$4.20
$5.80
bear $4.20
base $5.05
bull $5.80
margin of safety $4.10
market price* $3.43
Business condition attached to the margin of safety
The wind-down continues on its current track: no further deterioration in realized sale prices below the roughly 8% premium-to-book seen on dispositions completed in the first half of 2026, and no sign that the Dividend Trust Portfolio joint-venture partner is blocking or materially delaying use of the joint venture's buy/sell provision to monetize that stake.
SITE Centers today is a small, self-liquidating real-estate company, not a growing rental business. Since spinning off its convenience-retail properties into Curbline Properties in October 2024, SITE Centers has been selling its remaining shopping centers and returning the proceeds to shareholders, with management stating its intent to eventually dissolve the company. As of mid-2026, only eight wholly-owned shopping centers remain — in Long Beach, California; Parker, Colorado; Fort Walton Beach, Florida; two in Chicago, Illinois; two in Chapel Hill, North Carolina; and Portland, Oregon — plus a 20% stake in a ten-property joint venture (the Dividend Trust Portfolio). The company itself is an Ohio corporation with all of its own properties in the United States. The company has been very good at this so far. During 2025 it sold shopping centers including Winter Garden Village in Florida ($165 million), Nassau Park Pavilion in Princeton, New Jersey ($137.6 million), and The Promenade at Brentwood in Missouri ($71.6 million), among others, for combined proceeds of $752 million — a gain of roughly $320 million over book value, on top of a $633 million gain the year before. Alongside these sales, the company paid out $227.7 million of special dividends in 2025 and declared a further $1.00-per-share dividend (about $53 million) in the second quarter of 2026. The evidence has shifted, though, in a way that matters for what's left. In the first half of 2026, the company sold roughly $146 million of remaining real estate for a gain of only about $12 million — a premium of roughly 8%, far below the pattern from 2025. The company's own 2025 impairment charges of $114 million were explicitly tied to "changes in hold period assumptions" — in plain terms, marking properties down as their expected sale dates got closer. Occupancy at the wholly-owned properties also slipped from 90.6% to 83.7% over the same period. Taken together, this looks like the easiest, highest-value properties were sold first, and what remains commands a smaller premium over book value than the 2025 track record would suggest. We weighted this recent, dated evidence more heavily than the older full-year pattern in setting the base and bull cases, rather than assuming the strong 2025 premiums simply continue. We valued the shares as: today's book value per share (cash, the eight remaining properties at net book value, the joint-venture stake, less liabilities) plus a premium if the remaining properties sell above book, plus any premium on eventually monetizing the joint-venture stake — which has a contractual "buy/sell" mechanism that can force a resolution — minus the ongoing corporate costs (currently running near $36 million a year) and one-time wind-up expenses (severance, contract terminations, professional fees) the company has flagged it will incur before it can close down, all discounted at 6% over the one-to-two years we expect this process to take. As of June 30, 2026, book value already reflects roughly $53 million distributed to shareholders since year-end, and equity comes to about $5.38 per share. Our base case assumes the remaining properties sell at a moderate premium (around 25%, between the compressed first-half-2026 result and the stronger 2025 average), the joint-venture stake is realized at close to its current carrying value, and remaining corporate overhead consumes a modest net amount before the process wraps up in roughly a year — landing at about $5.05 per share. Our bear case assumes sales continue near the weak first-half-2026 pace, the joint venture is harder to unwind given the partner's consent rights, and the wind-up runs closer to two years with higher costs — about $4.20 per share. Our bull case assumes the remaining, more desirable properties (several are in larger coastal metro areas) recapture something closer to the historical premium, the joint venture is monetized on favorable terms, and the process concludes within about nine months — about $5.80 per share. What would have to be true for this to be wrong: if the compressed first-half-2026 sale premiums are simply the new normal rather than a temporary mix effect, or if corporate costs don't decline as the company shrinks, actual value could come in below even our bear case. Conversely, if the remaining eight properties — which include desirable West Coast and Sun Belt locations — command premiums closer to the stronger 2025 pattern, the bull case understates the outcome. Two items we could not fully pin down from the filings: management has not disclosed a specific dollar estimate for total remaining wind-up costs, and there is no disclosed timeline or expected price for monetizing the joint-venture stake. Both are treated as estimates built from the best evidence available, not disclosed facts.