siat 3.10.25__py3-none-any.whl → 3.10.125__py3-none-any.whl

siat/fama_french.py

@@ -32,14 +32,211 @@ if czxt in ['linux']: #website Jupyter
 # 解决保存图像时'-'显示为方块的问题
 plt.rcParams['axes.unicode_minus'] = False 
 #==============================================================================
+def convert2freq(df,new_freq='daily'):
+    """
+    功能：将月度或年度频度的数据均匀化为日度数据
+    """
+    dfcols=list(df)
+    
+    import pandas as pd
+    df['datetmp']=df.index
+    df['Date']=df['datetmp'].apply(lambda x: x.to_timestamp())
+    df.set_index('Date',inplace=True)
+    #df['Date']=df['date']
+    if not isinstance(df.index, pd.DatetimeIndex):
+        df.index = pd.to_datetime(df.index)
+    
+    if new_freq == 'daily':
+        df = df.asfreq('D')
+    elif new_freq == 'monthly':
+        df = df.asfreq('M')
+    else:
+        df = df.asfreq('Y')
+        
+    for c in dfcols:
+        df[c] = df[c].ffill()
+        df[c] = df[c].bfill()
+        
+    df.drop(columns=['datetmp'], inplace=True)
+    
+    #df=df[['Date']+dfcols]
+    
+    return df
+#==============================================================================
+#==============================================================================        
 if __name__=='__main__':
-    start='2016-1-1'
-    end='2020-12-31'
+    start='auto'; end='today'
+    
+    scope='US'
+    factor='FF3'
+    freq='yearly'
+    
+    show_ff_factors(ticker='US',indicator='FF3',freq='yearly')
+
+def show_ff_factors(market='US',indicator='FF3', \
+                    start='auto',end='today',freq='yearly'):
+    """
+    功能：套壳函数get_ff3_factors/get_ffc4_factors/get_ff5_factors
+    
+    【支持的因子种类(factor)】
+    FF3: FF3各个因子
+    FFC4: FFC4各个因子
+    FF5: FF5各个因子
+
+    【支持的国家/地区(scope)】
+    US: 美国
+    North_America：北美(美国+加拿大)
+    Global: 全球
+    Global_ex_US: 全球(除美国外)
+    Asia_Pacific_ex_Japan: 亚太(除日本外)
+    China: 中国
+    Japan: 日本
+    Europe: 欧洲
+
+    【支持的取样频率(freq)】
+    yearly: 年
+    monthly: 月
+    daily: 日
+    """
+    ticker=market
+    
+    indicator1=indicator.lower()
+    factorlist=['ff3','ffc4','ff5']
+    if not (indicator1 in factorlist):
+        print(f"  #Error(show_ff_factors): unsupported factor {indicator}")
+        print(f"  Supported factors for FF models: {factorlist}")
+        return
+    
+    import datetime; todaydt = datetime.date.today(); todaystr=str(todaydt)
+    if end == 'today': end=todaystr
+    
+    if start == 'auto':
+        if freq == 'yearly':
+            start=date_adjust(end,adjust=-365*5)
+        elif freq == 'monthly':
+            start=date_adjust(end,adjust=-31*8)
+        else: #freq=='daily'
+            start=date_adjust(end,adjust=-60)
+            
+    start,end=start_end_preprocess(start,end)
+    
+    freq=freq.lower()
+    if freq == 'daily':
+        print("Looking for daily factors, it may take time ...")
+    
+    if indicator1=='ff3':
+        df=get_ff3_factors(start=start,end=end,scope=ticker,freq=freq)
+    elif indicator1=='ffc4':
+        df=get_ffc4_factors(start=start,end=end,scope=ticker,freq=freq)
+    elif indicator1=='ff5':
+        df=get_ff5_factors(start=start,end=end,scope=ticker,freq=freq)
+    else:
+        pass
+    
+    #处理无数据情形
+    if df is None:
+        print(f"  #Warning(show_ff_factors): got none data from {start} to {end}")
+        print(f"  Suggestion: may specify an earlier start date and try again")
+        return
