macrosynergy 0.0.29

Macrosynergy Quant Research Package

Macrosynergy Quant Research

The Macrosynergy package supports financial market research and the development of trading strategies based on formats and conventions of the J.P. Morgan Macrosynergy Quantamental System (JPMaQS). JPMaQS provides quantitative-fundamental (quantamental) and market data in simple daily formats in accordance with the information state of markets. The Macrosynergy package consists of six sub-packages:

1. management: simulates, analyses and reshapes standard quantamental dataframes.
2. panel: analyses and visualizes panels of quantamental data.
3. signal: transforms quantamental indicators into trading signals and does naive analysis.
4. pnl: constructs portfolios based on signals, applies risk management and analyses realistic PnLs.
5. download: interface for downloading data from JP Morgan DataQuery, with main module jpmaqs.py.
6. dataquery: [DEPRECATED] interface for downloading data from JP Morgan DataQuery, with main module api.py.

Installation

The easiest method for installing the package is to use the PyPI installation method:

pip install macrosynergy

Alternatively for the cutting edge development version, install the package from the develop branch as

pip install git+https://github.com/macrosynergy/macrosynergy@develop

Usage

DataQuery Interface

To download data from JP Morgan DataQuery, you can use the JPMaQSDownload Object together with your OAuth authentication credentials (default):

import pandas as pd
from macrosynergy.download import JPMaQSDownload

with JPMaQSDownload(
        client_id="<dq_client_id>",
        client_secret="<dq_client_secret>"
) as downloader:
    data = downloader.download(tickers="EUR_FXXR_NSA", 
                                start_date="2022-01-01")

assert isinstance(data, pd.DataFrame) and not data.empty

assert data.shape[0] > 0
data.info()

Alternatively, you can also the certificate and private key pair, to access DataQuery as:

import pandas as pd
from macrosynergy.download import JPMaQSDownload

with JPMaQSDownload(
        oauth=False,
        username="<dq_username>",
        password="<dq_password>",
        crt="<path_to_dq_certificate>",
        key="<path_to_dq_key>"
) as downloader:
    data = downloader.download(tickers="EUR_FXXR_NSA", 
                                start_date="2022-01-01")

assert isinstance(data, pd.DataFrame) and not data.empty

assert data.shape[0] > 0
data.info()

Both of the above example will download a snippet of example data from the premium JPMaQS dataset of the daily timeseries of EUR FX excess returns.

Using the API you can also access a panel of tickers from different countries like so.

import pandas as pd
from macrosynergy.download import JPMaQSDownload

cids = ['EUR','GBP','USD']
xcats = ['FXXR_NSA','EQXR_NSA']
tickers = [cid+"_"+xcat for cid in cids for xcat in xcats]

with JPMaQSDownload(
        client_id="<dq_client_id>",
        client_secret="<dq_client_secret>"
) as downloader:
    data = downloader.download(tickers=tickers,
                                start_date="2022-01-01")

assert isinstance(data, pd.DataFrame) and not data.empty

assert data.shape[0] > 0
data.info()

It is also possible to use a proxy server with the Dataquery interface. Here's an example:

import pandas as pd
from macrosynergy.download import JPMaQSDownload

cids = ['EUR','GBP','USD']
xcats = ['FXXR_NSA','EQXR_NSA']
tickers = [cid+"_"+xcat for cid in cids for xcat in xcats]

oauth_proxy="https://secureproxy.example.com:port"
proxy = {"https": oauth_proxy}
# or proxy = {"http": "http://proxy.example.com:port"}
with JPMaQSDownload(
        client_id = "<dq_client_id>",
        client_secret = "<dq_client_secret>",
        proxy = proxy
) as downloader:
    data = downloader.download(tickers = tickers, start_date="2022-01-01")

assert isinstance(data, pd.DataFrame) and not df.empty

or,

...
proxies = {
    "http": "http://proxy.example.com:port",
    "https": "https://secucreproxy.example.com:port",
}
with JPMaQSDownload(
        client_id = "<dq_client_id>",
        client_secret = "<dq_client_secret>",
        proxy = proxies
) as downloader:
    data = downloader.download(tickers = tickers)
...

