日内因子¶

在本笔记本中，我们使用Alphalens分析日内因子的表现，该因子每日计算，股票在开盘时买入并在收盘时卖出，不持有隔夜头寸。

导入与设置¶

[2]:

import warnings
warnings.filterwarnings('ignore')

[3]:

import alphalens
import pandas as pd

[4]:

%matplotlib inline

加载数据¶

以下是一个针对少量大盘股的投资组合中股票代码与行业板块的简单映射关系。

[5]:

sector_names = {
    0 : "information_technology",
    1 : "financials",
    2 : "health_care",
    3 : "industrials",
    4 : "utilities",
    5 : "real_estate",
    6 : "materials",
    7 : "telecommunication_services",
    8 : "consumer_staples",
    9 : "consumer_discretionary",
    10 : "energy"
}

ticker_sector = {
    "ACN" : 0, "ATVI" : 0, "ADBE" : 0, "AMD" : 0, "AKAM" : 0, "ADS" : 0, "GOOGL" : 0, "GOOG" : 0,
    "APH" : 0, "ADI" : 0, "ANSS" : 0, "AAPL" : 0, "AMAT" : 0, "ADSK" : 0, "ADP" : 0, "AVGO" : 0,
    "AMG" : 1, "AFL" : 1, "ALL" : 1, "AXP" : 1, "AIG" : 1, "AMP" : 1, "AON" : 1, "AJG" : 1, "AIZ" : 1, "BAC" : 1,
    "BK" : 1, "BBT" : 1, "BRK.B" : 1, "BLK" : 1, "HRB" : 1, "BHF" : 1, "COF" : 1, "CBOE" : 1, "SCHW" : 1, "CB" : 1,
    "ABT" : 2, "ABBV" : 2, "AET" : 2, "A" : 2, "ALXN" : 2, "ALGN" : 2, "AGN" : 2, "ABC" : 2, "AMGN" : 2, "ANTM" : 2,
    "BCR" : 2, "BAX" : 2, "BDX" : 2, "BIIB" : 2, "BSX" : 2, "BMY" : 2, "CAH" : 2, "CELG" : 2, "CNC" : 2, "CERN" : 2,
    "MMM" : 3, "AYI" : 3, "ALK" : 3, "ALLE" : 3, "AAL" : 3, "AME" : 3, "AOS" : 3, "ARNC" : 3, "BA" : 3, "CHRW" : 3,
    "CAT" : 3, "CTAS" : 3, "CSX" : 3, "CMI" : 3, "DE" : 3, "DAL" : 3, "DOV" : 3, "ETN" : 3, "EMR" : 3, "EFX" : 3,
    "AES" : 4, "LNT" : 4, "AEE" : 4, "AEP" : 4, "AWK" : 4, "CNP" : 4, "CMS" : 4, "ED" : 4, "D" : 4, "DTE" : 4,
    "DUK" : 4, "EIX" : 4, "ETR" : 4, "ES" : 4, "EXC" : 4, "FE" : 4, "NEE" : 4, "NI" : 4, "NRG" : 4, "PCG" : 4,
    "ARE" : 5, "AMT" : 5, "AIV" : 5, "AVB" : 5, "BXP" : 5, "CBG" : 5, "CCI" : 5, "DLR" : 5, "DRE" : 5,
    "EQIX" : 5, "EQR" : 5, "ESS" : 5, "EXR" : 5, "FRT" : 5, "GGP" : 5, "HCP" : 5, "HST" : 5, "IRM" : 5, "KIM" : 5,
    "APD" : 6, "ALB" : 6, "AVY" : 6, "BLL" : 6, "CF" : 6, "DWDP" : 6, "EMN" : 6, "ECL" : 6, "FMC" : 6, "FCX" : 6,
    "IP" : 6, "IFF" : 6, "LYB" : 6, "MLM" : 6, "MON" : 6, "MOS" : 6, "NEM" : 6, "NUE" : 6, "PKG" : 6, "PPG" : 6,
    "T" : 7, "CTL" : 7, "VZ" : 7,
    "MO" : 8, "ADM" : 8, "BF.B" : 8, "CPB" : 8, "CHD" : 8, "CLX" : 8, "KO" : 8, "CL" : 8, "CAG" : 8,
    "STZ" : 8, "COST" : 8, "COTY" : 8, "CVS" : 8, "DPS" : 8, "EL" : 8, "GIS" : 8, "HSY" : 8, "HRL" : 8,
    "AAP" : 9, "AMZN" : 9, "APTV" : 9, "AZO" : 9, "BBY" : 9, "BWA" : 9, "KMX" : 9, "CCL" : 9,
    "APC" : 10, "ANDV" : 10, "APA" : 10, "BHGE" : 10, "COG" : 10, "CHK" : 10, "CVX" : 10, "XEC" : 10, "CXO" : 10,
    "COP" : 10, "DVN" : 10, "EOG" : 10, "EQT" : 10, "XOM" : 10, "HAL" : 10, "HP" : 10, "HES" : 10, "KMI" : 10
}

