predict.ffm: Predicts asset returns based on a fitted fundamental factor...
In factorAnalytics: Factor Analytics

Description Usage Arguments Details Value Author(s) See Also Examples

S3 predict method for object of class ffm.

1 2	## S3 method for class 'ffm' predict(object, newdata = NULL, pred.date = NULL, ...)

`object`	an object of class `ffm` produced by `fitFfm`.
`newdata`	data.frame containing the variables `asset.var`, `date.var` and the same exact `exposure.vars` used in the fitted `ffm` object. If omitted, the predictions are based on the data used for the fit.
`pred.date`	character; unique date used to base the predictions. Should be coercible to class `Date` and match one of the dates in the data used in the fiited `object`.
`...`	optional arguments passed to `predict.lm` or `predict.lmRob`.

The estimated factor returns and potentially new factor exposures are used to predict the asset returns during all dates from the fitted ffm object. For predictions based on estimated factor returns from a specific period use the pred.date argument.

predict.ffm produces a N x T matrix of predicted asset returns, where T is the number of time periods and N is the number of assets. T=1 if pred.date is specified.

Sangeetha Srinivasan

fitFfm, summary.ffm, predict.lm, predict.lmRob

# Load fundamental and return data
data(Stock.df)

# fit a fundamental factor model
exposure.vars <- c("BOOK2MARKET", "LOG.MARKETCAP")
fit <- fitFfm(data=stock, asset.var="TICKER", ret.var="RETURN", 
              date.var="DATE", exposure.vars=exposure.vars)

# generate random data
newdata <- as.data.frame(unique(stock$TICKER))
newdata$BOOK2MARKET <- rnorm(nrow(newdata))
newdata$LOG.MARKETCAP <- rnorm(nrow(newdata))
pred.fund <- predict(fit, newdata)