-
Notifications
You must be signed in to change notification settings - Fork 32
/
Copy pathscript.workshop2010.R
475 lines (409 loc) · 23.6 KB
/
script.workshop2010.R
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
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
### For R/Finance workshop on Portfolio Analytics
# Chicago, 16 April 2010
# Peter Carl and Brian Peterson
### Load the necessary packages
# Include optimizer and multi-core packages
library(PortfolioAnalytics)
require(xts)
require(DEoptim)
require(doMC)
registerDoMC()
require(TTR)
### Load the data
# Monthly total returns of four asset-class indexes
data(indexes)
#only look at 2000 onward
indexes<-indexes["2000::"]
#'## Review the data
postscript(file="WeightsVsRiskReview.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
# Generate charts to show 12m sma returns and CVaR by asset
charts.BarVaR(indexes[,1:4], p=(1-1/12), clean='boudt', show.cleaned=TRUE, methods=c("ModifiedVaR","ModifiedES"), colorset=rainbow6equal, cex.axis=1)
# Weights v %Contrib to CVaR
table = as.matrix(ES(indexes[,1:4], weights=rep(1/4,4), portfolio_method="component", p=(1-1/12))$pct_contrib_MES)
table = cbind(rep(1/4,4),table)
colnames(table) = c("Weights", "%Contrib to CVaR")
plot(table, ylim=c(-0.1,1), xlim=c(0,1), col=1:4, main="Weight and Contribution to Risk")
text(table[,1],table[,2],rownames(table), pos=4, cex = 0.8, col=1:4)
abline(a=0,b=1, col="darkgray", lty="dotted")
dev.off()
### Create a benchmark using equal-weighted portfolio returns
# Rebalance an equal-weight portfolio quarterly
dates=c(as.Date("1999-12-31"),time(indexes[endpoints(indexes, on="quarters")]))
weights = xts(matrix(rep(1/4,length(dates)*4), ncol=4), order.by=dates)
colnames(weights)= colnames(indexes[,1:4])
EqWgt = Return.rebalancing(indexes[,1:4],weights)
# Chart EqWgt Results
postscript(file="EqWgtPlot1.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.PerformanceSummary(EqWgt, main="Eq Wgt Portfolio", methods=c("ModifiedVaR", "ModifiedES"), p=(1-1/12), clean='boudt', show.cleaned=TRUE, gap=36, colorset=bluefocus, lwd=3)
dev.off()
### Chart the VaR Sensitivity
postscript(file="VaRSense.eps", height=6, width=6, paper="special", horizontal=FALSE, onefile=FALSE)
layout(matrix(c(1,2,3,4), nrow=2))
for(i in 1:4)
chart.VaRSensitivity(indexes[,i], methods=c("ModifiedES", "HistoricalES", "GaussianES"), colorset=bluemono, lwd=c(3,2,2), clean="boudt", main=paste("VaR Sensitivity for", colnames(indexes)[i], sep=" "), cex=.8)
dev.off()
# chart the assets
postscript(file="assetReturns.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.BarVaR(indexes[,1:4], methods=c("ModifiedVaR", "ModifiedES"), colorset=rep("black",4), clean="boudt", show.clean=TRUE)
dev.off()
#'# EXAMPLE 1: Constrained Mean-CVaR Portfolio
### Show an example of a constraint set
aConstraintObj <- constraint(assets = colnames(indexes[,1:4]),
min = .05, # minimum position weight
max = c(.85,.5,.5,.3), #1, # maximum position weight
min_sum=0.99, # minimum sum of weights must be equal to 1-ish
max_sum=1.01, # maximum sum must also be about 1
weight_seq = generatesequence()) # possible weights for random or brute force portfolios
### Add a return objective to a constraint
# Create a small weighted annualized trailing-period mean wrapper function
pamean <- function(n=12, R, weights, geometric=TRUE)
{ sum(Return.annualized(last(R,n), geometric=geometric)*weights) }
