A hyperscaler lease with a termination-for-convenience clause at year five reads like a fifteen-year hold until you find that clause on page 34.

This thread is well above beginner territory, so I want to flag upfront that I'm explaining how these structures work rather than advising on any specific deal or negotiation. Anyone analyzing actual LP documents at this level should have a real estate attorney read the lease and the operating agreement together.

For anyone following along who is newer to the vocabulary: a hyperscaler is one of the very large cloud and technology companies that leases entire data center facilities. A GP (general partner) is the operator who runs the deal and earns a promoted return, meaning a larger share of profits, once certain targets are met. An LP (limited partner) is the passive investor whose capital funds the deal.

The framing in this post is precise and worth understanding. A weighted average lease term is an average across all leases in a portfolio, weighted by rent. When a single tenant dominates the building and holds a termination right, that average can misrepresent real duration badly. The poster is right that an unconditional termination right gives the tenant a free option, and the GP's promote structure, typically built around a long hold and a sale at a stabilized cap rate, does not adjust automatically when that option exists.

On mitigation at term sheet stage: the levers I have seen discussed in deal structuring literature include termination fees sized to cover re-leasing costs and carry, a cure or performance threshold tied to uptime or capacity utilization before the right activates, and a right of first offer to re-lease before the tenant can exit. Each one costs the landlord something, either a lower headline rent, a longer free-rent period, or a more complex negotiation.

I am genuinely uncertain how often the unconditional version gets renegotiated versus accepted as market standard for hyperscaler credit. That would be worth asking someone with current leasing counsel experience in this sector specifically.

What is the structure of the deal you are reviewing, a direct LP interest or a fund-of-funds layer on top?

Recapture rights traded for a rent step that accelerates in years three through five sometimes buy the clause out structurally.

The mechanic worth adding: if the landlord can negotiate a termination fee sized to cover the fit-out gap, eighteen months of carry at the replacement tenant's assumed rent, plus the stranded capex on hyperscaler-specific infrastructure, the unconditional right becomes economically conditional even if it stays unconditional on paper. What the landlord typically gives up is the rent bump in the back half of the term, because the hyperscaler's treasury models the fee as a liability and prices it into the headline rent ask. So the trade is yield compression in years six through fifteen against a floor on the residual. Whether that trade makes sense depends entirely on what the GP's promoted return looks like if you strip years six through fifteen down to the termination fee recovery and nothing else. What does the waterfall look like if the fee replaces the terminal cap rate assumption?

The mitigation I have seen work at term sheet stage is a tiered liquidated damages schedule tied to the year of termination, where the penalty at year five is large enough to approximate the residual value gap the landlord would face remarketing a vacant shell. The hyperscaler accepts it because the clause still exists and their optionality is preserved, but the landlord converts an unconditional exit into an unconditional exit with a price, which reframes the promote math entirely. What the landlord typically gives up to get there is the rent step structure in years six through ten, essentially trading future upside that only materializes if the tenant stays for a present certainty that pays out if they leave. The tradeoff favors the liquidated damages structure when the landlord's basis is high relative to secondary market shell values, because the penalty floor substitutes for a residual value that the market may not support.

The assumption doing the most work in the original deal structure is that eighteen-month remarketing period, because if secondary demand has compressed to six months by year five, the clause is far less damaging than the cover deck implies, and if it has extended to thirty months, the damage is worse than modeled. What does the GP's sensitivity table show for vacancy duration, and does it stress the scenario where the replacement tenant requires a full power infrastructure refresh rather than a simple fit-out?