Skip to content

Commit

Permalink
ENH Allow non-daily timeseries to be passed into annual_return and an…
Browse files Browse the repository at this point in the history 
…nual_volatility calcs.

Also refactors how DAILY, MONTHLY and YEARYL frequencies are encoded.
  • Loading branch information
@twiecki
Calvin DeBoer authored and twiecki committed Oct 14, 2015
1 parent 889eba0 commit ee928a3
Show file tree
Hide file tree
Showing 3 changed files with 157 additions and 38 deletions.
51 changes: 40 additions & 11 deletions pyfolio/tests/test_timeseries.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -210,6 +210,21 @@ class TestStats(TestCase):
'2000-1-3',
periods=500,
freq='D'))

simple_week_rets = pd.Series(
[0.1] * 3 + [0] * 497,
pd.date_range(
'2000-1-31',
periods=500,
freq='W'))

simple_month_rets = pd.Series(
[0.1] * 3 + [0] * 497,
pd.date_range(
'2000-1-31',
periods=500,
freq='M'))

simple_benchmark = pd.Series(
[0.03] * 4 + [0] * 496,
pd.date_range(
Expand All @@ -221,25 +236,39 @@ class TestStats(TestCase):
dt = pd.date_range('2000-1-3', periods=3, freq='D')

@parameterized.expand([
(simple_rets, 'calendar', 0.10584000000000014),
(simple_rets, 'compound', 0.16317653888658334),
(simple_rets, 'calendar', 0.10584000000000014),
(simple_rets, 'compound', 0.16317653888658334)
(simple_rets, 'calendar', utils.DAILY, 0.10584000000000014),
(simple_rets, 'compound', utils.DAILY, 0.16317653888658334),
(simple_rets, 'calendar', utils.DAILY, 0.10584000000000014),
(simple_rets, 'compound', utils.DAILY, 0.16317653888658334),
(simple_week_rets, 'compound', utils.WEEKLY, 0.031682168889005213),
(simple_week_rets, 'calendar', utils.WEEKLY, 0.021840000000000033),
(simple_month_rets, 'compound', utils.MONTHLY, 0.0072238075842128158),
(simple_month_rets, 'calendar', utils.MONTHLY, 0.0050400000000000071)
])
def test_annual_ret(self, returns, style, expected):
def test_annual_ret(self, returns, style, period, expected):
self.assertEqual(
timeseries.annual_return(
returns,
style=style),
style=style, period=period),
expected)

@parameterized.expand([
(simple_rets, 0.12271674212427248),
(simple_rets, 0.12271674212427248)
(simple_rets, utils.DAILY, 0.12271674212427248),
(simple_rets, utils.DAILY, 0.12271674212427248),
(simple_week_rets, utils.WEEKLY, 0.055744909991675112),
(simple_week_rets, utils.WEEKLY, 0.055744909991675112),
(simple_month_rets, utils.MONTHLY, 0.026778988562993072),
(simple_month_rets, utils.MONTHLY, 0.026778988562993072)
])
def test_annual_volatility(self, returns, expected):
self.assertAlmostEqual(timeseries.annual_volatility(returns),
expected, DECIMAL_PLACES)
def test_annual_volatility(self, returns, period, expected):
self.assertAlmostEqual(
timeseries.annual_volatility(
returns,
period=period
),
expected,
DECIMAL_PLACES
)

@parameterized.expand([
(simple_rets, 'calendar', 0.8624740045072119),
Expand Down
130 changes: 103 additions & 27 deletions pyfolio/timeseries.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,7 @@

from . import utils
from .utils import APPROX_BDAYS_PER_MONTH, APPROX_BDAYS_PER_YEAR
from .utils import DAILY, WEEKLY, MONTHLY, YEARLY, ANNUALIZATION_FACTORS
from .interesting_periods import PERIODS


