I have been reading a bit about it, and AFAIK there is no real answer to your question. Transposon are very repetitive and usually a pain to map, and quantify. A few notes:

Mapping

1. If allowing multimapping reads do not use bowtie -a as this will artificially increased the reads mapping to repeat regions. Most published studies will use -m 1, retain only uniquely mapping reads, or -M 1 which chooses randomly a location for a multimmaping read (assuming equal Mapping qualities).
2. There is a new tool, ShortStack that estimates read density using uniquely mapping reads and then distributes multi-mappers accordingly. This will retain more reads comparatively to -m 1 and will be more accurate than -M 1. Looks good but untested for ChIP-seq. I would like to test it, but not time yet.

Regardless, when dealing with repeat regions, be very careful:

http://www.ncbi.nlm.nih.gov/pubmed/27046835

http://www.ncbi.nlm.nih.gov/pubmed/25805138

Quantification

If you mean having RNA-seq data wanting to know the expression of retrotransopsons, some people just intersect the reads with the annotation, for instance repeat masker track, and summarize counts per class or family of retrotransposon. Due to the mapping issues, quantifying individual genomic elements is tricky, but ShortStack also integrates a counting module. There are tools out there to quantify expression of Transposons, but sadly my paper database is not as organised as I thought, and I don't remember the names of these tools.

I normally use Repeat Enrichment Estimator. I downloaded the source code about a year ago and it's working fine for me.

I'd also like to hear about better tools if anyone has ideas.