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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
|
import datetime
import io
import pandas as pd
import requests
import zipfile
from serenitas.utils.db import serenitas_engine, dawn_engine
from serenitas.utils.env import DATA_DIR
from typing import Dict, Any
from utils.db import dbconn
from pickle import dumps, loads
from psycopg2.extensions import connection
def download_credit_slices(d: datetime.date) -> None:
for i in range(1, 400):
url = f"https://kgc0418-tdw-data2-0.s3.amazonaws.com/slices/SLICE_CREDITS_{d:%Y_%m_%d}_{i}.zip"
r = requests.get(url)
if r.status_code != 200:
continue
with zipfile.ZipFile(io.BytesIO(r.content)) as z:
z.extractall(path=DATA_DIR / "DTCC")
def download_cumulative_credit(d: datetime.date) -> None:
url = f"https://kgc0418-tdw-data2-0.s3.amazonaws.com/slices/CUMULATIVE_CREDITS_{d:%Y_%m_%d}.zip"
r = requests.get(url)
if r.status_code != 200:
return
with zipfile.ZipFile(io.BytesIO(r.content)) as z:
z.extractall(path=DATA_DIR / "DTCC")
def load_data():
base_dir = DATA_DIR / "DTCC"
df = pd.concat(
[
pd.read_csv(
f,
parse_dates=["EXECUTION_TIMESTAMP", "EFFECTIVE_DATE", "END_DATE"],
thousands=",",
index_col=["DISSEMINATION_ID"],
)
for f in base_dir.glob("*.csv")
]
)
for col in [
"ACTION",
"CLEARED",
"PRICE_NOTATION_TYPE",
"OPTION_TYPE",
"OPTION_CURRENCY",
"INDICATION_OF_COLLATERALIZATION",
"EXECUTION_VENUE",
"DAY_COUNT_CONVENTION",
"NOTIONAL_CURRENCY_1",
"SETTLEMENT_CURRENCY",
]:
df[col] = df[col].astype("category")
df.ORIGINAL_DISSEMINATION_ID = df.ORIGINAL_DISSEMINATION_ID.astype("Int64")
df.UNDERLYING_ASSET_1 = df.UNDERLYING_ASSET_1.str.rsplit(":", n=1, expand=True)[
1
].astype("str")
df = df[~df.index.isin(df.ORIGINAL_DISSEMINATION_ID)]
df = df[df.ACTION != "CANCEL"]
del df["ASSET_CLASS"]
del df["ACTION"]
return df
def apply_corrections(conn: connection, df):
with conn.cursor() as c:
c.execute("SELECT * FROM dtcc_corrections")
corrections = {did: correction for did, correction in c}
conn.commit()
for k, v in corrections.items():
v = loads(v)
for col, val in v.items():
df.at[k, col] = val
return df
def process_option_data(conn: connection, df):
df = df[df.OPTION_FAMILY.notnull()]
df = df.dropna(axis=1, how="all")
del df["OPTION_FAMILY"]
df = apply_corrections(conn, df)
for col in [
"INDICATION_OF_END_USER_EXCEPTION",
"INDICATION_OF_OTHER_PRICE_AFFECTING_TERM",
"BLOCK_TRADES_AND_LARGE_NOTIONAL_OFF-FACILITY_SWAPS",
]:
df[col] = df[col].map({"N": False, "Y": True})
for col in ["EFFECTIVE_DATE", "OPTION_EXPIRATION_DATE", "OPTION_LOCK_PERIOD"]:
df[col] = pd.to_datetime(df[col], errors="raise")
df = df.rename(
columns={
"OPTION_STRIKE_PRICE": "strike",
"OPTION_EXPIRATION_DATE": "expiration_date",
"UNDERLYING_ASSET_1": "redindexcode",
"ROUNDED_NOTIONAL_AMOUNT_1": "notional",
"OPTION_PREMIUM": "premium",
"OPTION_TYPE": "option_type",
"PRICE_NOTATION": "price",
"EXECUTION_TIMESTAMP": "trade_timestamp",
}
)
df.strike = df.strike.where(df.strike < 1000, df.strike / 100).where(
df.strike > 10, df.strike * 100
)
df.price = (df.price * 1e2).where(df.PRICE_NOTATION_TYPE == "Percentage", df.price)
conn = dbconn("serenitasdb")
df_indic = pd.read_sql_query(
"SELECT redindexcode, index, series, version FROM index_version", conn
)
conn.close()
df = df.merge(df_indic, on="redindexcode")
df = (
df.reset_index()