# Portfolio annualized exponential moving average monthly return
aConstraintObj <- add.objective(constraints=aConstraintObj,
type="return",
name="pamean",
enabled=TRUE,
multiplier=-1, # to maximize this objective using a minimizer
arguments = list(n=12))
### Add a risk objective to a constraint
aConstraintObj <- add.objective(aConstraintObj,
type="risk", # the kind of objective this is
name="CVaR", # the function to minimize
enabled=TRUE, # enable or disable the objective
arguments=list(p=(1-1/12), clean="boudt")
)
### Use the Random Portfolios engine
# Evaluate the constraint object with Random Portfolios
rndResult<-optimize.portfolio(R=indexes[,1:4],
constraints=aConstraintObj,
optimize_method='random',
search_size=1000, trace=TRUE, verbose=TRUE)
# Chart the results
postscript(file="rpPerformancePlot1.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.RP(rndResult, risk.col="CVaR", return.col="pamean", main="Constrained Mean-CVaR", neighbors=25)
dev.off()
# Chart the weights versus contribution to risk
postscript(file="WeightsVsRisk1.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
table = ES(indexes[,1:4],weights=rndResult$weights, portfolio_method="component", p=(1-1/12))$pct_contrib_MES
table = cbind(rndResult$weights,table)
colnames(table) = c("Weights", "%Contrib to CVaR")
plot(table, ylim=c(0,1), xlim=c(0,1), col=rainbow6equal, main="Weight and Contribution to Risk")
text(table[,1],table[,2],rownames(table), pos=4, cex = 1, col=rainbow6equal)
abline(a=0,b=1, col="darkgray", lty="dotted")
dev.off()
## Evaluate Constrained Mean-CVaR through time
#on one line for easy cut/paste/editing in interactive mode
registerDoMC()
rndResults<-optimize.portfolio.rebalancing(R=indexes[,1:4], constraints=aConstraintObj, optimize_method="random", trace=TRUE, rebalance_on='quarters', trailing_periods=NULL, training_period=36, search_size=1000)
# Chart the cumulative returns
weights1=extractWeights.rebal(rndResults)
Ex1=Return.rebalancing(indexes[,1:4], weights1)
index(weights1)<-as.Date(index(weights1))+1
colnames(Ex1)="Mean CVaR"
results = cbind(EqWgt,Ex1)
postscript(file="ReturnsEx1.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.PerformanceSummary(results[,2:1], main="Constrained Mean-CVaR", colorset=bluefocus[-3], lwd=c(3,2), method=c("ModifiedVaR","ModifiedES"), p=(1-1/12), gap=36)
dev.off()
# Chart the weights and contribution to risk through time
# First, extract weights and calculate ES
numColumns = length(rndResults[[1]]$weights)
numRows = length(rndResults)
contrib1 <- matrix(nrow=numRows, ncol=numColumns)
for(i in 1:numRows){
todate=paste("::",as.Date(names(rndResults)[i]), sep="")
contrib1[i,]=ES(indexes[todate,1:4], portfolio_method="component", weights=rndResults[[i]]$weights, clean="boudt", p=(1-1/12))$pct_contrib_MES
}
colnames(contrib1) = names(unlist(rndResults[[1]]$weights))
contrib1<-xts(contrib1,order.by=index(weights1))
# test
postscript(file="InSampleStats.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
retrisk1 <- matrix(nrow=numRows, ncol=2)
for(i in 1:numRows)
retrisk1[i,] = unlist(rndResults[[i]]$objective_measures)
rownames(retrisk1) = names(rndResults)
colnames(retrisk1) = c("pamean", "CVaR")
chart.TimeSeries(retrisk1, legend="topright", main="In Sample Estimates")
dev.off()
# end test
postscript(file="WeightsContribEx1.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
layout(rbind(1, 2, 3), height = c(3, 3, 1.2), width = 1)