+    if len(df) == 0:
+        print(f"  #Warning(show_ff_factors): got zero data from {start} to {end}")
+        print(f"  Suggestion: may specify an earlier start date and try again")
+        return    
+    
+    #处理RF小数位数
+    df['RF']=df['RF'].apply(lambda x: str(round(x,4)))
+    
+    #titletxt="Asset Pricing Factors: "+indicator.upper()+'Model, '+freq+' @ '+ticker
+    titletxt=indicator.upper()+'模型定价因子：'+freq.title()+' @'+ticker
+    
+    if freq == 'yearly':
+        ft0="说明：RF为年化无风险利率%(美短债收益率)"
+    elif freq == 'monthly':
+        ft0="说明：RF为月度无风险利率%(美短债收益率)"
+    else:
+        ft0="说明：RF为日频无风险利率%(基于短期美债收益率)"
+        
+    #footnote="Data source: Dartmouth College, "+todaystr
+    footnote=ft0+'\n'+"数据来源：Dartmouth College，"+todaystr
+    
+    df_display_CSS(df,titletxt=titletxt,footnote=footnote,facecolor='papayawhip', \
+                   decimals=2, \
+                   first_col_align='center',second_col_align='center', \
+                   last_col_align='center',other_col_align='center')    
+
+    return
+
+#==============================================================================        
+if __name__=='__main__':
+    start='2020-1-1'; end='2025-4-30'
+    
     scope='Europe'
     factor='Mom'
     freq='yearly'
+        
+    scope='US'
+    factor='FF3'
+    freq='yearly'
+    freq='daily'
+    
+    ff3d=get_ff_factors(start,end,scope,factor,freq='daily')
+    ff3mth=get_ff_factors(start,end,scope,factor,freq='monthly')
+    ff3y=get_ff_factors(start,end,scope,factor,freq='yearly')
+
+def get_ff_factors(start,end,scope='US',factor='FF3',freq='monthly',retry=10):
+    """
+    功能：套壳函数get_ff_factors0，应对freq='daily'经常失败的变通方法
+    """
+    import pandas as pd
+    startpd=pd.to_datetime(start)
+    endpd=pd.to_datetime(end)
+    
+    fff=None
+    for i in range(retry):
+        if not (fff is None): break
+        fff=get_ff_factors0(start=start,end=end,scope=scope,factor=factor,freq=freq)
+    
+    #多次尝试后仍未成功做变通处理：针对日度数据，使用月/年数据均匀化为日度数据
+    if fff is None:
+        if freq == 'daily':
+            start1=date_adjust(start,adjust=-31)
+            end1=date_adjust(end,adjust=31)
+            fff=get_ff_factors0(start=start1,end=end1,scope=scope,factor=factor,freq='monthly')
+            if not (fff is None):
+                ff_factors0=convert2freq(fff,new_freq='daily')
+                ff_factors=ff_factors0[(ff_factors0.index >=startpd) & (ff_factors0.index <=endpd)]
+            else:
+                ff_factors=fff
+
+        elif freq == 'monthly':
+            start1=date_adjust(start,adjust=-365)
+            end1=date_adjust(end,adjust=365)
+            fff=get_ff_factors0(start=start1,end=end1,scope=scope,factor=factor,freq='yearly') 
+            if not (fff is None):
+                ff_factors=convert2freq(fff,new_freq='monthly')
+            else:
+                ff_factors=fff
+
+        else:
+            start1=date_adjust(start,adjust=-365)
+            end1=date_adjust(end,adjust=365)
+            fff=get_ff_factors0(start=start1,end=end1,scope=scope,factor=factor,freq='yearly')
+            ff_factors=fff
+    else:
+        ff_factors=fff            
+
+    #加入日期
+    cols=list(ff_factors)
+    ff_factors['Date']=ff_factors.index
+    if freq == 'daily':
+        ff_factors['Date']=ff_factors['Date'].apply(lambda x: x.strftime("%Y-%m-%d"))
+    
+    ff_factors=ff_factors[['Date']+cols]
+    
+    return ff_factors
 
