Basic strategy temp

給你更多的策略模板，幫助你有更多靈感！

colab 範例檔案連結

安裝套件¶

In [1]:

!pip install finlab > log.txt
!pip install ta-lib-bin > log.txt

!pip install finlab > log.txt !pip install ta-lib-bin > log.txt

股價創新高動能¶

In [12]:

from finlab.backtest import sim
close = data.get("price:收盤價")
position = (close == close.rolling(250).max())
report = sim(position, resample="M", name="創年新高策略",upload=False)
report.display()

from finlab.backtest import sim close = data.get("price:收盤價") position = (close == close.rolling(250).max()) report = sim(position, resample="M", name="創年新高策略",upload=False) report.display()

00632R    0.125
00664R    0.125
00674R    0.125
00676R    0.125
00682U    0.125
00683L    0.125
00686R    0.125
1218      0.125
Name: 2022-07-31 00:00:00, dtype: float64

Timestamp('2022-07-31 00:00:00')

月營收公告截止日換股 & 本益成長比策略¶

position的dataframe用營收的index，不要用收盤價的index。 resample設D，就會是每月營收截止日換股（不一定是10日，遇假日可能往後），resample設D會依照position逐筆判斷進場訊號。

PEG策略教學

note:

position加上reindex月營收的日期索引，後會讓position日期變成每月營收截止日。

resample不須更改，使用預設"D"，這裡的"D"想成是每列而不是每日，因為使用reindex讓每列間隔一個月，所以執行resample="D"實際上是隔一個月。

In [7]:

from finlab import data
from finlab.backtest import sim

pe = data.get('price_earning_ratio:本益比')
rev = data.get('monthly_revenue:當月營收')
rev_ma3 = rev.average(3)
rev_ma12 = rev.average(12)
營業利益成長率 = data.get('fundamental_features:營業利益成長率')
peg = (pe/營業利益成長率)
cond1 = rev_ma3/rev_ma12 > 1.1
cond2 = rev/rev.shift(1) > 0.9

cond_all = cond1 & cond2
result = peg*(cond_all)

# 月營收截止日換股
position = result[result>0].is_smallest(10).reindex(rev.index_str_to_date().index, method='ffill')

report = sim(position=position, resample='D', name="策略教學範例:peg_rev", stop_loss=0.1, upload=False)
report.display()

from finlab import data from finlab.backtest import sim pe = data.get('price_earning_ratio:本益比') rev = data.get('monthly_revenue:當月營收') rev_ma3 = rev.average(3) rev_ma12 = rev.average(12) 營業利益成長率 = data.get('fundamental_features:營業利益成長率') peg = (pe/營業利益成長率) cond1 = rev_ma3/rev_ma12 > 1.1 cond2 = rev/rev.shift(1) > 0.9 cond_all = cond1 & cond2 result = peg*(cond_all) # 月營收截止日換股 position = result[result>0].is_smallest(10).reindex(rev.index_str_to_date().index, method='ffill') report = sim(position=position, resample='D', name="策略教學範例:peg_rev", stop_loss=0.1, upload=False) report.display()

1457    0.1
1906    0.1
2537    0.1
2603    0.1
3221    0.1
3706    0.1
4971    0.1
5309    0.1
6640    0.1
9906    0.1
Name: 2022-06-10 00:00:00, dtype: float64

Timestamp('2022-06-10 00:00:00')

技術指標選股¶

高 RSI 技術指標策略

上述策略會選出很多檔標的，要是我們只想要選出 20 檔要怎麼寫呢？這個策略我們可以將 RSI 最大的 20 檔股票納入組合，並且持有一週，來試試看效果如何。我們可以用 data.indicator 來計算所有股票的 RSI，也可以用 rsi.is_largest 來計算此股票的 RSI 是否是最大的 20 檔。