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(weights1, main="Mean-CVaR Weights", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(contrib1, main="Mean-CVaR Risk Contribution", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
plot.new()
par(oma = c(0, 0, 0, 0), mar=c(2,4,4,2)) #c(bottom, left, top, right)
legend("top", legend = colnames(weights1), fill = bluemono[c(-2,-4,-6)], ncol = 4, box.col="darkgray", border.col="white", cex=1)
dev.off()
#'# EXAMPLE 2: Mean-CVaR Risk Limit Portfolio
### Add a risk contribution constraint
# No more than 40% of the risk may be contributed by any one asset
# Reset the position constraints
aConstraintObj$max <-rep(1,4)
names(aConstraintObj$max) <- names(aConstraintObj$min)
aConstraintObj <- add.objective(aConstraintObj, type="risk_budget", name="CVaR", enabled=TRUE, min_prisk=-Inf, max_prisk=.4, arguments = list(clean='boudt', method="modified",p=(1-1/12)))
rndResult2<-optimize.portfolio(R=indexes[,1:4],
constraints=aConstraintObj,
optimize_method='random',
search_size=1000, trace=TRUE, verbose=TRUE)
### Chart the results
postscript(file="rpPerformancePlot2.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.RP(rndResult2, risk.col="CVaR", return.col="pamean", main="Mean-CVaR With Risk Limits", neighbors=25)
dev.off()
# Chart the weights versus contribution to risk
postscript(file="WeightsVsRisk2.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
table = cbind(rndResult2$weights, rndResult2$objective_measures$CVaR$pct_contrib_MES)
colnames(table) = c("Weights", "%Contrib to CVaR")
plot(table, ylim=c(0,1), xlim=c(0,1), col=rainbow6equal, main="Weight and Contribution to Risk")
text(table[,1],table[,2],rownames(table), pos=4, cex = 1, col=rainbow6equal)
abline(a=0,b=1, col="darkgray", lty="dotted")
dev.off()
### Evaluate Mean-CVaR Risk Limit through time
#on one line for easy cut/paste/editing in interactive mode
registerDoMC()
rndResults2<-optimize.portfolio.rebalancing(R=indexes[,1:4], constraints=aConstraintObj, optimize_method="random", trace=TRUE, rebalance_on='quarters', trailing_periods=NULL, training_period=36, search_size=3000)
### Chart the results
# Chart the cumulative returns
weights2=extractWeights.rebal(rndResults2)
Ex2=Return.rebalancing(indexes[,1:4], weights2)
index(weights2)<-as.Date(index(weights2)) + 1
colnames(Ex2) = "Risk Limit"
results = cbind(results, Ex2)
postscript(file="ReturnsEx2.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.PerformanceSummary(results[,3:1], main="Mean-CVaR Risk Limit", colorset=bluefocus[-3], lwd=c(3,2,2), method=c("ModifiedVaR","ModifiedES"), p=(1-1/12), gap=36)
dev.off()
# Chart the weights and contribution to risk through time
numColumns = length(rndResults2[[1]]$weights)
numRows = length(rndResults2)
contrib2 <- matrix(nrow=numRows, ncol=numColumns)
for(i in 1:numRows)
contrib2[i,] = unlist(rndResults2[[i]]$objective_measures$CVaR$pct_contrib_MES)
colnames(contrib2) = names(unlist(rndResults2[[1]]$objective_measures$CVaR$pct_contrib_MES))
contrib2<-xts(contrib2,order.by=index(weights1))
# op <- par(no.readonly = TRUE)
postscript(file="WeightsContribEx2.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
layout(rbind(1, 2, 3), height = c(3, 3, 1.2), width = 1)
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(weights2, main="Risk Limit Weights", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(contrib2, main="Risk Limit Risk Contribution", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
plot.new()