-def get_ff_factors(start,end,scope='US',factor='FF3',freq='yearly'):
+
+def get_ff_factors0(start,end,scope='US',factor='FF3',freq='yearly'):
    """
    【支持的因子种类(factor)】
    FF3: FF3各个因子
@@ -167,7 +364,7 @@ def get_ff_factors(start,end,scope='US',factor='FF3',freq='yearly'):
    try:
         ds = web.DataReader(symbol, source, start, end)
    except:
-        print("  #Error(get_ff_factors): inaccessible to ff data source (time out)")        
+        print(f"  #Warning(get_ff_factors): time out to get {freq} data, working around ...")        
         return None
     
    #提取希望的资产定价因子
@@ -731,7 +928,8 @@ def get_ffc4_factors(start,end,scope='US',freq='yearly'):
         Mom['date']=pd.to_datetime(Mom['date'].astype('str'))
         Mom.set_index('date',drop=True, inplace=True)
     
-    ffc4_factors=pd.merge(ff3,Mom,how='left',left_index=True,right_index=True)
+    #ffc4_factors=pd.merge(ff3,Mom,how='left',left_index=True,right_index=True)
+    ffc4_factors=pd.merge(ff3,Mom,how='inner',on='Date')
     #del ffc4_factors['date']
     
     return ffc4_factors
@@ -887,7 +1085,14 @@ if __name__=='__main__':
 
 
 #==============================================================================
-
+if __name__=='__main__':
+    ticker='BIDU'
+    start='2025-1-1'
+    end='2025-5-30'
+    scope='US'
+    graph=True
+    
+    
 def reg_ff3_betas(ticker,start,end,scope='US',graph=True):
     """
     功能：测试一只股票对于FF3各个因子的系数大小、符号方向和显著性
@@ -915,11 +1120,15 @@ def reg_ff3_betas(ticker,start,end,scope='US',graph=True):
     ret=ret.rename(columns={'Close':'dRet'})
 
     #抓取FF3因子
+    print("  Looking for factors for FF3 model ...")
     freq='daily'
     ff3=get_ff3_factors(start,end,scope,freq)
     if ff3 is None:
         print("  #Warning(reg_ff3_betas): factors unavailable for",scope,freq,'from',start,'to',end)
         return None
+    if len(ff3) == 0:
+        print("  #Warning(reg_ff3_betas): zero factors retrieved for",scope,freq,'from',start,'to',end)
+        return None
     
     #改造索引类型为时间戳索引（DatetimeIndex），便于与ret合成
     ff3['Date']=ff3.index.strftime("%m/%d/%Y")
@@ -944,7 +1153,7 @@ def reg_ff3_betas(ticker,start,end,scope='US',graph=True):
     #输出结果并绘制横向柱状图    
     if graph == True:
         gparms=parms.iloc[[1,2,3]]
-        print("\n",parms)
+        #print("\n",parms)
         title=ticker_name(ticker)+": FF3模型的贝塔系数"
         plt.title(title,fontsize=12,fontweight='bold')
         plt.ylabel("贝塔系数",fontsize=12,fontweight='bold')
@@ -1005,6 +1214,13 @@ if __name__=='__main__':
 
     
 #==============================================================================
+if __name__=='__main__':
+    ticker='MSFT'
+    start='2022-5-1'; end='2025-4-30'
+    scope='US'; graph=True
+    
+    
+    
 def reg_ffc4_betas(ticker,start,end,scope='US',graph=True):
     """
     功能：测试一只股票对于FFC4各个因子的系数大小、符号方向和显著性
@@ -1033,16 +1249,20 @@ def reg_ffc4_betas(ticker,start,end,scope='US',graph=True):
     ret=ret.rename(columns={'Close':'dRet'})
 
     #抓取每日FFC4因子
+    print("  Looking for factors for FFC4 model ...")
     freq='daily'
     ffc4=get_ffc4_factors(start,end,scope,freq)
     if ffc4 is None:
         print("  #Warning(reg_ffc4_betas): factors not available for",scope,freq,'from',start,'to',end)
         return None    
+    if len(ffc4) == 0:
+        print("  #Warning(reg_ffc4_betas): zero factors retrieved for",scope,freq,'from',start,'to',end)
+        return None    
     
     #改造索引类型为时间戳索引（DatetimeIndex），便于与ret合成
-    ffc4['Date']=ffc4.index.strftime("%m/%d/%Y")
+    #ffc4['Date']=ffc4.index.strftime("%m/%d/%Y")
     ffc4['Date']=pd.to_datetime(ffc4['Date'])
-    ffc4.set_index('Date',inplace=True)    
+    ffc4.set_index('Date',drop=True,inplace=True)    
 
     #合成股票日收益率+每日FF3因子
     sample=pd.merge(ret,ffc4,how='inner',left_index=True,right_index=True)
@@ -1051,7 +1271,10 @@ def reg_ffc4_betas(ticker,start,end,scope='US',graph=True):
     #回归FF3模型
     import statsmodels.api as sm
     y=sample['dRet']
-    X=sample[['Mkt-RF','SMB','HML','MOM']]
+    try:
+        X=sample[['Mkt-RF','SMB','HML','MOM']]
+    except:
+        X=sample[['Mkt-RF','SMB','HML','Mom']]
     X1=sm.add_constant(X)
     results=sm.OLS(y,X1).fit() 
     
@@ -1122,6 +1345,11 @@ if __name__=='__main__':
     test_ffc4_betas() 
     
