fitStatisticalFactorModel: Fit statistical factor model using principle components...
In R-Finance/FactorAnalytics: factor analysis

Description Usage Arguments Value Author(s) References Examples

Fit statistical factor model using principle components. This function is mainly adapted from S+FinMetric function mfactor.

1 2	fitStatisticalFactorModel(data, k = 1, refine = TRUE, check = FALSE, max.k = NULL, sig = 0.05, na.rm = FALSE)

`data`	a vector, matrix, data.frame, xts, timeSeries or zoo object with asset returns and factors retunrs names. If data does not have xts class, rownames must provide xts compatible time index.
`k`	numbers of factors if it is scalar or method of choosing optimal number of factors. "bn" represents Bai and Ng (2002) method and "ck" represents Connor and korajczyk (1993) method. Default is k = 1.
`refine`	`TRUE` By default, the APCA fit will use the Connor-Korajczyk refinement.
`check`	check if some variables has identical values. Default is FALSE.
`max.k`	scalar, select the number that maximum number of factors to be considered.
`sig`	significant level when ck method uses.
`na.rm`	if allow missing values. Default is FALSE.

factors T x K the estimated factors.
loadings K x N the asset specific factor loadings beta_i. estimated from regress the asset returns on factors.
alpha 1 x N the estimated intercepts alpha_i
ret.cov N x N asset returns sample variance covariance matrix.
r2 regression r square value from regress the asset returns on factors.
k the number of the facotrs.
eigen eigenvalues from the sample covariance matrix.
residuals T x N matrix of residuals from regression.
asset.ret asset returns
asset.fit List of regression lm class of individual returns on factors.
resid.variance vector of residual variances.
mimic N x K matrix of factor mimicking portfolio returns.

Eric Zivot and Yi-An Chen

Zivot and Wang, (2006) "Modeling Financial Time Series with S-PLUS, 2nd edition"

# load data for fitStatisticalFactorModel.r
# data from finmetric berndt.dat and folio.dat

data(stat.fm.data)
##
# sfm.dat is for pca
# sfm.apca.dat is for apca
class(sfm.dat)
class(sfm.apca.dat)

# pca
args(fitStatisticalFactorModel)
sfm.pca.fit <- fitStatisticalFactorModel(sfm.dat,k=2)
class(sfm.pca.fit)
names(sfm.pca.fit)
sfm.pca.fit$factors
sfm.pca.fit$loadings
sfm.pca.fit$r2
sfm.pca.fit$residuals
sfm.pca.fit$resid.variance
sfm.pca.fit$mimic
# apca
sfm.apca.fit <- fitStatisticalFactorModel(sfm.apca.dat,k=1)
names(sfm.apca.fit)
sfm.apca.res <- sfm.apca.fit$residuals
sfm.apca.mimic <- sfm.apca.fit$mimic
# apca with bai and Ng method
sfm.apca.fit.bn <- fitStatisticalFactorModel(sfm.apca.dat,k="bn")
class(sfm.apca.fit.bn)
names(sfm.apca.fit.bn)
sfm.apca.fit.bn$mimic

# apca with ck method
sfm.apca.fit.ck <- fitStatisticalFactorModel(sfm.apca.dat,k="ck")
class(sfm.apca.fit.ck)
names(sfm.apca.fit.ck)
sfm.apca.fit.ck$mimic