par(oma = c(0, 0, 0, 0), mar=c(2,4,4,2))
legend("top", legend = colnames(weights2), fill = bluemono[c(-2,-4,-6)], ncol = 4, box.col="darkgray", border.col="white", cex=1)
dev.off()
# par(op)
#'# EXAMPLE 3: Equal Risk Portfolio
### Constraints for an Equal risk contribution portfolio
EqRiskConstr <- constraint(assets = colnames(indexes[,1:4]), min = 0.05, max = c(0.85,0.5,0.5,0.3), min_sum=1, max_sum=1, weight_seq = generatesequence())
EqRiskConstr <- add.objective(EqRiskConstr, type="risk_budget", name="CVaR", enabled=TRUE, min_concentration=TRUE, arguments = list(clean='boudt', p=(1-1/12)))
EqRiskConstr <- add.objective(constraints=EqRiskConstr, type="return", name="pamean", enabled=TRUE, multiplier=0, arguments = list(n=12))
### Use DEoptim engine
EqRiskResultDE<-optimize.portfolio(R=indexes[,1:4], constraints=EqRiskConstr, optimize_method='DEoptim', search_size=2000, trace=TRUE, verbose=FALSE) #itermax=55, CR=0.99, F=0.5,
postscript(file="EqRiskDE.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.DE(EqRiskResultDE, return.col="pamean", risk.col="CVaR")
dev.off()
### Evaluate through time
EqRiskResultDERebal<-optimize.portfolio.rebalancing(R=indexes[,1:4],
constraints=EqRiskConstr, optimize_method="DEoptim", trace=FALSE, rebalance_on='quarters', trailing_periods=NULL, training_period=36, itermax=75, CR=0.99, F=0.5, search_size=3000)
### Chart results
# Panel 1: Equal Risk Performance Summary
EqRiskWeights=extractWeights.rebal(EqRiskResultDERebal)
EqRisk=Return.rebalancing(indexes, EqRiskWeights)
index(EqRiskWeights)<-as.Date(index(EqRiskWeights)) + 1
R=cbind(EqRisk,EqWgt)
colnames(R)=c("Equal Risk","Equal Weight")
postscript(file="EqRiskPerf.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.PerformanceSummary(R, methods=c("ModifiedVaR", "HistoricalVaR"), p=(1-1/12), colorset=bluefocus)
dev.off()
# Panel 2: Equal Risk Allocations
postscript(file="EqRiskBars.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
layout(rbind(1, 2, 3), height = c(3, 3, 1.2), width = 1)
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(EqRiskWeights, main="Equal Risk Weights", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
par(mar = c(2, 4, 4, 2) + 0.1)
### @TODO: Make this an extract function or calculate in the optim results
EqRiskPercContrCVaR=matrix(nrow=nrow(EqRiskWeights), ncol=ncol(EqRiskWeights))
for(i in 1:nrow(EqRiskWeights)){
dates = paste(index(indexes)[1], index(EqRiskWeights)[i], sep="::")
EqRiskPercContrCVaR[i,] = ES(indexes[dates,1:4], weights=EqRiskWeights[i,], p=(1-1/12), portfolio_method="component")$pct_contrib_MES
}
colnames(EqRiskPercContrCVaR) = names(EqRiskResultDERebal[[1]]$weights)
EqRiskPercContrCVaR<-xts(EqRiskPercContrCVaR,order.by=index(weights1))
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(EqRiskPercContrCVaR, main="Equal Risk Risk Contribution", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
plot.new()
par(oma = c(0, 0, 0, 0), mar=c(2,4,4,2))
legend("top", legend = colnames(weights1), fill = bluemono[c(-2,-4,-6)], ncol = 4, box.col="darkgray", border.col="white", cex=1)
dev.off()
#'## extended Equal Risk example
#' shart with the prior example
CDDConstr<-EqRiskConstr
#' turn back on the return objective
CDDConstr$objectives[[2]]$multiplier = -1
#' turn off risk_budget objective
CDDConstr$objectives[[1]]$multiplier = 0
#' add CDD objective
CDDConstr <- add.objective(CDDConstr, type="risk", name="CDD", enabled=TRUE, arguments = list(p=(1-1/12)))
