“…The problem arises because the number of assets that can be used in truly optimized optimal portfolio analysis should not exceed the limit of N-1 where N is the number of scenarios. For example, Ando et al (2018) conducted optimal portfolio analysis by creating 24 subunits for the U.S. portion of the Prairie Pothole Region, 7 subregions for birds of the Eastern U.S., and 11 subregions for Southern Appalachian salamanders across 71, 8, and 12 climate scenarios, respectively, where the number of subunits (or assets) is smaller than the number of scenarios (or observations per subunit).…”