Usually, the ratio of Het/Hom is expected to be 2:1. As homozygous cases would be either due to mendelian transmission due to inbreeding like effects in a small, finite population or due to some founder effects, so the expected homozygous cases are quite lower. This ratio is, however, true only when the reference panel and the sample you are analysing come from the same population, and the population structure is not admixed. For admixed cases, the ratio obviously lowers than 2:1.
For more: 1. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4551921/ 2. http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1000160 (Beware, this paper used low coverage and hence sampling errors could be present.