In [9]:

from finlab import data
from finlab.backtest import sim

# 選出 RSI 最大的 20 檔股票
rsi = data.indicator('RSI', timeperiod=20)
position = rsi.is_largest(20)

# 回測，每週(W)調整一次
report = sim(position, resample='W', name="高RSI策略", upload=False)
report.display()

from finlab import data from finlab.backtest import sim # 選出 RSI 最大的 20 檔股票 rsi = data.indicator('RSI', timeperiod=20) position = rsi.is_largest(20) # 回測，每週(W)調整一次 report = sim(position, resample='W', name="高RSI策略", upload=False) report.display()

00632R    0.05
00643K    0.05
00664R    0.05
00674R    0.05
00676R    0.05
00686R    0.05
020004    0.05
020015    0.05
02001R    0.05
2364      0.05
2936      0.05
2939      0.05
3205      0.05
5102      0.05
5904      0.05
6655      0.05
6796      0.05
8234      0.05
8427      0.05
9188      0.05
Name: 2022-07-10 00:00:00, dtype: float64

Timestamp('2022-07-10 00:00:00')

小資族優等生策略¶

上過 HaHow 上 FinLab Python 小資族課程嗎？

也可以把課程的內容改寫過來喔～

In [6]:

from finlab import data
from finlab.backtest import sim

股本 = data.get('financial_statement:股本')
price = data.get('price:收盤價')
市值 = 股本 * price / 10 * 1000

df1 = data.get('financial_statement:投資活動之淨現金流入_流出')
df2 = data.get('financial_statement:營業活動之淨現金流入_流出')
自由現金流 = (df1 + df2).rolling(4).mean()

稅後淨利 = data.get('fundamental_features:經常稅後淨利')
權益總計 = data.get('financial_statement:股東權益總額')
股東權益報酬率 = 稅後淨利/ 權益總計

營業利益成長率 = data.get('fundamental_features:營業利益成長率')

當月營收 = data.get('monthly_revenue:當月營收')* 1000
當季營收 = 當月營收.rolling(4).sum()
市值營收比 = 市值 / 當季營收


condition1 = (市值 < 1e10)
condition2 = 自由現金流 > 0
condition3 = 股東權益報酬率 > 0
condition4 = 營業利益成長率 > 0
condition5 = 市值營收比 < 5

position = condition1 & condition2 & condition3 & condition4 & condition5
report = sim(position, resample='M', name="優等生策略", upload=False)
report.display()

from finlab import data from finlab.backtest import sim 股本 = data.get('financial_statement:股本') price = data.get('price:收盤價') 市值 = 股本 * price / 10 * 1000 df1 = data.get('financial_statement:投資活動之淨現金流入_流出') df2 = data.get('financial_statement:營業活動之淨現金流入_流出') 自由現金流 = (df1 + df2).rolling(4).mean() 稅後淨利 = data.get('fundamental_features:經常稅後淨利') 權益總計 = data.get('financial_statement:股東權益總額') 股東權益報酬率 = 稅後淨利/ 權益總計 營業利益成長率 = data.get('fundamental_features:營業利益成長率') 當月營收 = data.get('monthly_revenue:當月營收')* 1000 當季營收 = 當月營收.rolling(4).sum() 市值營收比 = 市值 / 當季營收 condition1 = (市值 < 1e10) condition2 = 自由現金流 > 0 condition3 = 股東權益報酬率 > 0 condition4 = 營業利益成長率 > 0 condition5 = 市值營收比 < 5 position = condition1 & condition2 & condition3 & condition4 & condition5 report = sim(position, resample='M', name="優等生策略", upload=False) report.display()

INFO:finlab.data:financial_statement:股本 -- Daily data usage: 647.7 / 5000 MB
INFO:finlab.data:financial_statement:投資活動之淨現金流入_流出 -- Daily data usage: 649.0 / 5000 MB
INFO:finlab.data:financial_statement:營業活動之淨現金流入_流出 -- Daily data usage: 650.4 / 5000 MB
INFO:finlab.data:fundamental_features:經常稅後淨利 -- Daily data usage: 651.7 / 5000 MB
INFO:finlab.data:financial_statement:股東權益總額 -- Daily data usage: 653.0 / 5000 MB

stock_id
1108    0.005814
1203    0.005814
1217    0.005814
1258    0.005814
1336    0.005814
          ...   
8916    0.005814
8923    0.005814
8928    0.005814
8941    0.005814
9927    0.005814
Name: 2022-07-31 00:00:00, Length: 172, dtype: float64