 #==============================================================================
+if __name__=='__main__':
+    ticker='MSFT'
+    start='2024-5-1'; end='2025-4-30'
+    scope='US'; graph=True
+
 def reg_ff5_betas(ticker,start,end,scope='US',graph=True):
     """
     功能：测试一只股票对于FF5各个因子的系数大小、符号方向和显著性
@@ -1149,12 +1377,16 @@ def reg_ff5_betas(ticker,start,end,scope='US',graph=True):
     #命名日收益率字段为dRet
     ret=ret.rename(columns={'Close':'dRet'})
 
-    #抓取每日FF3因子
+    #抓取每日FF5因子
+    print("  Looking for factors for FF5 model ...")
     freq='daily'
     ff5=get_ff5_factors(start,end,scope,freq)
     if ff5 is None:
         print("  #Warning(reg_ff5_betas): factors not available for",scope,freq,'from',start,'to',end)
         return None       
+    if len(ff5) is None:
+        print("  #Warning(reg_ff5_betas): zero factors retrieved for",scope,freq,'from',start,'to',end)
+        return None       
     
     #改造索引类型为时间戳索引（DatetimeIndex），便于与ret合成
     ff5['Date']=ff5.index.strftime("%m/%d/%Y")
@@ -1238,4 +1470,52 @@ def test_ff5_betas():
 if __name__=='__main__':
     test_ff5_betas() 
     
-#==============================================================================    
+#==============================================================================
+
+
+
+def reg_ff_betas(ticker,start,end='today',indicator='FF3',market='US'):
+    """
+    功能：套壳函数reg_ff3_betas/reg_ffc4_betas/reg_ff5_betas
+    参数：
+    indicator：默认'FF3',还可为'FFC4'或'FF5'
+    scope：默认'US',还可为'China'或'Japan'或'Europe'等
+    """
+    scope=market
+    
+    start,end=start_end_preprocess(start,end)
+    
+    indicator1=indicator.lower()
+    factorlist=['ff3','ffc4','ff5']
+    if not (indicator1 in factorlist):
+        print(f"  #Error(show_ff_factors): unsupported factor {indicator}")
+        print(f"  Supported factors for FF models: {factorlist}")
+        return    
+        
+    if indicator1=='ff3':
+        df=reg_ff3_betas(ticker=ticker,start=start,end=end,scope=scope)
+    elif indicator1=='ffc4':
+        df=reg_ffc4_betas(ticker=ticker,start=start,end=end,scope=scope)
+    elif indicator1=='ff5':
+        df=reg_ff5_betas(ticker=ticker,start=start,end=end,scope=scope)
+    else:
+        pass   
+    
+    #打印回归结果
+    cols=list(df)
+    df['const']=df.index
+    df=df[['const']+cols]
+    
+    titletxt=ticker_name(ticker)+"："+indicator.upper()+'模型回归结果'  
+    import datetime; todaydt = datetime.date.today(); todaystr=str(todaydt)
+    footnote="数据来源：Dartmouth College/sina/stooq/Yahoo等，"+todaystr    
+    
+    df.rename(columns={'const':'Constant','coef':'Coefficient', \
+                       'sig':'Significance'},inplace=True)
+    
+    df_display_CSS(df,titletxt=titletxt,footnote=footnote,facecolor='papayawhip', \
+                   decimals=4, \
+                   first_col_align='center',second_col_align='center', \
+                   last_col_align='center',other_col_align='center')       
+        
+    return df

siat/financial_statements.py

@@ -52,14 +52,8 @@ def get_balance_sheet(symbol):
     try:
         stmta=stock.balance_sheet()  # Defaults to Annual
     except:
-        print("  Searching for annual report failed, recovering......")
-        sleep_random(max_sleep=60)
-        
-        try:
-            stmta=stock.balance_sheet()
-        except:
-            print("  #Warning(get_balance_sheet): no annual info available for",symbol)
-            return None
+        print("  #Warning(get_balance_sheet): no annual info available for",symbol)
+        return None
         
     if isinstance(stmta,dict):
             print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
@@ -76,14 +70,8 @@ def get_balance_sheet(symbol):
     try:
         stmtq=stock.balance_sheet(frequency="q")
     except:
-        print("  Searching for quarterly report failed, recovering......")
-        sleep_random(max_sleep=60)
-        
-        try:
-            stmtq=stock.balance_sheet(frequency="q")
-        except:
-            print("  #Warning(get_balance_sheet): no quarterly info available for",symbol)
-            return None
+        print("  #Warning(get_balance_sheet): no quarterly info available for",symbol)
+        return None
         
     if isinstance(stmtq,dict):
             print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
@@ -196,17 +184,12 @@ def get_income_statements(symbol):
     try:
         stmta=stock.income_statement()  # Defaults to Annual
     except:
-        print("  Searching for annual report failed, recovering......")
-        sleep_random(max_sleep=60)
-        try:
-            stmta=stock.income_statement()
-        except:
-            print("  #Warning(get_income_statements): no annual info available for",symbol)
-            return None
+        print("  #Warning(get_income_statements): no annual info available for",symbol)
+        return None
 
     if isinstance(stmta,dict):
-            print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
-            return None
+        print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
+        return None
         
