell -= positions[instrument]
            if cash_for_sell <= 0:
                break

    # 3. 生成买入订单:按机器学习算法预测的排序,买入前面的stock_count只股票
    buy_cash_weights = context.stock_weights
    buy_instruments = list(ranker_prediction.instrument[:len(buy_cash_weights)])
    max_cash_per_instrument = context.portfolio.portfolio_value * context.max_cash_per_instrument
    for i, instrument in enumerate(buy_instruments):
        cash = cash_for_buy * buy_cash_weights[i]
        if cash > max_cash_per_instrument - positions.get(instrument, 0):
            # 确保股票持仓量不会超过每次股票最大的占用资金量
            cash = max_cash_per_instrument - positions.get(instrument, 0)
        if cash > 0:
            current_price = data.current(context.symbol(instrument), 'price')
            amount = math.floor(cash / current_price - cash / current_price % 100)
            context.order(context.symbol(instrument), amount)

# 回测引擎:准备数据,只执行一次
def m12_prepare_bigquant_run(context):
    pass

# 回测引擎:初始化函数,只执行一次
def m12_initialize_bigquant_run(context):
    # 加载预测数据
    context.ranker_prediction = context.options['data'].read_df()

    # 系统已经设置了默认的交易手续费和滑点,要修改手续费可使用如下函数
    context.set_commission(PerOrder(buy_cost=0.0003, sell_cost=0.0013, min_cost=5))
    # 预测数据,通过options传入进来,使用 read_df 函数,加载到内存 (DataFrame)
    # 设置买入的股票数量,这里买入预测股票列表排名靠前的5只
    stock_count = 1
    # 每只的股票的权重,如下的权重分配会使得靠前的股票分配多一点的资金,[0.339160, 0.213986, 0.169580, ..]
    context.stock_weights = T.norm([1 / math.log(i + 2) for i in range(0, stock_count)])
    # 设置每只股票占用的最大资金比例
    context.max_cash_per_instrument = 0.7
    context.options['hold_days'] = 1

m12 = M.trade.v3(
    instruments=m9.data,
    options_data=m8.predictions,
    start_date='',
    end_date='',
    handle_data=m12_handle_data_bigquant_run,
    prepare=m12_prepare_bigquant_run,
    initialize=m12_initialize_bigquant_run,
    volume_limit=0.025,
    order_price_field_buy='open',
    order_price_field_sell='close',
    capital_base=100000,
    benchmark='000300.SHA',
    auto_cancel_non_tradable_orders=True,
    data_frequency='daily',
    price_type='真实价格',
    plot_charts=True,
    backtest_only=False,
    amount_integer=False
)
  • 收益率130.19%
  • 年化收益率31.32%
  • 基准收益率2.28%
  • 阿尔法0.31
  • 贝塔0.9
  • 夏普比率0.85
  • 胜率0.5
  • 盈亏比1.12
  • 收益波动率36.26%
  • 信息比率0.06
  • 最大回撤56.94%

(华尔街的猫) #2

大盘是一个很重要的因素啊,16年大盘涨的比17年好


(chenys101) #3

有没有办法大盘跌,策略不跌啊


(think) #4

能做到这个的,就是超级好策略了,加油


(chenys101) #5

我的描述有误,是不希望策略走向跟大盘走,不能熊市亏太多,毕竟穷哈哈


(冰柠檬) #6

可以做个大盘择时 llt和rsrs 都挺不错的 都可以有效规避熊市 不过rsrs对牛市不太敏感 现在 好的大盘择时策略太少了。


(chenys101) #7

谢谢你的建议,我有空研究下哈