### Use DEoptim engine
CDDResultDE2<-optimize.portfolio(R=indexes[,1:4], constraints=CDDConstr, optimize_method='DEoptim', search_size=3000, trace=TRUE, verbose=FALSE, itermax=75, CR=0.99, F=0.5)
postscript(file="CDDDE2.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.DE(CDDResultDE2, return.col="pamean", risk.col="CDD")
dev.off()
### Evaluate through time
CDDResultDERebal2<-optimize.portfolio.rebalancing(R=indexes[,1:4],
constraints=CDDConstr, optimize_method="DEoptim", trace=FALSE, rebalance_on='quarters', trailing_periods=NULL, training_period=36, itermax=75, CR=0.99, F=0.5, search_size=3000)
### Chart results
# Panel 1: CDD Performance Summary
CDDweights=extractWeights.rebal(CDDResultDERebal2)
CDDRet=Return.rebalancing(indexes, CDDweights)
index(CDDweights)<-as.Date(index(CDDweights)) + 1
R=cbind(CDDRet,EqWgt)
colnames(R)=c("CDD","Equal Weight")
postscript(file="CDDPerf.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.PerformanceSummary(R, methods=c("ModifiedVaR", "HistoricalVaR"), p=(1-1/12), colorset=bluefocus)
dev.off()
# Panel 2: CDD Allocations
postscript(file="CDDBars2.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
layout(rbind(1, 2, 3), height = c(3, 3, 1.2), width = 1)
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(CDDweights, main="CDD w/ Return Obj. Weights", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
par(mar = c(2, 4, 4, 2) + 0.1)
### @TODO: Make this an extract function or calculate in the optim results
CDDPercContrCVaR2=matrix(nrow=nrow(CDDweights), ncol=ncol(CDDweights))
for(i in 1:nrow(CDDweights)){
dates = paste(index(indexes)[1], index(CDDweights)[i], sep="::")
CDDPercContrCVaR2[i,] = ES(indexes[dates,1:4], weights=CDDweights[i,], p=(1-1/12), portfolio_method="component")$pct_contrib_MES
}
colnames(CDDPercContrCVaR2) = names(unlist(CDDResultDERebal2[[1]]$weights))
CDDPercContrCVaR2<-xts(CDDPercContrCVaR2,order.by=index(weights1))
par(mar = c(2, 4, 4, 2) + 0.1)
PerformanceAnalytics:::chart.StackedBar.xts(CDDPercContrCVaR2, main="CDD w/ Ret. Obj. Risk Contribution", legend.loc=NULL, cex.axis=1, colorset=bluemono[c(-2,-4,-6)], space=0, border="darkgray")
plot.new()
par(oma = c(0, 0, 0, 0), mar=c(2,4,4,2))
legend("top", legend = colnames(weights1), fill = bluemono[c(-2,-4,-6)], ncol = 4, box.col="darkgray", border.col="white", cex=1)
dev.off()
#'## APPENDIX EXAMPLES? ###
stopifnot(isTRUE(extended_ex),"halting")
## Markowitz-like constrained mean-variance
MeanVarConstr <- constraint(assets = colnames(indexes[,1:4]),
min = 0, max = 1, min_sum=0.99, max_sum=1.01,
weight_seq = generatesequence())
MeanVarConstr <- add.objective(constraints=MeanVarConstr, type="return", name="mean", enabled=TRUE, multiplier=-1, arguments = list())
MeanVarConstr <- add.objective(MeanVarConstr,
type="risk", # the kind of objective this is
name="sd",
enabled=TRUE, # enable or disable the objective
arguments= list())
MeanVarResultRP<-optimize.portfolio(R=indexes[,1:4],
constraints=MeanVarConstr,
optimize_method='random',
search_size=1000,
trace=TRUE, verbose=TRUE)
# Chart the results
postscript(file="mvRP.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
charts.RP(MeanVarResultRP, risk.col="sd", return.col="mean", main="Mean Variance", neighbors=25)
dev.off()
MeanVarResultRPRebal<-optimize.portfolio.rebalancing(R=indexes[,1:4],
constraints=MeanVarConstr,
optimize_method="random",
trace=FALSE,
rebalance_on='quarters',
trailing_periods=NULL,
training_period=36,