.set_index(["index", "series", "version", "trade_timestamp"])
.sort_index()
)
return df[
["expiration_date", "notional", "strike", "option_type", "premium", "price"]
]
def process_tranche_data(df):
df = df[df.TAXONOMY.str.startswith("Credit:IndexTranche")]
df = df.loc[:, ~df.columns.str.contains("OPTION")]
df.sort_values("EXECUTION_TIMESTAMP", inplace=True)
return df
def map_tranche(df):
idx_ver = pd.read_sql_query(
"select index, series, redindexcode, maturity, tenor " "from index_desc",
serenitas_engine,
parse_dates="maturity",
index_col=["redindexcode", "maturity"],
)
idx_ver = idx_ver.loc[idx_ver.index.dropna()]
markit = pd.read_sql_query(
"select * from markit_tranche_quotes a "
"left join (select series, index, basketid "
"as id from index_version) b on a.basketid=b.id",
serenitas_engine,
parse_dates="quotedate",
)
df = df.reset_index().set_index(["UNDERLYING_ASSET_1", "END_DATE"])
df.rename_axis(
index={"UNDERLYING_ASSET_1": "redindexcode", "END_DATE": "maturity"},
inplace=True,
)
df = df.merge(idx_ver, left_index=True, right_index=True, how="left")
df["EXECUTION_TIMESTAMP"] = df.EXECUTION_TIMESTAMP.dt.normalize()
# Normalize the price format
df.PRICE_NOTATION = (df.PRICE_NOTATION / 1e2).where(
df.PRICE_NOTATION_TYPE == "Percentage", df.PRICE_NOTATION
)
df["price"] = (df.PRICE_NOTATION / 1e2).where(
df["index"].isin(["IG", "EU", "XO"]), 1 - df.PRICE_NOTATION / 1e2
)
# Allow for bigger bid/offers for equity tranche
markit.upfront_bid = (markit.upfront_mid - 0.01).where(
markit.attach == 0, markit.upfront_bid
)
markit.upfront_ask = (markit.upfront_mid + 0.01).where(
markit.attach == 0, markit.upfront_ask
)
df = df.merge(
markit,
left_on=["index", "series", "tenor", "EXECUTION_TIMESTAMP"],
right_on=["index", "series", "tenor", "quotedate"],
how="outer",
)
df["identified"] = df.apply(
lambda df: (df.price > df.upfront_bid) & (df.price < df.upfront_ask)
if df.upfront_bid > 0
else (df.price < df.upfront_bid) & (df.price > df.upfront_ask),
axis=1,
)
df = df[df.identified]
df = df[~df.duplicated(["DISSEMINATION_ID"], keep=False)]
return df.sort_values("EXECUTION_TIMESTAMP")
def insert_correction(conn: connection, dissemination_id: int, **kwargs) -> None:
with conn.cursor() as c:
c.execute(
"INSERT INTO dtcc_corrections VALUES(%s, %s)",
(dissemination_id, dumps(kwargs)),
)
conn.commit()
def get_correction(conn: connection, dissemination_id: int) -> Dict[str, Any]:
with conn.cursor() as c:
c.execute(
"SELECT correction FROM dtcc_corrections WHERE dissemination_id=%s",
(dissemination_id,),
)
(pick,) = c.fetchone()
conn.commit()
return loads(pick)
if __name__ == "__main__":
import argparse
parser = argparse.ArgumentParser()
parser.add_argument(
"workdate",
nargs="?",
type=lambda s: pd.datetime.strptime(s, "%Y-%m-%d").date(),
default=datetime.date.today(),
)
parser.add_argument("-s", "--slice", action="store_true", help="download slice")
args = parser.parse_args()
if args.slice:
download_credit_slices(args.workdate)
else:
download_cumulative_credit(args.workdate)
# from utils.db import serenitas_pool
# serenitasdb = serenitas_pool.getconn()
# insert_correction(serenitasdb, 107282774, OPTION_EXPIRATION_DATE="2019-09-18")
# get_correction(serenitasdb, 107282774)
dr = pd.bdate_range("2018-01-01", "2019-02-11")
for d in dr:
download_cumulative_credit(d)
df = load_data()
df = process_tranche_data(df)
df = map_tranche(df)
|
