#*****************************************************************
# Load historical data
#*****************************************************************
library(SIT)
load.packages('quantmod')

# load saved Proxies Raw Data, data.proxy.raw, to extend DBC and SHY
# please see http://systematicinvestor.github.io/Data-Proxy/ for more details
load('data/data.proxy.raw.Rdata')

tickers = '
LQD + VWESX
DBC + CRB
VTI +VTSMX # (or SPY)
ICF + VGSIX # (or IYR)
CASH = SHY + TB3Y
'

data <- new.env()
getSymbols.extra(tickers, src = 'yahoo', from = '1970-01-01', env = data, raw.data = data.proxy.raw, set.symbolnames = T, auto.assign = T)
for(i in data$symbolnames) data[[i]] = adjustOHLC(data[[i]], use.Adjusted=T)

#print(bt.start.dates(data))
bt.prep(data, align='remove.na', fill.gaps = T)

# Check data
plota.matplot(scale.one(data$prices),main='Asset Perfromance')

#*****************************************************************
# Setup
#*****************************************************************
data$universe = data$prices > 0
  # do not allocate to CASH, or BENCH
  data$universe$CASH = NA

prices = data$prices * data$universe
  n = ncol(prices)
  nperiods = nrow(prices)


frequency = 'months'
# find period ends, can be 'weeks', 'months', 'quarters', 'years'
period.ends = endpoints(prices, frequency)
  period.ends = period.ends[period.ends > 0]

models = list()

commission = list(cps = 0.01, fixed = 10.0, percentage = 0.0)

# lag prices by 1 day
#prices = mlag(prices)

#*****************************************************************
# Equal Weight each re-balancing period
#******************************************************************
data$weight[] = NA
  data$weight[period.ends,] = ntop(prices[period.ends,], n)
models$ew = bt.run.share(data, clean.signal=F, commission = commission, trade.summary=T, silent=T)

#*****************************************************************
# Risk Parity each re-balancing period
#******************************************************************
ret = diff(log(prices))
hist.vol = bt.apply.matrix(ret, runSD, n = 20)

# risk-parity
weight = 1 / hist.vol
rp.weight = weight / rowSums(weight, na.rm=T)

data$weight[] = NA
  data$weight[period.ends,] = rp.weight[period.ends,]
models$rp = bt.run.share(data, clean.signal=F, commission = commission, trade.summary=T, silent=T)

#*****************************************************************
# Strategy:
#
# 1) Use 60,120,180, 252-day percentile channels
# - corresponding to 3,6,9 and 12 months in the momentum literature- 
# (4 separate systems) with a .75 long entry and .25 exit threshold with 
# long triggered above .75 and holding through until exiting below .25 
# (just like in the previous post) - no shorts!!!
#
# 2) If the indicator shows that you should be in cash, hold SHY
#
# 3) Use 20-day historical volatility for risk parity position-sizing 
# among active assets (no leverage is used). This is 1/volatility (asset A) 
# divided by the sum of 1/volatility for all assets to determine the position size.
#******************************************************************
allocation = 0 * ifna(prices, 0)
for(lockback.len in c(60,120,180, 252)) {
  high.channel = bt.apply.matrix(data$prices, runQuantile, lockback.len, probs=0.75)
  low.channel = bt.apply.matrix(data$prices, runQuantile, lockback.len, probs=0.25)
  signal = iif(cross.up(prices, high.channel), 1, iif(cross.dn(prices, low.channel), -1, NA))
  allocation = allocation + ifna( bt.apply.matrix(signal, ifna.prev), 0) 
}

# (A) Channel score
allocation = ifna(allocation / 4, 0)

# equal-weight
weight = abs(allocation) / rowSums(abs(allocation))
	weight[allocation < 0] = 0
weight$CASH = 1 - rowSums(weight, na.rm=T)

data$weight[] = NA
  data$weight[period.ends,] = weight[period.ends,]
models$channel.ew = bt.run.share(data, clean.signal=F, commission = commission, trade.summary=T, silent=T)



# risk-parity: (C)
weight = allocation * 1 / hist.vol
weight = abs(weight) / rowSums(abs(weight), na.rm=T)
	weight[allocation < 0] = 0
weight$CASH = 1 - rowSums(weight, na.rm=T)

data$weight[] = NA
  data$weight[period.ends,] = ifna(weight[period.ends,], 0)
models$channel.rp = bt.run.share(data, clean.signal=F, commission = commission, trade.summary=T, silent=T)


# let's verify
last.period = last(period.ends)
print(allocation[last.period,])

print(to.percent(last(hist.vol[last.period,])))

print(to.percent(last(weight[last.period,])))

#*****************************************************************
#The [Quantitative Approach To Tactical Asset Allocation Strategy(QATAA) by Mebane T. Faber](http://mebfaber.com/timing-model/)
#[SSRN paper](http://papers.ssrn.com/sol3/papers.cfm?abstract_id=962461)
#******************************************************************
# compute 10 month moving average
sma = bt.apply.matrix(prices, SMA, 200)

# go to cash if prices falls below 10 month moving average
go2cash = prices < sma
  go2cash = ifna(go2cash, T)

# equal weight target allocation
target.allocation = ntop(prices,n)

# If asset is above it's 10 month moving average it gets allocation
weight = iif(go2cash, 0, target.allocation)

# otherwise, it's weight is allocated to cash
weight$CASH = 1 - rowSums(weight)

data$weight[] = NA
  data$weight[period.ends,] = weight[period.ends,]
models$QATAA = bt.run.share(data, clean.signal=F, commission = commission, trade.summary=T, silent=T)



#*****************************************************************
# Report
#*****************************************************************
#strategy.performance.snapshoot(models, T)
plotbt(models, plotX = T, log = 'y', LeftMargin = 3, main = NULL)
	mtext('Cumulative Performance', side = 2, line = 1)

print(plotbt.strategy.sidebyside(models, make.plot=F, return.table=T))

for(m in names(models)) {
  print('#', m, 'strategy:')
  plotbt.transition.map(models[[m]]$weight, name=m)
    legend('topright', legend = m, bty = 'n')
                
  print(plotbt.monthly.table(models[[m]]$equity, make.plot = F))
   
  print(to.percent(last(models[[m]]$weight)))
}

#  plotbt(models[3], xfun = function(x) { 100 * compute.drawdown(x$equity) })

models.2010 = bt.trim(models, dates = '2010::')

plotbt(models.2010, plotX = T, log = 'y', LeftMargin = 3, main = NULL)
	mtext('Cumulative Performance', side = 2, line = 1)

print(plotbt.strategy.sidebyside(models.2010, make.plot=F, return.table=T))