[Python知识库]Python获取全部基金前十大持仓股并进行选股分析

选股逻辑

股市有风险，投资需谨慎。

投资方向：跟随基金机构进行选股，简单来说，就是优先筛选那些基金公司重仓持有的股票。

目标设定

获取全部基金前十大持仓股的持股数、市值等信息，分析出排名比较靠前的基金重仓股。

爬取全部基金代码信息

注：使用python内置sqlite数据库

建表存储基金代码信息

-- 全部标的信息
create table if not exists targets_all
(
    uid      string,
    name     string,
    id       string,
    category string,
    tag      string,
    type     string,
    version  datetime not null default (datetime('now', 'localtime'))
);

参考：https://juejin.cn/post/6993525305713541151

根据基金代码获取相应持仓信息

建表存储持仓信息

-- 基金前十大持仓股
create table if not exists top10_stocks_in_fund
(
    uid string,
    fund_id string,
    rank_num integer,
    stock_id string,
    stock_name string,
    hold_rate float,
    hold_count float,
    hold_value float,
    update_time string
)

爬取并存入数据库

# coding:utf-8
import logging
from collections import defaultdict
from bs4 import BeautifulSoup as BS
import requests
import uuid

from common.constant import CONN_KEY
from util import db_util
from common import global_var as gv

logging.basicConfig(level=logging.INFO)


# 将请求到的数据进行解析
def _parse_content(content):
    result = content.replace("var apidata={", "")
    result = result[:-2]

    content = result.split("\",arryear:")[0].replace("content:\"", "")
    content = BS(content).find_all("tr")

    result = defaultdict(list)

    for tr in content:
        date = tr.parent.parent.previous_sibling.previous_sibling.find("font").text
        td = tr.find_all("td")
        if len(td) == 9:
            result[date].append({
                "rank_num": td[0].text,
                "stock_id": td[1].text,
                "stock_name": td[2].text,
                "hold_rate": float(td[6].text.replace("%", ""))/100,
                "hold_count": td[7].text,
                "hold_value": eval(td[8].text.replace(",", ""))
            })
        elif len(td) == 7:
            result[date].append({
                "rank_num": td[0].text,
                "stock_id": td[1].text,
                "stock_name": td[2].text,
                "hold_rate": float(td[4].text.replace("%", ""))/100,
                "hold_count": td[5].text,
                "hold_value": eval(td[6].text.replace(",", ""))
            })

    return result


# 根据基金代码获取基金持仓
def get_top10_stocks_by_fund(fund_code):
    base_url = "https://fundf10.eastmoney.com/FundArchivesDatas.aspx?type=jjcc&code={fund_code}&topline=10"
    url = base_url.format(fund_code=fund_code)
    response = str(requests.get(url).content, "utf-8")
    return _parse_content(response)


# 判断当前基金的股票持仓是否存在
def _is_exist(conn, fund_id, stock_id, update_time):
    query_sql = f"select 1 from top10_stocks_in_fund " \
                f"where fund_id = '{fund_id}'" \
                f"and stock_id = '{stock_id}'" \
                f"and update_time = '{update_time}'"

    res = conn.query(query_sql)
    if res.fetchone():
        return True
    return False


# 更新基金前10大持仓股信息
def update_top10_stocks_by_fund():
    # 初始化数据库连接池
    db_util.init_conn_pool()
    conn = gv.get_value(CONN_KEY)
    query_sql = "select * from targets_all where type = '基金'"
    targets = conn.query(query_sql)

    # 记录更新成功和失败的条数
    success_count = 0
    failure_count = 0

    for (uid, fund_name, fund_id, category, tag, type, version) in targets:
        fund_id = str(fund_id).zfill(6)
        logging.info(f"正在请求{fund_name}-{fund_id}的持仓股信息...")
        stocks_in_fund = get_top10_stocks_by_fund(fund_id)

        # 获取持仓信息
        for update_time, stocks in stocks_in_fund.items():
            # 遍历各条持股信息
            for stock in stocks:
                # 若当前持股信息已存在，则不再插入
                if _is_exist(conn, fund_id, stock["stock_id"], update_time):
                    continue

                # 插入新的持股信息
                insert_sql = f"insert into top10_stocks_in_fund" \
                             f"(uid,fund_id,rank_num,stock_id,stock_name,hold_rate,hold_count,hold_value,update_time)" \
                             f"values('{uuid.uuid4()}','{fund_id}',{stock['rank_num']}," \
                             f"'{stock['stock_id']}','{stock['stock_name']}',{stock['hold_rate']}," \
                             f"{stock['hold_count']},{stock['hold_value']},'{update_time}')"

                if conn.operate(insert_sql):
                    logging.info("插入成功！")
                    success_count = success_count+1
                else:
                    logging.warning(f"插入失败：{insert_sql}")
                    failure_count = failure_count+1

    logging.info(f"全部基金持仓信息更新完毕，成功：{success_count}条，失败：{failure_count}条。")


if __name__ == "__main__":
    update_top10_stocks_by_fund()

分析第一/二季度基金排名靠前的重仓股

select stock_id "股票代码", stock_name "股票名称", cast(sum(hold_value) as double) "基金持有总市值"
from top10_stocks_in_fund
where update_time = '2021-06-30'
group by 1, 2
order by 3 desc

select stock_id "股票代码", stock_name "股票名称", cast(sum(hold_count) as double) "基金持有数(万股)"
from top10_stocks_in_fund
where update_time = '2021-06-30'
group by 1, 2
order by 3 desc

结论

金融市场偏好能源消费、银行券商等行业龙头股，深入研究持股数和市值变化，可以为加仓换股提供比较有意义的参考。（注：基金持仓信息大多是每季度末更新，距离越久，参考意义越不大）