Don't believe that I've ever found myself in disagreement with Perry's analyses before, but this DC is far from free.
Minor calculation error aside, there are three problems with the analysis:
1. Inflation was neglected. The mfs (annual dues) are $9000 today. That number will not remain static for one year, much less nine.
2. The cost of money was hand-waved away. This is most unPerry-like, as neglecting this cost runs counter to other valuation models he has described.
3. The costs have already increased as additional properties have been added. The Traveler membership is currently $125000 with $9000 annual fees, up from $100000/$8500 the date of the initial post.
Assuming a static 5% cost of money, the $125000 deposit costs $6250 per year. Further assume 3.5% inflation on the initial annual dues of $9000. The annual cost for the 15 days of use that this membership allows is therefore $15250 in year 1, $15565 in year 2, ... and $18101 in year 9. The total cumulative annual costs in year 1 through year 9 is $149,566, exclusive of the $125,000 deposit.
Assuming you opt out when the membership deposit has increased to $155,000, you will receive $139,500 (90% of the current membership cost). This is a net gain of $139,500 - $125,000 = $14,500, which is less than the annual cost during the first year of ownership.
There is indeed no such thing as a free lunch, or a free DC.
The minor calculation error in the initial post was the assumption that when purchasing at $100,000 and selling at $155,000, that 90% of the $55,000 difference ($49,500) is the net gain to the purchaser. The net gain is actually 90% x 155,000 - 100,000 = $39,500. Considerably more than at the current higher pricing, almost 3 years of annual fees using the modeling summarized above with the $100,000/$8,500 costs valid at the time of the initial post.