/
strategy_optimised.py
391 lines (288 loc) · 13.6 KB
/
strategy_optimised.py
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
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
"""Optimised version of strategy.py to calculate performance of
a leveraged investment strategy"""
import datetime as dt
import math
import time
from itertools import product
import numpy as np
import pandas as pd
import simulate
from debt import Debt
debt_available = {'SU': Debt(), 'Nordnet': Debt()}
def determine_investment(phase, pv_u, tv_u, s, td, pi_rf, dst, g, period):
# returns cash, new_equity, new_debt
if phase == 1:
# Check if gearing cap has been reached
equity = tv_u + s - td
if td > (equity * g):
new_debt = 0
else:
new_debt = nd(g, s, tv_u, td, dst, period)
return 0, s, new_debt
if phase == 2:
stocks_sold = max(pv_u - dst, 0)
debt_repayment = min(td, s + stocks_sold)
repayment_left = debt_repayment
if 'Nordnet' in debt_available.keys():
repayment = min(debt_repayment, debt_available['Nordnet'].debt_amount)
debt_available['Nordnet'].prepayment(repayment)
repayment_left = debt_repayment - repayment
if 'SU' in debt_available.keys():
debt_available['SU'].prepayment(repayment_left)
leftover_savings = max(s - debt_repayment - stocks_sold, 0)
return 0, leftover_savings, -debt_repayment
if phase == 3:
return 0, s, 0
if phase == 4:
desired_cash = (1 - pi_rf) * (tv_u + s)
desired_savings = (pi_rf) * (tv_u + s)
change_in_stock = desired_savings - pv_u
return desired_cash, change_in_stock, 0
# Function assumes monthly periods
def nd(g, s, tv_u, td, dst, period):
equity = tv_u + s - td
total_desired_debt = min(g / (g + 1) * dst, equity * g)
remaining_debt_needed = max(0, total_desired_debt - td)
SU_amount, Nordnet_amount = 0, 0
if period <= 60 and 'SU' in debt_available.keys():
# Has SU already been taken?
SU_amount = min(3234, remaining_debt_needed)
debt_available['SU'].add_debt(SU_amount)
remaining_debt_needed -= SU_amount
if 'Nordnet' in debt_available.keys():
# Has Nordnet already been taken?
Nordnet_amount = min(max(0, g * equity), remaining_debt_needed)
debt_available['Nordnet'].add_debt(Nordnet_amount)
return SU_amount + Nordnet_amount
def interest_all_debt():
interest_bill = 0
for debt in debt_available.values():
interest_bill += debt.calculate_interest()
return interest_bill
def phase_check(phase, pi_rf, pi_rm, pi_hat, td):
if phase == 4:
return 4
if td > 0:
# has target not been reached?
if pi_hat < pi_rm and phase <= 1:
return 1
# if target has been reached once and debt remains, stay in phase 2
return 2
# if target has been reached and no debt remains
# is the value still above the target?
if pi_hat < pi_rf:
return 3
return 4
def calc_pi(gamma, sigma, mr, rate, cost=0):
return (mr - cost - rate) / (gamma * sigma)
def calculate_return(savings_in, returns, gearing_cap, pi_rf, pi_rm, rf):
# Running controls
len_savings = len(savings_in)
assert len_savings == len(returns), 'Investment plan should be same no of periods as market'
# Setting up constants and dataframe for calculation
ses_val = savings_in.sum() # Possibly add more sophisticated discounting
ist = pi_rm * ses_val
columns = ['period', 'savings', 'cash', 'new_equity', 'new_debt', 'total_debt', 'nip', 'pv_p',
'interest', 'market_returns', 'pv_u', 'tv_u', 'equity', 'dst', 'phase', 'pi_hat',
'g_hat', 'SU_debt', 'Nordnet_debt']
len_columns = len(columns)
pp = np.zeros((len_savings, len_columns))
period, savings, cash, new_equity, new_debt, total_debt, nip, pv_p, interest, market_returns, pv_u, tv_u, equity, dst, phase, pi_hat, g_hat, SU_debt, Nordnet_debt = range(
len_columns)
pp[:, period] = range(len_savings)
pp[:, market_returns] = returns
pp[:, savings] = savings_in
pp[0, market_returns] = 0
# Initializing debt objects
if 'SU' in debt_available.keys():
debt_available['SU'] = Debt(rate_structure=[[0, 0, 0.04]], rate_structure_type='relative', initial_debt=0)
if 'Nordnet' in debt_available.keys():
debt_available['Nordnet'] = Debt(rate_structure=[[0, .4, 0.02], [.4, .6, 0.03], [.6, 0, 0.07]],
rate_structure_type='relative', initial_debt=0)
# Period 0 primo
pp[0, cash] = 0
pp[0, new_equity] = pp[0, savings]
pp[0, new_debt] = pp[0, new_equity] * gearing_cap
pp[0, total_debt] = pp[0, new_debt]
pp[0, SU_debt] = min(pp[0, new_debt], 3248)
pp[0, Nordnet_debt] = max(0, pp[0, new_debt] - 3248)
pp[0, nip] = pp[0, new_debt] + pp[0, new_equity]
pp[0, pv_p] = pp[0, nip]
pp[0, pi_hat] = pp[0, pv_p] / ses_val
# Period 0 ultimo
pp[0, interest] = pp[0, new_debt] * max(interest_all_debt(), 0)
pp[0, pv_u] = pp[0, pv_p]
pp[0, tv_u] = pp[0, pv_u] + pp[0, cash]
pp[0, equity] = pp[0, tv_u] - pp[0, total_debt]
pp[0, dst] = ist
pp[0, phase] = 1
# Looping over all reminaning periods
for i in range(1, len_savings):
# Period t > 0 primo
if not (pp[i - 1, tv_u] <= 0 and (pp[i - 1, interest] > pp[i, savings])):
pp[i, cash] = pp[i - 1, cash] * (1 + rf)
pp[i, cash], pp[i, new_equity], pp[i, new_debt] = determine_investment(
pp[i - 1, phase], pp[i - 1, pv_u],
pp[i - 1, tv_u], pp[i, savings], pp[i - 1, total_debt],
pi_rf, pp[i - 1, dst], gearing_cap, pp[i, period])
if 'SU' in debt_available.keys():
pp[i, SU_debt] = debt_available['SU'].debt_amount
if 'Nordnet' in debt_available.keys():
pp[i, Nordnet_debt] = debt_available['Nordnet'].debt_amount
pp[i, total_debt] = pp[i - 1, total_debt] + pp[i, new_debt]
pp[i, nip] = pp[i, new_equity] + max(0, pp[i, new_debt])
pp[i, pv_p] = pp[i - 1, pv_u] + pp[i, nip]
# Period t > 0 ultimo
if pp[i, period] == 60 and 'SU' in debt_available.keys():
debt_available['SU'].rate_structure = [[0, 0, 0.01]]
pp[i, interest] = max(interest_all_debt(), 0)
pp[i, pv_u] = pp[i, pv_p] * (1 + pp[i, market_returns]) - pp[i, interest]
pp[i, tv_u] = pp[i, pv_u] + pp[i, cash]
pp[i, equity] = pp[i, tv_u] - pp[i, total_debt]
pp[i, pi_hat] = min(pp[i, pv_u] / ses_val, pp[i, pv_u] / pp[i, tv_u])