YFinance数据下载¶

[6]:

import yfinance as yf
import pandas_datareader.data as web
yf.pdr_override()

tickers = list(ticker_sector.keys())
df = web.get_data_yahoo(tickers, start='2017-01-01',  end='2017-06-01')
df.index = pd.to_datetime(df.index, utc=True)

[*********************100%***********************]  182 of 182 completed

19 Failed downloads:
- CHK: Data doesn't exist for startDate = 1483250400, endDate = 1496293200
- BF.B: No data found for this date range, symbol may be delisted
- BRK.B: No data found, symbol may be delisted
- MON: Data doesn't exist for startDate = 1483250400, endDate = 1496293200
- CELG: No data found, symbol may be delisted
- APC: No data found, symbol may be delisted
- ARNC: Data doesn't exist for startDate = 1483250400, endDate = 1496293200
- CBG: No data found for this date range, symbol may be delisted
- CTL: No data found, symbol may be delisted
- BCR: No data found for this date range, symbol may be delisted
- DWDP: No data found, symbol may be delisted
- BHF: Data doesn't exist for startDate = 1483250400, endDate = 1496293200
- GGP: No data found for this date range, symbol may be delisted
- CXO: No data found, symbol may be delisted
- HCP: No data found, symbol may be delisted
- DPS: No data found for this date range, symbol may be delisted
- AGN: No data found, symbol may be delisted
- BBT: No data found, symbol may be delisted
- BHGE: No data found, symbol may be delisted

[7]:

df = df.stack()
df.index.names = ['date', 'asset']
df.info()

<class 'pandas.core.frame.DataFrame'>
MultiIndex: 16789 entries, (Timestamp('2017-01-03 00:00:00+0000', tz='UTC'), 'A') to (Timestamp('2017-05-31 00:00:00+0000', tz='UTC'), 'XOM')
Data columns (total 6 columns):
 #   Column     Non-Null Count  Dtype
---  ------     --------------  -----
 0   Adj Close  16789 non-null  float64
 1   Close      16789 non-null  float64
 2   High       16789 non-null  float64
 3   Low        16789 non-null  float64
 4   Open       16789 non-null  float64
 5   Volume     16789 non-null  float64
dtypes: float64(6)
memory usage: 842.6+ KB

因子计算¶

我们的示例因子根据股票的隔夜价格缺口（昨日收盘价至今日开盘价）对股票进行排序。我们将观察该因子是否具有阿尔法收益，还是纯粹的噪音。

[8]:

available_tickers = df.index.unique('asset')
ticker_sector = {k: v for k, v in ticker_sector.items() if k in available_tickers}

[9]:

today_open = df.loc[:, 'Open'].unstack('asset')
today_close = df.loc[:, 'Close'].unstack('asset')
yesterday_close = today_close.shift(1)

[10]:

factor = (today_open - yesterday_close) / yesterday_close

传递给alphalens的定价数据应包含资产的入场价格，因此必须反映在给定时间戳观察到因子值后的下一个可用价格。这些价格不得用于计算该时间点的因子值。务必反复检查，确保您的研究没有引入前瞻性偏差。

价格数据还必须包含资产的退出价格，对于周期1将使用下一个时间戳的价格，对于周期2将使用2个时间戳后的价格，依此类推。

对于计算因子时使用的时间频率没有限制/假设，对于交易因子的具体时间也没有限制（开盘交易、收盘交易或日内交易），唯一的要求是根据上述规则确保因子和价格数据框正确对齐。

在我们的示例中，我们希望以开盘价买入股票，因此需要与因子值时间戳完全匹配的开盘价数据；同时我们计划以收盘价卖出股票，所以还需添加收盘价数据——这些收盘价将用于计算周期1的远期收益，因为它们正好出现在因子值时间戳之后。因此Alphalens计算的收益率将基于资产从开盘价到收盘价的价差。

如果我们有其他价格数据，就可以计算其他时间段的收益率，例如开盘后一小时、两小时等。我们可以在开盘价之后直接添加这些价格，并指示Alphalens计算1、2、3...等多个周期的收益率，而不仅限于本例中的周期1。

数据格式化¶

时间调整¶

[11]:

# Fix time as Yahoo doesn't set it
today_open.index += pd.Timedelta('9h30m')
today_close.index += pd.Timedelta('16h')
# pricing will contain both open and close
pricing = pd.concat([today_open, today_close]).sort_index()

[12]:

pricing.head()

[12]:

资产	A	AAL	AAP	AAPL	ABBV	ABC	ABT	ACN	ADBE	ADI	...	NUE	PCG	PKG	PPG	SCHW	STZ	T	VZ	XEC	XOM
日期
2017-01-03 09:30:00+00:00	45.930000	47.279999	170.779999	28.950001	62.919998	78.510002	38.630001	117.379997	103.430000	72.599998	...	59.740002	60.810001	85.160004	95.430000	40.049999	155.009995	42.689999	53.959999	137.529999	90.940002
2017-01-03 16:00:00+00:00	46.490002	46.299999	170.600006	29.037500	62.410000	82.610001	39.049999	116.459999	103.480003	72.510002	...	59.610001	60.369999	85.000000	95.250000	40.200001	154.750000	43.020000	54.580002	138.789993	90.889999
2017-01-04 09:30:00+00:00	46.930000	46.630001	170.369995	28.962500	62.639999	82.599998	39.060001	116.910004	103.739998	72.769997	...	59.759998	60.610001	85.440002	95.709999	40.400002	157.149994	42.939999	54.549999	138.479996	91.120003
2017-01-04 16:00:00+00:00	47.099998	46.700001	172.000000	29.004999	63.290001	84.660004	39.360001	116.739998	104.139999	72.360001	...	61.250000	60.590000	86.370003	97.269997	41.220001	157.990005	42.770000	54.520000	138.500000	89.889999
2017-01-05 09:30:00+00:00	47.049999	46.520000	170.869995	28.980000	63.380001	84.379997	39.240002	116.980003	104.129997	72.410004	...	61.119999	60.660000	86.370003	96.459999	40.970001	150.550003	42.849998	54.779999	138.500000	90.190002

5行 × 163列

对齐因子与价格数据¶

[13]:

# Align factor to open price
factor.index += pd.Timedelta('9h30m')
factor = factor.stack()
factor.index = factor.index.set_names(['date', 'asset'])

[14]:

factor.unstack().head()

[14]:

资产	A	AAL	AAP	AAPL	ABBV	ABC	ABT	ACN	ADBE	ADI	...	NUE	PCG	PKG	PPG	SCHW	STZ	T	VZ	XEC	XOM
日期
2017-01-04 09:30:00+00:00	0.009464	0.007127	-0.001348	-0.002583	0.003685	-0.000121	0.000256	0.003864	0.002513	0.003586	...	0.002516	0.003976	0.005176	0.004829	0.004975	0.015509	-0.001860	-0.000550	-0.002234	0.002531
2017-01-05 09:30:00+00:00	-0.001062	-0.003854	-0.006570	-0.000862	0.001422	-0.003307	-0.003049	0.002056	-0.000096	0.000691	...	-0.002122	0.001155	0.000000	-0.008327	-0.006065	-0.047092	0.001870	0.004769	0.000000	0.003337
2017-01-06 09:30:00+00:00	0.001934	-0.000872	-0.003258	0.001458	0.001725	-0.001793	0.000000	0.000000	0.000661	0.003646	...	-0.000328	-0.003304	0.000469	0.001569	0.008055	0.001635	-0.015709	-0.017753	0.001784	0.002710
2017-01-09 09:30:00+00:00	0.000417	-0.004328	0.002535	0.000339	0.000157	-0.002359	0.000245	-0.001376	-0.003139	0.000559	...	0.009786	-0.000327	-0.003440	-0.005649	-0.005578	0.003747	-0.000726	-0.000751	-0.011285	-0.003164
2017-01-10 09:30:00+00:00	0.004155	-0.001699	-0.004896	-0.001849	-0.002492	-0.004446	0.001718	-0.000522	0.000000	-0.000973	...	0.012743	0.000498	-0.000798	-0.002906	0.001702	-0.001797	-0.003431	0.000190	0.004664	0.001494

5行 × 163列

运行Alphalens¶

周期1将显示从市场开盘到收盘的回报，而周期2将显示从今日开盘到明日开盘的回报

获取Alphalens输入¶

[15]:

non_predictive_factor_data = alphalens.utils.get_clean_factor_and_forward_returns(factor,
                                                                                  pricing,
                                                                                  periods=(1,2),
                                                                                  groupby=ticker_sector,
                                                                                  groupby_labels=sector_names)

Dropped 2.9% entries from factor data: 1.0% in forward returns computation and 2.0% in binning phase (set max_loss=0 to see potentially suppressed Exceptions).
max_loss is 35.0%, not exceeded: OK!

收益分析报告¶

[16]:

alphalens.tears.create_returns_tear_sheet(non_predictive_factor_data)

Returns Analysis

	6小时30分钟	1天
年化Alpha	0.324	-0.046
beta	0.177	0.179
最高分位数的平均周期回报率 (基点)	-7.776	-2.096
底部十分位数平均周期收益率 (基点)	-0.445	0.697
周期平均价差(基点)	-7.331	-2.795

<Figure size 432x288 with 0 Axes>

../_images/notebooks_intraday_factor_30_3.png

[17]:

alphalens.tears.create_event_returns_tear_sheet(non_predictive_factor_data, pricing);

<Figure size 432x288 with 0 Axes>

../_images/notebooks_intraday_factor_31_1.png

概述事件研究