日経225の全銘柄で、配当日の一ヶ月前に買って、直前に売ったら儲かるか、過去5年分で検証してみた。

import yfinance as yf
import pandas as pd
import sqlite3
import matplotlib.pyplot as plt
import time

# データベースファイルのパス
db_path = 'profit_loss.db'

# 日経225銘柄リスト
ticker_symbols = ['7272','2914','4452','4911','6326','3659','4704','5108','1605','2502','3405','3436','4631','8804','2503','4004','4324','4519','5101','5201','5301','6302','6361','6723','7751','2501','4578','5214','9147','1928','6506','8233','9602','3086','3382','8267','7741','8697','4689','6861','8795','8411','9064','4568','5831','7832','2768','3099','6301','6758','6902','7733','8035','9104','3401','4005','4523','4902','5406','5801','6501','6594','8031','8253','9101','9434','9613','9984','4061','4503','4507','5019','6103','6178','6473','6701','6762','7004','7186','7270','8316','9020','9433','4042','5401','5711','5713','6471','6645','6770','6857','6988','7013','7211','7267','7269','8001','8002','8015','8053','8058','8304','8630','8766','9001','9022','9107','9432','9532','9843','2432','4043','5333','6702','7735','7752','7974','8309','8591','8725','9021','9201','2802','2871','3402','3407','4183','4188','4506','5332','5411','5541','5631','6098','6305','6504','6752','6841','6971','7205','8252','8306','8308','8750','9735','1332','1801','2002','2282','3289','4021','4502','4543','5020','5232','5802','6113','6472','6724','6753','7012','7202','7261','8331','8604','9202','1721','1802','1803','1808','1812','1925','1963','2269','2413','2531','2801','3861','3863','4041','4208','4661','5706','5714','6273','6367','6479','6503','6674','6952','6981','7011','7731','7762','7911','7912','8354','8601','8801','8802','9005','9007','9008','9009','9301','9501','9502','9503','9531','9766','4063','4901','5233','6954','6976','7201','7203','7951','8830','5803','4385','6920','9983','4751','4151']

# SQLiteデータベースに接続
conn = sqlite3.connect(db_path)

# 各銘柄について損益を計算
for ticker_symbol in ticker_symbols:
    ticker = yf.Ticker(ticker_symbol + '.T')
    
    # 過去5年間の配当データと株価データを取得
    dividends = ticker.dividends.last('5YE')
    
    for ex_date in dividends.index:
        buy_date = ex_date - pd.Timedelta(days=30)
        sell_date = ex_date - pd.Timedelta(days=1)
        
        # 株価データを取得
        hist = ticker.history(start=buy_date, end=sell_date)
        time.sleep(1) 

        # 購入日と売却日の株価を取得
        if not hist.empty and len(hist) > 1:
            buy_price = hist.iloc[0]['Close']
            sell_price = hist.iloc[-1]['Close']
            profit = (sell_price - buy_price) / buy_price * 100  # 損益率(%)
            
            # 損益率と配当支払日をデータベースに保存
            pd.DataFrame({
                'Ticker': [ticker_symbol],
                'AvgProfit': [profit],
                'Date': [ex_date.strftime('%Y-%m-%d')]  # 日付を文字列形式で保存
            }).to_sql('profit_loss', conn, if_exists='append', index=False)

# データベース接続を閉じる
conn.close()

# データベースから損益データを読み込むための関数
def load_data(db_path):
    conn = sqlite3.connect(db_path)
    df = pd.read_sql('SELECT * FROM profit_loss', conn)
    conn.close()
    df['Year'] = pd.to_datetime(df['Date']).dt.year
    return df

# データベースから損益データを読み込む
profit_loss_df = load_data(db_path)

# 銘柄ごとにプロット
plt.figure(figsize=(12, 8))

for ticker in profit_loss_df['Ticker'].unique():
    data = profit_loss_df[profit_loss_df['Ticker'] == ticker]
    years = data['Year'].unique()
    
    profits_by_year = data.groupby('Year')['AvgProfit'].mean().reindex(years).fillna(0)
    