The deprecated DataQuery interface internally makes use of the JPMaQSDownload object. Using the deprecated path will raise a deprecation warning. The path and module will be removed in v0.1.0. Here's an example of the deprecated interface:

...
from macrosynergy.dataquery import api
...
with api.Interface(
        oauth=True,
        username="<dq_username>",
        password="<dq_password>"
) as dq:
    data = dq.download(tickers=tickers, start_date="2022-01-01")
...

Management

In order to use the rest of the package without access to the API you can simulate quantamental data using the management sub-package.

from macrosynergy.management.simulate_quantamental_data import make_qdf

cids = ['AUD', 'GBP', 'NZD', 'USD']
xcats = ['FXXR_NSA', 'FXCRY_NSA', 'FXCRR_NSA', 'EQXR_NSA', 'EQCRY_NSA', 'EQCRR_NSA',
             'FXWBASE_NSA', 'EQWBASE_NSA']

df_cids = pd.DataFrame(index=cids, columns=['earliest', 'latest', 'mean_add',
                                                'sd_mult'])

df_cids.loc['AUD'] = ['2000-01-01', '2022-03-14', 0, 1]
df_cids.loc['GBP'] = ['2001-01-01', '2022-03-14', 0, 2]
df_cids.loc['NZD'] = ['2002-01-01', '2022-03-14', 0, 3]
df_cids.loc['USD'] = ['2000-01-01', '2022-03-14', 0, 4]

 df_xcats = pd.DataFrame(index=xcats, columns=['earliest', 'latest', 'mean_add',
                                                  'sd_mult', 'ar_coef', 'back_coef'])
df_xcats.loc['FXXR_NSA'] = ['2010-01-01', '2022-03-14', 0, 1, 0, 0.2]
df_xcats.loc['FXCRY_NSA'] = ['2010-01-01', '2022-03-14', 1, 1, 0.9, 0.2]
df_xcats.loc['FXCRR_NSA'] = ['2010-01-01', '2022-03-14', 0.5, 0.8, 0.9, 0.2]
df_xcats.loc['EQXR_NSA'] = ['2010-01-01', '2022-03-14', 0.5, 2, 0, 0.2]
df_xcats.loc['EQCRY_NSA'] = ['2010-01-01', '2022-03-14', 2, 1.5, 0.9, 0.5]
df_xcats.loc['EQCRR_NSA'] = ['2010-01-01', '2022-03-14', 1.5, 1.5, 0.9, 0.5]
df_xcats.loc['FXWBASE_NSA'] = ['2010-01-01', '2022-02-01', 1, 1.5, 0.8, 0.5]
df_xcats.loc['EQWBASE_NSA'] = ['2010-01-01', '2022-02-01', 1, 1.5, 0.9, 0.5]
data = make_qdf(df_cids, df_xcats, back_ar=0.75)

The management sub-package can also be used to check which data is available in the dataframe.

from macrosynergy.management.check_availability import check_availability
filt_na = (data['cid'] == 'USD') & (data['real_date'] < '2015-01-01')
data_filt.loc[filt_na, 'value'] = np.nan
check_availability(df=data_filt, xcats=xcats, cids=cids)

You can also use the built-in function to reshape the data depending on the dates or tickers of your choice.

data_reduced = reduce_df(data, xcats=xcats[:-1], cids=cids[0],
                       start='2012-01-01', end='2018-01-31')

Panel

Basket

The basket class is used to calculate the returns and carries of financial contracts using various methods, a basket is created as so.

from macrosynergy.panel.basket import Basket

black = {'AUD': ['2010-01-01', '2013-12-31'], 'GBP': ['2010-01-01', '2013-12-31']}
contracts = ['AUD_FX', 'AUD_EQ', 'NZD_FX', 'GBP_EQ', 'USD_EQ']
gdp_figures = [17.0, 17.0, 41.0, 9.0, 250.0]
basket_1 = Basket(
    df=data, contracts=contracts_1, ret="XR_NSA", cry=["CRY_NSA", "CRR_NSA"],
    blacklist=black
)
basket_1.make_basket(weight_meth="equal", max_weight=0.55, basket_name="GLB_EQUAL")

Using the basket class you have access to the methods such as visulasing the weights associated with each contract, or returning the weight or basket.

basket_1.return_basket()
basket_1.return_weights()
basket_1.weight_visualiser(basket_name="GLB_EQUAL")

You can also calculate and visualise the following and more with built-in functions.