search_size=1000)
## Risk budget
RiskBudget.constr <- constraint(assets = colnames(indexes[,1:4]), min = 0, max = 1, min_sum=1, max_sum=1, weight_seq = generatesequence())
RiskBudget.constr <- add.objective(RiskBudget.constr, type="risk_budget", name="CVaR", enabled=TRUE, min_prisk=-Inf, max_prisk=.4, arguments = list(clean='boudt', method="modified",p=.95))
RiskBudget.constr <- add.objective(constraints=RiskBudget.constr, type="risk", name="CVaR", enabled=TRUE, arguments = list(method="modified", portfolio_method="single", enabled=TRUE, p=.95, clean="boudt"))
RiskBudget.optimResult<-optimize.portfolio(R=indexes[,1:4], constraints=RiskBudget.constr,
optimize_method='DEoptim', search_size=1000, verbose=TRUE)
## Equal risk
# First method: minimize risk concentration across assets
EqRiskConstr1 <- constraint(assets = colnames(indexes[,1:4]), min = 0.05, max = c(0.85,0.5,0.5,0.3), min_sum=0.99, max_sum=1.01, weight_seq = generatesequence())
EqRiskConstr1 <- add.objective(EqRiskConstr1, type="risk_budget", name="CVaR", enabled=TRUE, min_concentration=TRUE, arguments = list(clean='boudt', p=(1-1/12)))
EqRiskResultDE1<-optimize.portfolio(R=indexes[,1:4], constraints=EqRiskConstr1, optimize_method='DEoptim', search_size=2000, trace=TRUE, verbose=FALSE) #itermax=55, CR=0.99, F=0.5,
# Second method: force risk levels between limits
EqRiskConstr <- constraint(assets = colnames(indexes[,1:4]), min = 0.05, max = c(0.85,0.5,0.5,0.3), min_sum=1, max_sum=1, weight_seq = generatesequence())
EqRiskConstr <- add.objective(EqRiskConstr, type="risk_budget", name="CVaR", enabled=TRUE, arguments = list(p=(1-1/12), clean="boudt"), min_prisk=0.24, max_prisk=0.26)
EqRiskResultDE<-optimize.portfolio(R=indexes[,1:4], constraints=EqRiskConstr, optimize_method='DEoptim', search_size=2000, trace=TRUE, verbose=FALSE)
#EqRiskResultRP<-optimize.portfolio(R=indexes[,1:4], constraints=EqRiskConstr, optimize_method='random', search_size=1000, trace=TRUE, verbose=TRUE)
# Chart the results
# postscript(file="EqRiskRP.eps", height=6, width=5, paper="special", horizontal=FALSE, onefile=FALSE)
# charts.RP(EqRiskResultRP, risk.col="CVaR", return.col="mean", main="Mean Variance", neighbors=25)
# dev.off()
EqRiskResultDERebal<-optimize.portfolio.rebalancing(R=indexes[,1:4],
constraints=EqRiskConstr, optimize_method="DEoptim", trace=FALSE, rebalance_on='quarters', trailing_periods=NULL, training_period=36, itermax=45, CR=0.99, F=0.5, search_size=1000)
# Panel 1: Equal Risk Performance Summary
EqRiskWeights=extractWeights.rebal(EqRiskResultDERebal)
EqRisk=Return.rebalancing(indexes, EqRiskWeights)
index(EqRiskWeights)<-as.Date(index(EqRiskWeights)) + 1
R=cbind(EqRisk,EqWgt)
colnames(R)=c("Equal Risk","Equal Weight")
charts.PerformanceSummary(R, methods=c("ModifiedVaR", "HistoricalVaR"), p=(1-1/12), colorset=redfocus)
# Panel 2: Equal Risk Allocations
chart.StackedBar(EqRiskWeights)
## @TODO: Make this an extract function or calculate in the optim results
EqRiskPercContrCVaR=matrix(nrow=nrow(EqRiskWeights), ncol=ncol(EqRiskWeights))
for(i in 1:nrow(EqRiskWeights)){
dates = paste(index(indexes)[1], index(EqRiskWeights)[i], sep="::")
EqRiskPercContrCVaR[i,] = ES(indexes[dates,1:4], weights=EqRiskWeights[i,], p=(1-1/12), portfolio_method="component")$pct_contrib_MES
}
colnames(EqRiskPercContrCVaR) = names(unlist(EqRiskResultDERebal[[1]]$weights))
rownames(EqRiskPercContrCVaR) = colnames(weights1)
chart.StackedBar(EqRiskPercContrCVaR)
