1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
|
import os
from common import root
import csv
import datetime
from db import serenitasdb
import re, sys
from pandas.tseries.offsets import BDay
import pandas as pd
import numpy as np
import psycopg2
from sqlalchemy import create_engine
from collections import defaultdict
def convert(x):
try:
return float(x[:-1])
except ValueError:
return None
index_list = ['HY9', 'HY10'] + ['HY' + str(s) for s in range(15, 25)] + ['IG9'] + \
['IG' + str(s) for s in range(16, 25)] + ['XO22', 'XO23', 'EU9'] + \
['EU' +str(s) for s in range(19, 24)]
DOC_CLAUSE_MAPPING14 = {'Full Restructuring': 'MM14',
'No Restructuring': 'XR14',
'Modified Modified Restructurin': 'MM14'}
DOC_CLAUSE_MAPPING = {'Full Restructuring': 'MM',
'No Restructuring': 'XR',
'Modified Modified Restructurin': 'MM'}
def get_markit_bbg_mapping(database, basketid_list, workdate):
if workdate>=datetime.date(2014, 9, 19):
doc_clause_mapping = DOC_CLAUSE_MAPPING14
else:
doc_clause_mapping = DOC_CLAUSE_MAPPING
markit_bbg_mapping = defaultdict(set)
all_tickers = set([])
with database.cursor() as c:
c.execute("SELECT markit_ticker, markit_tier, spread, currency, cds_curve, " \
" doc_clause FROM historical_cds_issuers(%s) where index_list && %s",
(workdate, basketid_list))
for line in c:
all_tickers.add((line['markit_ticker'], line['markit_tier']))
key = (line['markit_ticker'], line['markit_tier'], line['currency'],
doc_clause_mapping[line['doc_clause']], float(line['spread'])/10000)
if key==('CESEOP', 'SNRFOR', 'USD', 'XR14', 0.05):
key=('CESEOP', 'SNRFOR', 'USD', 'XR', 0.05)
## each markit ticker can be mapped to multiple bbg tickers
## these bbg tickers can have different curves (ok)
## or same curves (not ok since date, curve_ticker needs to be unique)
## therefore we keep them in a set strucutre
markit_bbg_mapping[key].add(tuple(line['cds_curve']))
database.commit()
return (all_tickers, markit_bbg_mapping)
def get_basketids(database, index_list, workdate):
r = []
with database.cursor() as c:
for index in index_list:
c.execute("SELECT * FROM nameToBasketID(%s, %s)", (index, workdate))
r.append(c.fetchone()[0])
database.commit()
return r
def get_current_tickers(database, workdate):
basketid_list = get_basketids(database, index_list, workdate)
return get_markit_bbg_mapping(database, basketid_list, workdate)
def insert_cds(database, workdate):
all_tickers, markit_bbg_mapping = get_current_tickers(database, workdate)
filename = "cds eod {0}.csv".format(datetime.datetime.strftime(workdate, "%Y%m%d"))
colnames = ['Upfront'+tenor for tenor in ['6m', '1y', '2y', '3y', '4y', '5y', '7y', '10y']]
sqlstr = "INSERT INTO cds_quotes(date, curve_ticker, upfrontbid, upfrontask," \
"runningbid, runningask, source, recovery) VALUES(%s, %s, %s, %s, %s, %s, %s, %s)"
tickers_found = set([])
with database.cursor() as c:
c.execute("DELETE from cds_quotes where date=%s", (workdate,))
database.commit()
with open(os.path.join(root, "Tranche_data", "CDS", filename)) as fh:
csvreader = csv.DictReader(fh)
with database.cursor() as c:
for line in csvreader:
k = (line['Ticker'], line['Tier'], line['Ccy'], line['DocClause'], float(line['RunningCoupon']))
if k in markit_bbg_mapping:
for curves in markit_bbg_mapping[k]:
c.executemany(sqlstr,
[(workdate, t, convert(line[colnames[i]]), convert(line[colnames[i]]),
float(line['RunningCoupon'])*10000, float(line['RunningCoupon'])*10000,
'MKIT', convert(line['RealRecovery'])/100)
for i, t in enumerate(curves)])
tickers_found.add((line['Ticker'], line['Tier']))
database.commit()
print(all_tickers-tickers_found)
def insert_index(engine, workdate):
basedir = os.path.join(root, 'Tranche_data', 'Composite_reports')
filenames = [os.path.join(basedir, f) for f in os.listdir(basedir) if 'Indices' in f]
name_mapping = {"CDXNAHY":"HY", "CDXNAIG":"IG",'iTraxx Eur': "EU", 'iTraxx Eur Xover': "XO"}
cols = ['closeprice', 'closespread', 'modelprice', 'modelspread']
colmapping={'Date':'date', 'Name': 'index', 'Series': 'series', 'Version': 'version',
'Term': 'tenor', 'Composite Price': 'closeprice', 'Composite Spread': 'closespread',
'Model Price': 'modelprice', 'Model Spread': 'modelspread'}
ext_cols = ['date', 'index', 'series', 'version', 'tenor'] + cols + \
['adjcloseprice', 'adjmodelprice']
for f in filenames:
if datetime.datetime.fromtimestamp(os.path.getmtime(f)).date()==(workdate+BDay(1)).date():
data = pd.read_csv(f, skiprows=2, parse_dates=[0,7], engine='python')
data.rename(columns=colmapping, inplace=True)
data.dropna(subset=['closeprice'], inplace=True)
for col in cols:
data[col] = data[col].str.replace('%', '').astype('float')
data['tenor'] = data['tenor'].apply(lambda x: x.lower()+'r')
data['index'] = data['index'].apply(lambda x: name_mapping[x] if x in name_mapping else np.NaN)
data = data.dropna(subset=['index'])
data.set_index('index', drop=False, inplace=True)
data['closespread'] *= 100
data['modelspread'] *= 100
## we renumbered the version for HY9, 10 and 11
data.loc[data.series.isin([9, 10, 11]) & (data.index=='HY'),'version'] -= 3
data['adjcloseprice'] = data['closeprice']
data['adjmodelprice'] = data['modelprice']
data = data.groupby(['index', 'series', 'tenor', 'date']).last()
data.reset_index(inplace=True)
data[ext_cols].to_sql('index_quotes', engine, if_exists='append', index=False)
if __name__=="__main__":
if len(sys.argv)>=2:
workdate = datetime.datetime.strptime(sys.argv[1], "%Y-%m-%d")
else:
workdate = datetime.datetime.today()-BDay(1)
workdate = workdate.date()
engine = create_engine('postgresql://serenitas_user@debian/serenitasdb')
#insert_cds(serenitasdb, workdate)
insert_index(engine, workdate)
#serenitasdb.close()
# for f in os.listdir(os.path.join(root, "Tranche_data", "CDS")):
# if f.endswith("csv"):
# workdate = datetime.datetime.strptime(f.split(" ")[2].split(".")[0], "%Y%m%d")
# workdate = workdate.date()
# insert_cds(serenitasdb, workdate)
# serenitasdb.close()
|