1. historic volume
2. z-scores
3. beta values
4. timeline

from macrosynergy.panel.historic_vol import historic_vol
data_historic = historic_vol(
    data, cids=cids, xcat='FXXR_NSA', lback_periods=21, lback_meth='ma', half_life=11,
    remove_zeros=True)

from macrosynergy.panel.make_zn_scores import make_zn_scores
z_mean = make_zn_scores(data, xcat='FXXR_NSA', sequential=True, cids=cids,
                      blacklist=black, iis=False, neutral='mean',
                      pan_weight=0.5, min_obs=261, est_freq="w")
z_median = make_zn_scores(data, xcat='FXXR_NSA', sequential=True, cids=cids,
                      blacklist=black, iis=False, neutral='median',
                      pan_weight=0.5, min_obs=261, est_freq="d")

from macrosynergy.panel.return_beta import return_beta
benchmark_return = "USD_FXXR_NSA"
data_hedge = return_beta(df=data, xcat='FXXR_NSA', cids=cids,
                       benchmark_return=benchmark_return, start='2010-01-01',
                       end='2020-10-30',
                       blacklist=black, meth='ols', oos=True,
                       refreq='w', min_obs=24, hedged_returns=True)
print(df_hedge)
beta_display(df_hedge=df_hedge, subplots=False)

view_timelines(data, xcats=['FXXR_NSA','FXCRY_NSA'], cids=cids[0],
                   size=(10, 5), title='AUD Return and Carry')

Signal

Signal Return Relations

The SignalReturnRelations class analyses and visualises signal and return series.

from macrosynergy.signal.signal_return import SignalReturnRelations

srn = SignalReturnRelations(data, ret="EQXR_NSA", sig="EQCRY_NSA", rival_sigs=None,
                                sig_neg=True, cosp=True, freq="M", start="2002-01-01")
srn.summary_table()

In the creation of the class you can also indicate rival signals for basic relational statistics.

r_sigs = [ "EQCRR_NSA"]
srn = SignalReturnRelations(data, "EQXR_NSA", sig="EQCRY_NSA", rival_sigs=r_sigs,
                            sig_neg=True, cosp=True, freq="M", start="2002-01-01")
df_sigs = srn.signals_table(sigs=['EQCRY_NSA_NEG', 'EQCRR_NSA_NEG'])

df_sigs_all = srn.signals_table()

Using the class you can plot accuracy bars between returns and signals.

srn.accuracy_bars(type="signals", title="Accuracy measure between target return, EQXR_NSA,"
                                        " and the respective signals, ['EQCRY_NSA_NEG', "
                                        " 'EQCRR_NSA_NEG'].")

PnL

Naive pnl

The NaivePnL class computes Pnls with limited signal options and disregarding transaction costs.

from macrosynergy.pnl.naive_pnl import NaivePnL
pnl = NaivePnL(data, ret="EQXR_NSA", sigs=["CRY", "GROWTH"], cids=cids,
        start="2000-01-01", bms=["EUR_EQXR_NSA", "USD_EQXR_NSA"])

You can then make the pnl and see a list of key pnl statistics.

pnl.make_pnl(
        sig="GROWTH", sig_op="zn_score_pan", sig_neg=True, rebal_freq="monthly",
        vol_scale=5, rebal_slip=1, min_obs=250, thresh=2)
df_eval = pnl.evaluate_pnls(
        pnl_cats=["PNL_GROWTH_NEG"], start="2015-01-01", end="2020-12-31")

Documentation

The official documentation can be found at our website: https://docs.macrosynergy.com