     """
     #判断是否抓取到了数据
@@ -219,17 +202,12 @@ def get_income_statements(symbol):
     try:
         stmtq=stock.income_statement(frequency="q")
     except:
-        print("  Searching for quarterly report failed, recovering......")
-        sleep_random(max_sleep=60)
-        try:
-            stmtq=stock.income_statement(frequency="q")
-        except:
-            print("  #Warning(get_income_statements): no quarterly info available for",symbol)
-            return None
+        print("  #Warning(get_income_statements): no quarterly info available for",symbol)
+        return None
         
     if isinstance(stmtq,dict):
-            print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
-            return None
+        print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
+        return None
         
     """
     if (len(stmta)==0) and (len(stmtq)==0):
@@ -311,17 +289,12 @@ def get_cashflow_statements(symbol):
     try:
         stmta=stock.cash_flow()  # Defaults to Annual
     except:
-        print("  Searching for annual report failed, recovering......")
-        sleep_random(max_sleep=60)
-        try:
-            stmta=stock.cash_flow()
-        except:
-            print("  #Warning(get_cashflow_statements): no annual info available for",symbol)
-            return None
+        print("  #Warning(get_cashflow_statements): no annual info available for",symbol)
+        return None
 
     if isinstance(stmta,dict):
-            print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
-            return None    
+        print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
+        return None    
         
     """
     #判断是否抓取到了数据
@@ -334,17 +307,12 @@ def get_cashflow_statements(symbol):
     try:
         stmtq=stock.cash_flow(frequency="q")
     except:
-        print("  Searching for quarterly report failed, recovering......")
-        sleep_random(max_sleep=60)
-        try:
-            stmtq=stock.cash_flow(frequency="q")
-        except:
-            print("  #Warning(get_cashflow_statements): no quarterly info available for",symbol)
-            return None
+        print("  #Warning(get_cashflow_statements): no quarterly info available for",symbol)
+        return None
         
     if isinstance(stmtq,dict):
-            print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
-            return None
+        print("  #Warning(get_balance_sheet): Yahoo Finance currently unaccessable for",symbol)
+        return None
         
     """
     if (len(stmta)==0) and (len(stmtq)==0):
@@ -408,17 +376,8 @@ def get_financial_statements(ticker):
     try:
         fbs = get_balance_sheet(ticker)
     except:
-        print("  Retrieving fin info failed, trying to recover...")
-        sleep_random(max_sleep=60)
-        try:
-            fbs = get_balance_sheet(ticker)
-        except:
-            sleep_random(max_sleep=60)
-            try:
-                fbs = get_balance_sheet(ticker)
-            except:
-                print("  #Error(get_financial_statements): balance sheet not available for",ticker)
-                return None
+        print("  #Error(get_financial_statements): balance sheet not available for",ticker)
+        return None
     if fbs is None:
         print("  #Error(get_financial_statements): financial statements not available for",ticker)
         return None
@@ -427,17 +386,8 @@ def get_financial_statements(ticker):
     try:
         fis = get_income_statements(ticker)
     except:
-        print("      Failed, recovering...")
-        sleep_random(max_sleep=60)
-        try:
-            fis = get_income_statements(ticker)
-        except:
-            sleep_random(max_sleep=60)
-            try:
-                fis = get_income_statements(ticker)
-            except:
-                print("  #Error(get_financial_statements): income info not available for",ticker)
-                return None
+        print("  #Error(get_financial_statements): income info not available for",ticker)
+        return None
     if fis is None:
         print("  #Error(get_financial_statements): financial statements not available for",ticker)
         return None
@@ -446,17 +396,8 @@ def get_financial_statements(ticker):
     try:
         fcf = get_cashflow_statements(ticker)
     except:
-        print("      Failed, recovering...")
-        sleep_random(max_sleep=60)
-        try:
-            fcf = get_cashflow_statements(ticker)
-        except:
-            sleep_random(max_sleep=60)
-            try:
-                fcf = get_cashflow_statements(ticker)
-            except:
-                print("  #Error(get_financial_statements): cash flow info not available for",ticker)
-                return None
+        print("  #Error(get_financial_statements): cash flow info not available for",ticker)
+        return None
     if fcf is None:
         print("  #Error(get_financial_statements): financial statements not available for",ticker)
         return None