Timestamp('2022-07-31 00:00:00')

多空並行回測¶

若策略想同時做多和做空，可參考以下範例，範例是做多 2330 一半的部位，做空 1101 一半的部位。

In [8]:

from finlab import data
from finlab import backtest

close = data.get('price:收盤價')

position = close < 0
position['2330'] = 0.5
position['1101'] = -0.5

report = backtest.sim(position, upload=False)
report.display()

from finlab import data from finlab import backtest close = data.get('price:收盤價') position = close < 0 position['2330'] = 0.5 position['1101'] = -0.5 report = backtest.sim(position, upload=False) report.display()

1101   -0.5
2330    0.5
Name: 2022-07-06 00:00:00, dtype: float64

Timestamp('2022-07-06 00:00:00')

營建產業限定策略¶

運用營建股"合約負債"的特性進行限定產業選股。

詳細文章內容

In [11]:

from finlab import data
from finlab.backtest import sim
import pandas as pd

# 限定回測產業範圍
with data.universe(category='建材營造'):

    # 資料引入
    close = data.get('price:收盤價')
    contract_debt = data.get('financial_statement:合約負債_流動')
    contract_debt_gr = (contract_debt/(contract_debt.shift())-1)
    equity = data.get('financial_statement:股本')

    # 計算合約負債佔股本的比率，比率越高代表未來潛在貢獻每股營收越高
    ce_ratio = (contract_debt / equity)

    # 選股條件
    cond1 = ce_ratio > 0.5
    cond2 = (contract_debt_gr > 0.05) & (contract_debt_gr<0.5)
    cond_all = cond1 & cond2

    result = ce_ratio * (cond_all)
    position = result[result > 0].is_largest(5)
    
    # position_limit 限定單檔標的最多持有30%部位，避免標的數過少時產生重壓的非系統性風險
    report = sim(position,resample='M',position_limit=0.3,stop_loss=0.1,fee_ratio=1.425/1000*0.3,name="營建股合約負債策略",upload=False)

report.display()

from finlab import data from finlab.backtest import sim import pandas as pd # 限定回測產業範圍 with data.universe(category='建材營造'): # 資料引入 close = data.get('price:收盤價') contract_debt = data.get('financial_statement:合約負債_流動') contract_debt_gr = (contract_debt/(contract_debt.shift())-1) equity = data.get('financial_statement:股本') # 計算合約負債佔股本的比率，比率越高代表未來潛在貢獻每股營收越高 ce_ratio = (contract_debt / equity) # 選股條件 cond1 = ce_ratio > 0.5 cond2 = (contract_debt_gr > 0.05) & (contract_debt_gr<0.5) cond_all = cond1 & cond2 result = ce_ratio * (cond_all) position = result[result > 0].is_largest(5) # position_limit 限定單檔標的最多持有30%部位，避免標的數過少時產生重壓的非系統性風險 report = sim(position,resample='M',position_limit=0.3,stop_loss=0.1,fee_ratio=1.425/1000*0.3,name="營建股合約負債策略",upload=False) report.display()

stock_id
2540    0.2
2548    0.2
3056    0.2
5213    0.2
6186    0.2
Name: 2022-05-31 00:00:00, dtype: float64

Timestamp('2022-05-31 00:00:00')

乖離率和財報濾網¶

In [13]:

from finlab import data
from finlab.backtest import sim

# 下載 ROE 跟收盤價
roe = data.get("fundamental_features:ROE稅後")
close = data.get("price:收盤價")
position = ((close / close.shift(60)).is_largest(30) & (roe > 0))
# 回測，每月(M)調整一次
report = sim(position, resample="M", upload=False)
report.display()

from finlab import data from finlab.backtest import sim # 下載 ROE 跟收盤價 roe = data.get("fundamental_features:ROE稅後") close = data.get("price:收盤價") position = ((close / close.shift(60)).is_largest(30) & (roe > 0)) # 回測，每月(M)調整一次 report = sim(position, resample="M", upload=False) report.display()