    # 年ごとの損益を折れ線グラフでプロット
    plt.plot(profits_by_year.index, profits_by_year.values, '-o', label=f'{ticker} Yearly')

plt.xlabel('Year / Ticker Symbol')
plt.ylabel('Profit (%)')
plt.title('Yearly Profit/Loss for Each Ticker Over 5 Years')
plt.xticks(rotation=45)  # X軸のラベルを45度回転
plt.legend()
plt.tight_layout()
plt.show()

import yfinance as yf

import pandas as pd

import sqlite3

import matplotlib.pyplot as plt

import time

# データベースファイルのパス

db_path = 'profit_loss.db'

# 日経225銘柄リスト

ticker_symbols = ['7272','2914','4452','4911','6326','3659','4704','5108','1605','2502','3405','3436','4631','8804','2503','4004','4324','4519','5101','5201','5301','6302','6361','6723','7751','2501','4578','5214','9147','1928','6506','8233','9602','3086','3382','8267','7741','8697','4689','6861','8795','8411','9064','4568','5831','7832','2768','3099','6301','6758','6902','7733','8035','9104','3401','4005','4523','4902','5406','5801','6501','6594','8031','8253','9101','9434','9613','9984','4061','4503','4507','5019','6103','6178','6473','6701','6762','7004','7186','7270','8316','9020','9433','4042','5401','5711','5713','6471','6645','6770','6857','6988','7013','7211','7267','7269','8001','8002','8015','8053','8058','8304','8630','8766','9001','9022','9107','9432','9532','9843','2432','4043','5333','6702','7735','7752','7974','8309','8591','8725','9021','9201','2802','2871','3402','3407','4183','4188','4506','5332','5411','5541','5631','6098','6305','6504','6752','6841','6971','7205','8252','8306','8308','8750','9735','1332','1801','2002','2282','3289','4021','4502','4543','5020','5232','5802','6113','6472','6724','6753','7012','7202','7261','8331','8604','9202','1721','1802','1803','1808','1812','1925','1963','2269','2413','2531','2801','3861','3863','4041','4208','4661','5706','5714','6273','6367','6479','6503','6674','6952','6981','7011','7731','7762','7911','7912','8354','8601','8801','8802','9005','9007','9008','9009','9301','9501','9502','9503','9531','9766','4063','4901','5233','6954','6976','7201','7203','7951','8830','5803','4385','6920','9983','4751','4151']

# SQLiteデータベースに接続

conn = sqlite3.connect(db_path)

# 各銘柄について損益を計算

for ticker_symbol in ticker_symbols:

ticker = yf.Ticker(ticker_symbol + '.T')

# 過去5年間の配当データと株価データを取得

dividends = ticker.dividends.last('5YE')

for ex_date in dividends.index:

buy_date = ex_date - pd.Timedelta(days=30)

sell_date = ex_date - pd.Timedelta(days=1)

# 株価データを取得

hist = ticker.history(start=buy_date, end=sell_date)

time.sleep(1)

# 購入日と売却日の株価を取得

if not hist.empty and len(hist) > 1:

buy_price = hist.iloc[0]['Close']

sell_price = hist.iloc[-1]['Close']

profit = (sell_price - buy_price) / buy_price * 100 # 損益率(%)

# 損益率と配当支払日をデータベースに保存

pd.DataFrame({

'Ticker': [ticker_symbol],

'AvgProfit': [profit],

'Date': [ex_date.strftime('%Y-%m-%d')] # 日付を文字列形式で保存

}).to_sql('profit_loss', conn, if_exists='append', index=False)

# データベース接続を閉じる

conn.close()

# データベースから損益データを読み込むための関数

def load_data(db_path):

conn = sqlite3.connect(db_path)

df = pd.read_sql('SELECT * FROM profit_loss', conn)

conn.close()

df['Year'] = pd.to_datetime(df['Date']).dt.year

return df

# データベースから損益データを読み込む

profit_loss_df = load_data(db_path)