## Return target with risk budget
## You want to do what?
### ---------------------- Scratch area ---------------------- ###
# Single period DEOptim results
assets.pema=matrix(nrow=1,ncol=4)
# for(i in 1:4){ assets.pema[,i] = paEMA(n=12,indexes[,i],1) }
for(i in 1:4){ assets.pema[,i] = Return.annualized(last(indexes[,i],12),geometric=TRUE) }
colnames(assets.pema)=colnames(indexes[,1:4])
rownames(assets.pema)="Tr 12m Ann Return"
assets.CVaR=ES(indexes[,1:4], invert=FALSE)
assets = rbind(assets.CVaR,assets.pema)
pdf()
plot(t(assets))
points(optimResult$objective_measures[1], optimResult$objective_measures[2])
text(t(assets), colnames(assets))
text(optimResult$objective_measures[1], optimResult$objective_measures[2], "Optimal")
dev.off()
paEMA <- function(n=10, R, weights, ...)
{# call Exponential Moving Average from TTR, return the last observation
sum((12*last(apply(R,2,FUN=TTR::EMA,n=n)))*weights)
}
# xtract = extractStats.rp(rndResult)
# plot(xtract[,6],xtract[,7], xlab="CVaR", ylab="pEMA", col="lightgray", main="Random Portfolios")
# points(xtract[1,6],xtract[1,7], col="orange", pch=16) # equal weighted (seed)
# points(rndResult$constrained_objective[[1]], rndResult$constrained_objective[[2]], col="red", pch=16)
## Use DEoptim as an engine
optimResult<-optimize.portfolio(R=indexes[,1:4],
constraints=aConstraintObj,
optimize_method='DEoptim',
itermax=45, CR=0.99, F=0.5,
search_size=2000 #, verbose=TRUE
)
### Run it several times
optimResultList <- foreach(ii=iter(1:20),# find 20 sol'ns
.errorhandling='pass') %dopar%
optimize.portfolio(R=indexes[,1:4], aConstraintObj,
optimize_method='DEoptim', trace=TRUE,
itermax=25, CR=0.99, F=0.5, search_size=1000)
rndResultList <- foreach(ii=iter(1:20),# 120 sol'ns
.errorhandling='pass') %dopar%
optimize.portfolio(R=indexes[,1:4], aConstraintObj,
optimize_method='random', trace=TRUE,
search_size=1000, verbose=TRUE)
### Evaluate through time
#on one line for easy cut/paste/editing in interactive mode
registerDoMC()
rndResults<-optimize.portfolio.rebalancing(R=indexes[,1:4], constraints=aConstraintObj, optimize_method="random", trace=TRUE, rebalance_on='quarters', trailing_periods=NULL, training_period=36, search_size=1000)
# on multiple, for demo
rndResults<-optimize.portfolio.rebalancing(R=indexes[,1:4],
constraints=aConstraintObj, # our constraints object
optimize_method="random", # allows kitchen sink sol'ns
trace=FALSE, # set verbosity for tracking
rebalance_on='quarters', # any xts 'endpoints'
trailing_periods=NULL, # calculation from inception
training_period=36, # starting period for calculation
search_size=1000) # how many portfolios to test