INFO:finlab.data:fundamental_features:ROE稅後 -- Daily data usage: 655.7 / 5000 MB

1319    0.055556
1522    0.055556
1524    0.055556
2305    0.055556
2364    0.055556
2482    0.055556
2736    0.055556
2929    0.055556
2939    0.055556
4543    0.055556
5258    0.055556
6113    0.055556
6199    0.055556
6472    0.055556
6655    0.055556
6715    0.055556
8097    0.055556
9802    0.055556
Name: 2022-07-31 00:00:00, dtype: float64

Timestamp('2022-07-31 00:00:00')

加入大盤指標濾網¶

大盤指標轉空時，出清持股

In [51]:

from finlab import data
from finlab.backtest import sim
import numpy as np


# 市場多空排列家數
def ls_order_position(short=5,mid=10,long=30):
    close = data.get("price:收盤價")
    short_ma = close.average(short)
    mid_ma = close.average(mid)
    long_ma = close.average(long)
    long_order = (short_ma>=mid_ma) & (mid_ma>=long_ma)
    long_order = long_order.sum(1)
    short_order = (short_ma<mid_ma) & (mid_ma<long_ma)
    short_order = short_order.sum(1)
    entry = long_order > short_order
    cond = ~close.isna()
    position = cond & entry
    return position


close = data.get("price:收盤價")
sma20 = close.average(20)
rev = data.get('monthly_revenue:當月營收')

cond1 = (close > close.shift(20)) & (close > close.shift(60))
cond2 = (rev.average(3) > rev.average(12)).sustain(3)

buy = cond1 & cond2
sell = sma20 < sma20.shift()

position = pd.DataFrame(np.nan, index=buy.index, columns=buy.columns)
position[buy] = 1
position[sell] = 0
position = position.ffill().fillna(0).astype(bool)
position = position & ls_order_position()
report = sim(position, resample="D", name="ls_order_position",upload=False)
report.display()

from finlab import data from finlab.backtest import sim import numpy as np # 市場多空排列家數 def ls_order_position(short=5,mid=10,long=30): close = data.get("price:收盤價") short_ma = close.average(short) mid_ma = close.average(mid) long_ma = close.average(long) long_order = (short_ma>=mid_ma) & (mid_ma>=long_ma) long_order = long_order.sum(1) short_order = (short_ma short_order cond = ~close.isna() position = cond & entry return position close = data.get("price:收盤價") sma20 = close.average(20) rev = data.get('monthly_revenue:當月營收') cond1 = (close > close.shift(20)) & (close > close.shift(60)) cond2 = (rev.average(3) > rev.average(12)).sustain(3) buy = cond1 & cond2 sell = sma20 < sma20.shift() position = pd.DataFrame(np.nan, index=buy.index, columns=buy.columns) position[buy] = 1 position[sell] = 0 position = position.ffill().fillna(0).astype(bool) position = position & ls_order_position() report = sim(position, resample="D", name="ls_order_position",upload=False) report.display()

Series([], Name: 2022-07-06 00:00:00, dtype: float64)

Timestamp('2022-07-06 00:00:00')

營收動能 & 過濾全額交割股¶

買進訊號隔日若為全額交割股，則不交易。 全額交割股清單每日0815更新。

In [16]:

from finlab import data
from finlab.backtest import sim
import pandas as pd

rev = data.get("monthly_revenue:當月營收")
rev_rf = data.get("monthly_revenue:去年同月增減(%)")
vol = data.get("price:成交股數")/1000