pp[i, phase] = phase_check(pp[i - 1, phase], pi_rf, pi_rm, pp[i, pi_hat], pp[i, total_debt])
target_pi = pi_rm if pp[i - 1, phase] < 3 else pi_rf
pp[i, dst] = max(pp[i, tv_u] * target_pi, ist) # Moving stock target
# pp[i, dst] = max(pp[i-1, dst], max(pp[i, tv_u]*target_pi, ist)) # Locked stock target at highest previous position
else:
print('Warning: Fatal wipeout')
pp[i:, [savings, cash, new_equity, new_debt, nip, pv_p,
interest, pv_u, tv_u, pi_hat, g_hat]] = 0
cols = [total_debt, SU_debt, Nordnet_debt, equity, dst, phase]
pp[i:, cols] = pp[i - 1, cols]
break
pp[:, g_hat] = pp[:, total_debt] / pp[:, equity]
pp = pd.DataFrame(pp, columns=columns)
return pp
def calculate100return(savings_in, returns):
# Running controls
len_savings = len(savings_in)
assert len_savings == len(returns), 'Investment plan should be same no of periods as market'
columns = ['period', 'savings', 'pv_p', 'market_returns', 'tv_u']
pp = np.empty((len_savings, len(columns)))
period, savings, pv_p, market_returns, tv_u = range(5)
pp[:, period] = range(len_savings)
pp[:, market_returns] = returns
pp[:, savings] = savings_in
pp[0, market_returns] = 0
pp[0, pv_p] = pp[0, savings]
pp[0, tv_u] = pp[0, savings]
for i in range(1, len_savings):
# Period t > 0 primo
pp[i, pv_p] = pp[i - 1, tv_u] + pp[i, savings]
# Period t > 0 ultimo
pp[i, tv_u] = pp[i, pv_p] * (1 + pp[i, market_returns])
pp = pd.DataFrame(pp, columns=columns)
return pp
def calculate9050return(savings_in, returns, rf):
# Strategy where 90% of value is initially invested in stocks, rest in risk free asset
# Ratio of stocks falls linearly to 50% by age 65 and stays there
# Running controls
len_savings = len(savings_in)
assert len_savings == len(returns), 'Investment plan should be same no of periods as market'
columns = ['period', 'savings', 'cash', 'pv_p', 'market_returns', 'pv_u', 'tv_u', 'ratio']
len_columns = len(columns)
pp = np.empty((len_savings, len_columns))
period, savings, cash, pv_p, market_returns, pv_u, tv_u, ratio = range(len_columns)
pp[:, period] = range(len_savings)
pp[:, market_returns] = returns
pp[:, savings] = savings_in
pp[0, market_returns] = 0
pp[0, pv_p] = pp[0, savings] * 0.9
pp[0, cash] = pp[0, savings] * 0.1
pp[0, pv_u] = pp[0, pv_p]
pp[0, tv_u] = pp[0, savings]
pp[0, ratio] = 90
for i in range(1, len_savings):
ratio_val = max(90 - pp[i, period] / 12, 50)
pp[i, ratio] = ratio_val
# Period t > 0 primo
pp[i, pv_p] = pp[i - 1, pv_u] + pp[i, savings] * (ratio_val / 100)
pp[i, cash] = pp[i - 1, cash] * (1 + rf) + pp[i, savings] * (1 - ratio_val / 100)
# Period t > 0 ultimo
pp[i, pv_u] = pp[i, pv_p] * (1 + pp[i, market_returns])
pp[i, tv_u] = pp[i, pv_u] + pp[i, cash]
pp = pd.DataFrame(pp, columns=columns)
return pp
def summary_stats(returns, values, h, annual_rf):
mean = ((values[-1] / values[0]) ** (1 / h) - 1) * 3.46410
std = returns.std() * 3.46410 # sqrt(12) = 3.46410
return np.array([mean, std, (mean - annual_rf) / std, values[-1]]).transpose()
def main(investments_in, sim_type, random_state, gearing_cap, gamma, sigma, mr,
yearly_rf, yearly_rm, cost):
vars_for_name = (sim_type, random_state, gearing_cap, gamma, sigma, mr, yearly_rf, yearly_rm, cost)
vars_for_name_len = len(vars_for_name)-1
out_str = [str(x) + '_' if vars_for_name.index(x) != vars_for_name_len else str(x) for x in vars_for_name]
try:
pd.read_pickle('sims/' + sim_type + '/' + ''.join(out_str) + '.bz2')
print('skipping...')