# 銘柄ごとにプロット

plt.figure(figsize=(12, 8))

for ticker in profit_loss_df['Ticker'].unique():

data = profit_loss_df[profit_loss_df['Ticker'] == ticker]

years = data['Year'].unique()

profits_by_year = data.groupby('Year')['AvgProfit'].mean().reindex(years).fillna(0)

# 年ごとの損益を折れ線グラフでプロット

plt.plot(profits_by_year.index, profits_by_year.values, '-o', label=f'{ticker} Yearly')

plt.xlabel('Year / Ticker Symbol')

plt.ylabel('Profit (%)')

plt.title('Yearly Profit/Loss for Each Ticker Over 5 Years')

plt.xticks(rotation=45) # X軸のラベルを45度回転

plt.legend()

plt.tight_layout()

plt.show()

過去5年で、日経２２５の全銘柄を配当日の一ヶ月前に買って、直前で売ったら
1.42432634843033 = 42.43%増えるっぽい。

select avg(AvgProfit)
from profit_loss

1 2	select avg(AvgProfit) from profit_loss

ホンマかいな？
去年1年間に限定すると、1.06758290731527で、6.75%しか増えない…。

select avg(AvgProfit)
from profit_loss
where Date BETWEEN '2023-01-01' and '2023-12-31'

select avg(AvgProfit)

from profit_loss

where Date BETWEEN '2023-01-01' and '2023-12-31'

年によってバラバラだな…。

SELECT DATE_FORMAT(date,"%Y"), avg(AvgProfit) 
FROM `profit_loss` 
GROUP BY DATE_FORMAT(date,"%Y")

SELECT DATE_FORMAT(date,"%Y"), avg(AvgProfit)

FROM `profit_loss`

GROUP BY DATE_FORMAT(date,"%Y")

2006 0.42932260082157503
2007 -1.50511207941048
2008 -7.348637230573734
2009 -4.184183437433469
2010 -2.565141554643453
2017 9.49097995712965
2018 -0.28980099664524017
2019 2.2362537576919905
2020 -3.066188644443086
2021 5.620801464833454
2022 0.22597778630588108
2023 1.0675829073152694
2024 5.302092547982628

発想を変えて、儲かった銘柄コードを算出してみよう。

select Ticker, avg(AvgProfit),max(AvgProfit),min(AvgProfit), STDDEV(AvgProfit)
from profit_loss
group by `Ticker`
HAVING avg(AvgProfit) > 0 -- 損益がプラス
and (avg(AvgProfit) + min(AvgProfit)) > 1 -- 大幅に下がっていない銘柄
order by min(AvgProfit) DESC;

select Ticker, avg(AvgProfit),max(AvgProfit),min(AvgProfit), STDDEV(AvgProfit)

from profit_loss

group by `Ticker`

HAVING avg(AvgProfit) > 0 -- 損益がプラス

and (avg(AvgProfit) + min(AvgProfit)) > 1 -- 大幅に下がっていない銘柄

order by min(AvgProfit) DESC;

Ticker avg(AvgProfit) max(AvgProfit) min(AvgProfit) STDDEV(AvgProfit)
9007 8.334906878377723 17.5605253773565 1.1253958826826 5.467529813553868
9001 5.340144025773639 21.5384611810114 -0.755299923599934 6.109776193653475
9005 4.363215281218556 8.94256111476747 -1.97461395145588 3.378942894043412
5401 4.025981746611886 17.9037925960584 -2.1673903065123 6.911068836329029
7974 3.592718641567016 13.8745076898528 -2.3507327826203 4.992528819197802
5406 4.522780145854256 13.6862311086385 -2.49222084341221 5.82302628381912

過去5年間のデータ上では、この5銘柄が、配当権利確定日の一ヶ月前に買って、直前で売れば儲かりそう…。
なんだけど、実際のチャートを見ると全然そんな事ないな
9007　小田急電鉄
9001　東武鉄道
9005　東急
5401　日本製鉄
7974　任天堂
5406　神戸製鋼所

関連記事