# 過濾全額交割股
full_cash_delivery_stock_filter = data.get("etl:full_cash_delivery_stock_filter")

rev_recent_3 = rev.rolling(3).sum()
vol_avg = vol.average(10)
cond1 = (rev_recent_3/rev_recent_3.rolling(24, min_periods=12).max())==1
cond2 = vol_avg > 300
cond_all = cond1 & cond2 & full_cash_delivery_stock_filter 
result = rev_rf*(cond_all)
position = result[result>0].is_largest(10).reindex(rev.index_str_to_date().index, method="ffill")
report = sim(position=position,stop_loss=0.3,position_limit=0.1, upload=False)
report.display()

from finlab import data from finlab.backtest import sim import pandas as pd rev = data.get("monthly_revenue:當月營收") rev_rf = data.get("monthly_revenue:去年同月增減(%)") vol = data.get("price:成交股數")/1000 # 過濾全額交割股 full_cash_delivery_stock_filter = data.get("etl:full_cash_delivery_stock_filter") rev_recent_3 = rev.rolling(3).sum() vol_avg = vol.average(10) cond1 = (rev_recent_3/rev_recent_3.rolling(24, min_periods=12).max())==1 cond2 = vol_avg > 300 cond_all = cond1 & cond2 & full_cash_delivery_stock_filter result = rev_rf*(cond_all) position = result[result>0].is_largest(10).reindex(rev.index_str_to_date().index, method="ffill") report = sim(position=position,stop_loss=0.3,position_limit=0.1, upload=False) report.display()

INFO:finlab.data:etl:full_cash_delivery_stock_filter -- Daily data usage: 659.8 / 5000 MB

1906    0.1
2530    0.1
2607    0.1
4946    0.1
6111    0.1
6241    0.1
6259    0.1
6446    0.1
6715    0.1
8097    0.1
Name: 2022-06-10 00:00:00, dtype: float64

Timestamp('2022-06-10 00:00:00')

現金流量策略 & 過濾 KY股¶

在台第一上市的海外公司經常爆出財務問題，對此有疑慮者，可在財報策略排除KY股。

基本面指標使用財報公告截止日。

In [53]:

from finlab import data
from finlab.backtest import sim

營業現金流 = data.get("financial_statement:營業活動之淨現金流入_流出").deadline()
投資現金流 = data.get("financial_statement:投資活動之淨現金流入_流出").deadline()
融資現金流 = data.get("financial_statement:籌資活動之淨現金流入_流出").deadline()
roe = data.get("fundamental_features:ROE稅後").deadline()
close = data.get("price:收盤價")

position = (營業現金流 > 0) & (投資現金流 > 0) & (融資現金流 < 0) & (roe > 5) & (close > 10)

# 過濾 KY股
sc = data.get("security_categories")
position_col = position.columns
ky_filter = position_col[~position_col.isin(list(sc[sc["name"].str.contains("KY")]["stock_id"]))]
position = position[ky_filter]

report = sim(position,resample="M",name="現金流正數", upload=False)
report.display()

from finlab import data from finlab.backtest import sim 營業現金流 = data.get("financial_statement:營業活動之淨現金流入_流出").deadline() 投資現金流 = data.get("financial_statement:投資活動之淨現金流入_流出").deadline() 融資現金流 = data.get("financial_statement:籌資活動之淨現金流入_流出").deadline() roe = data.get("fundamental_features:ROE稅後").deadline() close = data.get("price:收盤價") position = (營業現金流 > 0) & (投資現金流 > 0) & (融資現金流 < 0) & (roe > 5) & (close > 10) # 過濾 KY股 sc = data.get("security_categories") position_col = position.columns ky_filter = position_col[~position_col.isin(list(sc[sc["name"].str.contains("KY")]["stock_id"]))] position = position[ky_filter] report = sim(position,resample="M",name="現金流正數", upload=False) report.display()

stock_id
1539    0.043478
2303    0.043478
2504    0.043478
2636    0.043478
3093    0.043478
3131    0.043478
3169    0.043478
3557    0.043478
4506    0.043478
4945    0.043478
5534    0.043478
6195    0.043478
6278    0.043478
6281    0.043478
6294    0.043478
6506    0.043478
6542    0.043478
6683    0.043478
6706    0.043478
6747    0.043478
6761    0.043478
8084    0.043478
8421    0.043478
Name: 2022-07-31 00:00:00, dtype: float64

Timestamp('2022-07-31 00:00:00')