Expand Down Expand Up @@ -135,19 +136,21 @@ def aggregate_returns(df_daily_rets, convert_to):
def cumulate_returns(x):
return cum_returns(x)[-1]

if convert_to == 'weekly':
if convert_to == WEEKLY:
return df_daily_rets.groupby(
[lambda x: x.year,
lambda x: x.month,
lambda x: x.isocalendar()[1]]).apply(cumulate_returns)
elif convert_to == 'monthly':
elif convert_to == MONTHLY:
return df_daily_rets.groupby(
[lambda x: x.year, lambda x: x.month]).apply(cumulate_returns)
elif convert_to == 'yearly':
elif convert_to == YEARLY:
return df_daily_rets.groupby(
[lambda x: x.year]).apply(cumulate_returns)
else:
ValueError('convert_to must be weekly, monthly or yearly')
ValueError(
'convert_to must be {}, {} or {}'.format(WEEKLY, MONTHLY, YEARLY)
)


def max_drawdown(returns):
Expand Down Expand Up @@ -188,20 +191,24 @@ def max_drawdown(returns):
return -1 * MDD


def annual_return(returns, style='compound'):
def annual_return(returns, style='compound', period=DAILY):
"""Determines the annual returns of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
Periodic returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
style : str, optional
- If 'compound', then return will be calculated in geometric
terms: (1+mean(all_daily_returns))^252 - 1.
- If 'calendar', then return will be calculated as
((last_value - start_value)/start_value)/num_of_years.
- Otherwise, return is simply mean(all_daily_returns)*252.
period : str, optional
- defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'
- defaults to 'daily'.
Returns
-------
Expand All @@ -213,27 +220,41 @@ def annual_return(returns, style='compound'):
if returns.size < 1:
return np.nan

try:
ann_factor = ANNUALIZATION_FACTORS[period]
except KeyError:
raise ValueError(
"period cannot be '{}'. "
"Must be '{}', '{}', or '{}'".format(
period, DAILY, WEEKLY, MONTHLY
)
)

if style == 'calendar':
num_years = len(returns) / APPROX_BDAYS_PER_YEAR
num_years = len(returns) / ann_factor
df_cum_rets = cum_returns(returns, starting_value=100)
start_value = df_cum_rets[0]
end_value = df_cum_rets[-1]
return ((end_value - start_value) / start_value) / num_years
if style == 'compound':
return pow((1 + returns.mean()), APPROX_BDAYS_PER_YEAR) - 1
return pow((1 + returns.mean()), ann_factor) - 1
else:
return returns.mean() * APPROX_BDAYS_PER_YEAR
return returns.mean() * ann_factor


def annual_volatility(returns):
def annual_volatility(returns, period=DAILY):
"""
Determines the annual volatility of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
Periodic returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
period : str, optional
- defines the periodicity of the 'returns' data for purposes of
annualizing volatility. Can be 'monthly' or 'weekly' or 'daily'.
- defaults to 'daily'
Returns
-------
Expand All @@ -244,10 +265,20 @@ def annual_volatility(returns):
if returns.size < 2:
return np.nan

return returns.std() * np.sqrt(APPROX_BDAYS_PER_YEAR)
try:
ann_factor = ANNUALIZATION_FACTORS[period]
except KeyError:
raise ValueError(
"period cannot be: '{}'."
" Must be '{}', '{}', or '{}'".format(
period, DAILY, WEEKLY, MONTHLY
)
)

return returns.std() * np.sqrt(ann_factor)


def calmar_ratio(returns, returns_style='calendar'):
def calmar_ratio(returns, returns_style='calendar', period=DAILY):
"""
Determines the Calmar ratio, or drawdown ratio, of a strategy.
Expand All @@ -258,6 +289,11 @@ def calmar_ratio(returns, returns_style='calendar'):
- See full explanation in tears.create_full_tear_sheet.
returns_style : str, optional
See annual_returns' style
period : str, optional
- defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'
- defaults to 'daily'.
Returns
-------
Expand All @@ -273,7 +309,9 @@ def calmar_ratio(returns, returns_style='calendar'):
if temp_max_dd < 0:
temp = annual_return(
returns=returns,
style=returns_style) / abs(max_drawdown(returns=returns))
style=returns_style,
period=period
) / abs(max_drawdown(returns=returns))
else:
return np.nan

Expand Down Expand Up @@ -321,7 +359,7 @@ def omega_ratio(returns, annual_return_threshhold=0.0):
return np.nan


def sortino_ratio(returns, required_return=0):
def sortino_ratio(returns, required_return=0, period=DAILY):

