Question

I am trying to find months in S&P 500 which yielded more than x%.
The following code find such instances:

getSymbols('^GSPC', from='2010-01-01')
G <- monthlyReturn(Cl(GSPC))
names(G) <- 'RET.M'
G[G$RET.M>.05]

Now, I would like to find out what is the yearly return of such instances?
I know I can calculate the yearly return as follow:

G$RET.Y <- yearlyReturn(Cl(GSPC))

But now, I want to query G$RET.Y to get the yearly return for months that yielded 5% or more.

Any suggestions?

Answer 1

You can merge them, fill backwards using na.locf with fromLast=TRUE, then subset as normal

G$RET.Y <- yearlyReturn(Cl(GSPC))
G$RET.Y <- na.locf(G$RET.Y, fromLast=TRUE)
G$RET.Y[G$RET.M > 0.05]
#                   RET.Y
#2010-03-31  0.1100186233
#2010-07-30  0.1100186233
#2010-09-30  0.1100186233
#2010-12-31  0.1100186233
#2011-10-31 -0.0000318056
#2013-01-31  0.0643252302

Or, you could format the index as a character representation of the year — format(index(G[G$RET.M > 0.05]), "%Y") — then just subset the yearly data with that

do.call(rbind, lapply(format(index(G[G$RET.M > 0.05]), "%Y"), function(x) {
  yearlyReturn(Cl(GSPC))[x]
}))
#           yearly.returns
#2010-12-31   0.1100186233
#2010-12-31   0.1100186233
#2010-12-31   0.1100186233
#2010-12-31   0.1100186233
#2011-12-30  -0.0000318056
#2013-02-08   0.0643252302