Is it pretty common to have "reverse complemented alleles" when lifting over a vcf from hg19 to hg38?

Question

[curious]

I just lifted over a VCF from hg19 to hg38 using LiftoverVcf from GATK.

I noticed a lot of the alleles have the field ReverseComplementedAlleles in the lifted VCF I create

For example, one record of hg19 is chrom:pos:ref:alt this:

21:11019500:T:C

but lifts over to this:

chr21:10492957:A:G

if I use UCSC's genome browser and search that position on hg 19 and hg38, the ref allele change seems to check out: https://genome.ucsc.edu/cgi-bin/hgGateway

Is this a pretty normal phenomenon during liftover?

Looks like it happened to about 246,033 of 247,915 of my variants