"""
Determines the Sortino ratio of a strategy.
Expand All @@ -331,9 +369,14 @@ def sortino_ratio(returns, required_return=0):
returns : pd.Series or pd.DataFrame
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
returns_style : str, optional
See annual_returns' style
required_return: float / series
minimum acceptable return
period : str, optional
- defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'
- defaults to 'daily'.
Returns
-------
Expand All @@ -344,14 +387,24 @@ def sortino_ratio(returns, required_return=0):
Annualized Sortino ratio.
"""
try:
ann_factor = ANNUALIZATION_FACTORS[period]
except KeyError:
raise ValueError(
"period cannot be: '{}'."
" Must be '{}', '{}', or '{}'".format(
period, DAILY, WEEKLY, MONTHLY
)
)

mu = np.nanmean(returns - required_return, axis=0)
sortino = mu / downside_risk(returns, required_return)
if len(returns.shape) == 2:
sortino = pd.Series(sortino, index=returns.columns)
return sortino * APPROX_BDAYS_PER_YEAR
return sortino * ann_factor


def downside_risk(returns, required_return=0):
def downside_risk(returns, required_return=0, period=DAILY):
"""
Determines the downside deviation below a threshold
Expand All @@ -363,6 +416,10 @@ def downside_risk(returns, required_return=0):
required_return: float / series
minimum acceptable return
period : str, optional
- defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'
- defaults to 'daily'.
Returns
-------
Expand All @@ -373,18 +430,28 @@ def downside_risk(returns, required_return=0):
Annualized downside deviation
"""
try:
ann_factor = ANNUALIZATION_FACTORS[period]
except KeyError:
raise ValueError(
"period cannot be: '{}'."
" Must be '{}', '{}', or '{}'".format(
period, DAILY, WEEKLY, MONTHLY
)
)

downside_diff = returns - required_return
mask = downside_diff > 0
downside_diff[mask] = 0.0
squares = np.square(downside_diff)
mean_squares = np.nanmean(squares, axis=0)
dside_risk = np.sqrt(mean_squares) * np.sqrt(APPROX_BDAYS_PER_YEAR)
dside_risk = np.sqrt(mean_squares) * np.sqrt(ann_factor)
if len(returns.shape) == 2:
dside_risk = pd.Series(dside_risk, index=returns.columns)
return dside_risk


def sharpe_ratio(returns, returns_style='compound'):
def sharpe_ratio(returns, returns_style='compound', period=DAILY):
"""
Determines the Sharpe ratio of a strategy.
Expand All @@ -395,6 +462,10 @@ def sharpe_ratio(returns, returns_style='compound'):
- See full explanation in tears.create_full_tear_sheet.
returns_style : str, optional
See annual_returns' style
period : str, optional
- defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'
- defaults to 'daily'.
Returns
-------
Expand All @@ -406,8 +477,8 @@ def sharpe_ratio(returns, returns_style='compound'):
See https://en.wikipedia.org/wiki/Sharpe_ratio for more details.
"""

numer = annual_return(returns, style=returns_style)
denom = annual_volatility(returns)
numer = annual_return(returns, style=returns_style, period=period)
denom = annual_volatility(returns, period=period)

if denom > 0.0:
return numer / denom
Expand Down Expand Up @@ -661,7 +732,8 @@ def calc_alpha_beta(returns, factor_returns):
def perf_stats(
returns,
returns_style='compound',
return_as_dict=False):
return_as_dict=False,
period=DAILY):
"""Calculates various performance metrics of a strategy, for use in
plotting.show_perf_stats.
Expand All @@ -674,6 +746,10 @@ def perf_stats(
See annual_returns' style
return_as_dict : boolean, optional
If True, returns the computed metrics in a dictionary.
period : str, optional
- defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'
- defaults to 'daily'.
Returns
-------
Expand All @@ -685,14 +761,14 @@ def perf_stats(
all_stats = OrderedDict()
all_stats['annual_return'] = annual_return(
returns,
style=returns_style)
all_stats['annual_volatility'] = annual_volatility(returns)
style=returns_style, period=period)
all_stats['annual_volatility'] = annual_volatility(returns, period=period)
all_stats['sharpe_ratio'] = sharpe_ratio(
returns,
returns_style=returns_style)
returns_style=returns_style, period=period)
all_stats['calmar_ratio'] = calmar_ratio(
returns,
returns_style=returns_style)
returns_style=returns_style, period=period)
all_stats['stability'] = stability_of_timeseries(returns)
all_stats['max_drawdown'] = max_drawdown(returns)
all_stats['omega_ratio'] = omega_ratio(returns)
Expand Down
Loading

0 comments on commit ee928a3

Please sign in to comment.