except FileNotFoundError:
returns = np.load('market_lookup/' + sim_type + '/' + str(random_state) + '.npy')
rf = math.exp(yearly_rf / 12) - 1
pi_rf = calc_pi(gamma, sigma, mr, yearly_rf, cost)
pi_rm = calc_pi(gamma, sigma, mr, yearly_rm, cost)
port = calculate_return(investments_in, returns, gearing_cap, pi_rf, pi_rm, rf)
port100 = calculate100return(investments_in, returns)
port9050 = calculate9050return(investments_in, returns, rf)
# Joining normal strategies on to geared
port['100'] = port100['tv_u']
port['9050'] = port9050['tv_u']
# Reducing size of port
# Setting period as index
port.set_index('period', drop=True, inplace=True)
# Dropping non-essential columns
port.drop(columns=['nip', 'pv_u', 'equity', 'pi_hat', 'g_hat'], inplace=True)
# Convert selected float columns to integer values
flt_cols = ['savings', 'cash', 'new_equity', 'new_debt', 'total_debt',
'pv_p', 'interest', 'tv_u', 'dst', 'phase', '100', '9050']
port.loc[:, flt_cols] = port.loc[:, flt_cols].astype(int)
for debt in ['SU_debt', 'Nordnet_debt']:
try:
port.loc[:, [debt]] = port.loc[:, [debt]].astype(int)
except KeyError:
pass
# Using compressed pickle to store data efficiently
port.to_pickle('sims/' + sim_type + '/' + ''.join(out_str) + '.bz2', compression="bz2")
print('dumped sim')
if __name__ == "__main__":
from multiprocessing.pool import Pool
# Creating returns to simulate
spx = pd.read_csv('^GSPC.csv', index_col=0)
savings_year = pd.read_csv('investment_plan_year.csv', sep=';', index_col=0)
savings_year.index = pd.to_datetime(savings_year.index, format='%Y')
savings_month = (savings_year.resample('BMS').pad() / 12)['Earnings'].values
YEARLY_RF = 0.015
YEARLY_RM = 0.04 # Weighted average of margin rates
# --- Fixed parameters ----
investments = savings_month * 0.05
START = dt.date(2020, 1, 1)
END = dt.date(2070, 1, 31)
Market = simulate.Market(spx.iloc[-7500:, -2], START, END)
GAMMA = 2.5
SIGMA = Market.yearly.pct_change().std() ** 2
MR = (Market.yearly[-1] / Market.yearly[0]) ** (1 / len(Market.yearly)) - 1
COST = 0
# --- End fixed parameters ----
# Creating list of arguments
a = [[investments], ['garch', 'norm', 'draw', 't'], range(100), (1, 1.5),
(1.8, 2.0, 2.2), [SIGMA], [MR], (0.01, 0.02, 0.03), (0.04, 0.05, 0.06, 0.07, 0.08), (0, 0.002)]
#main(investments_in, sim_type, random_state, gearing_cap, gamma, sigma, mr,
# yearly_rf, yearly_rm, cost)
comb_args = tuple(product(*a))
arg_iter = (i for i in comb_args)
num_sims = sum(1 for _ in arg_iter)
print('number of simulations to run: ', num_sims)
#import cProfile, pstats
#profiler = cProfile.Profile()
#profiler.enable()
#main(investments, 'garch', 1, 1.5, 2.0, SIGMA, MR, 0.01, 0.05, 0)
#profiler.disable()
#stats = pstats.Stats(profiler)
#stats.strip_dirs()
#stats.sort_stats('cumtime')
#stats.print_stats()
with Pool() as p:
tic = time.perf_counter()
p.starmap(main, comb_args)
toc = time.perf_counter()
print(f"Script took {toc - tic:0.5f} seconds")