percent; Detroit, 10 percent; Seattle, 10 percent; Chicago, 6 percent; and Los Angeles, 3

percent.

 Approximately three percent of all retail sales represent sale-leaseback transactions.

This yields 340 sale-leaseback transactions in the retail sample, which is similar to the

number of such transactions observed with office sales (342). Because of the small

number of sale-leaseback transactions and their relatively equal distribution across

markets there are only two markets where I can expect to observe results that are

statistically meaningful. These markets are Chicago, where a total of 56 sale-leasebacks

are observed and Phoenix, which has 50 such transactions.

 Finally, portfolio sales represent on average 3 percent of all retail sales. This yields

a sample of 347 portfolio transactions. In contrast there were only 222 portfolio sales in

the office sample. The market with the largest percentage of portfolio sales is Tucson,

where these transactions represent 10 percent of all observations. Another market with a

relatively high percentage of bulk purchases is Dallas / Forth Worth, where portfolio

sales represent 8 percent of all sales. Finally, there are four additional markets with larger

numbers of portfolio sales: Denver (33 observations), Detroit (30 observations), Los

Angeles (40 observations) and Seattle (37 observations).

 Table 19 presents t-tests for differences between the mean price of retail properties

in a comparison group composed of all properties not associated with any conditions or

atypical motivation, and the groups of properties that are the subject of interest. The

EXREPL group represents properties that are part of replacement exchange only and that

are not associated with other conditions of sale. The EXRELQ group represents

relinquished exchanges only; RELQ REPL is a group of properties that were part of two