The Reserve Stack Nobody Prices Correctly: How Lenders Layer Taxes, Insurance, CapEx, and Interest Escrows—and Why Your "Proceeds" Aren't Real Leverage

When a lender approves a $50 million transitional loan, the check you receive is rarely $50 million. What arrives is the proceeds net of a reserve stack: holdbacks for property taxes, insurance, capital expenditures, interest reserves, and sometimes tenant improvement allowances or lease-up contingencies. Each reserve layer is mathematically defensible. Collectively, they redefine what leverage actually means.

This isn't new. What's structural is that most sponsors still price their deals around the loan amount, not the net proceeds. The reserve stack is the gap between the number in the term sheet and the money available to execute the business plan. Understanding its architecture is the difference between a feasible underwriting and a refinance trap.

The Escrow Hierarchy: Order Matters More Than You Think

Lenders don't hold reserves randomly. They're sequenced by priority and by when the lender perceives risk. The typical stack runs: interest reserves first, then property tax and insurance reserves, then capital expenditure reserves, then lease-up or operational contingency reserves. Some lenders front-load lease-up reserves; others tier them to performance milestones. The order determines which reserves are drawn first and which sit idle longest, accruing opportunity cost against your equity.

Interest reserves are the most transparent. A 12-month SOFR-plus-spread reserve at 7% all-in on a $50 million loan is roughly $3.5 million held from day one. You don't see it. You can't deploy it. It sits in a lender-controlled account, earning the lender's sweep rate (usually zero) while you pay interest on the full $50 million. That's the structural tax on leverage: you pay interest on capital you don't control.

Property tax and insurance reserves are where sponsors systematically misunderstand the real cost. A lender will hold 12 to 24 months of estimated taxes and insurance. On a $10 million stabilized value asset with a 1.2% effective tax rate and 0.4% insurance cost, that's roughly $160,000 annually. A 24-month reserve is $320,000 held upfront. If the property is partially occupied or in lease-up, the lender may estimate full stabilized taxes and insurance anyway, assuming you'll bear the carry until occupancy justifies the reserve release.

Capital Expenditure Reserves: The Structural Underestimate

CapEx reserves are where the reserve stack becomes a constraint on execution. Lenders typically hold 5 to 15 percent of the total project cost as a replacement reserve, depending on asset age and condition. On a $100 million value-add project with a stated 8% CapEx reserve, that's $8 million. But the lender's definition of "eligible CapEx" often excludes items you've already budgeted: soft costs, permitting delays, code upgrades triggered during construction, or lease commissions to backfill occupancy shortfalls.

The structural problem: your GMP or construction budget assumes a certain scope of work. The lender's CapEx reserve assumes a different scope, often more conservative. When scope creep occurs, as it does in every refurbishment, you're funding the gap from equity or operational cash flow, not from the CapEx reserve. The reserve sits, untouched and unavailable, while you're out of pocket for the actual cost overruns.

This is why sponsors with $50 million in stated proceeds often have only $35 million available to execute. The remaining $15 million is locked in reserves, governed by release conditions you didn't fully price into your underwriting.

Release Conditions: When Reserves Become Covenants

Every reserve has a release mechanism. Property tax and insurance reserves release when the lender's servicer confirms payment from operating cash flow. Interest reserves release when DSCR stabilizes above a threshold, or when the loan refinances. CapEx reserves release upon completion certification, cost reconciliation, and lien waivers. Lease-up reserves release based on occupancy, rent achievement, or tenant quality metrics.

The structural constraint is that release conditions are backward-looking. You must prove performance first, then access the reserve. If your lease-up is slower than underwritten, the lease-up reserve remains trapped. If your CapEx costs exceed budget, you can request a reserve release variance, but the lender will typically demand additional equity contribution or a higher interest rate as compensation. The reserve is collateral against underwriting risk. Its release is conditional, not automatic.

The Compounding Tax on Proceeds

Here's the arithmetic that matters. A $100 million loan with a typical reserve stack might break down as follows: interest reserve ($3.5 million), property tax and insurance reserve ($500,000), CapEx reserve ($8 million), lease-up contingency ($2 million), and miscellaneous holdbacks ($1 million). Total: $15 million. Your net proceeds are $85 million, but you're paying interest on $100 million. Your effective all-in cost of capital is therefore 15 percent higher than the stated interest rate suggests.

Over a three-year hold, that compounds. If the reserve stack earns zero and costs you 7% per annum in interest expense, the cumulative drag is roughly $3.15 million in foregone capital or additional cost. That's not a rounding error. It's the difference between a 2.0x equity multiple and a 1.7x multiple on a typical value-add deal.

The structural reality: lenders price this into the advance rate, but most sponsors don't. They compare their loan-to-value against comparable transactions and assume apples-to-apples leverage. They're comparing stated loan amounts, not net proceeds. The lender knows this gap. It's why they're comfortable with slightly lower spreads on transitional deals than you'd expect: the reserve stack is the real yield.

Sponsor Optionality and the Path Dependency Problem

The reserve stack is also a constraint on your optionality. If you're forced to refinance early because of market conditions or sponsor capital constraints, you're refinancing based on the as-is or as-stabilized value, not on proceeds. But the reserve stack reduces the amount of CapEx and operational investment you've actually deployed. The property may look less improved than the underwriting assumed, which depresses the refinance value and increases the likelihood of a cash-out shortfall.

This is path dependency at work. The reserve stack isn't a one-time cost. It's a structural constraint that echoes through the hold period, limiting your ability to respond to market changes or operational challenges without additional equity. Over decades of cycles, sponsors with disciplined reserve underwriting outperform those who treat reserves as fiction to be released at closing.

What I'm watching next

• Lender reserve release language post-rate cuts: As refinance windows widen, watch whether lenders loosen interest reserve release conditions or demand higher spreads to hold reserves longer. The reserve stack's opportunity cost will shift.
• CapEx reserve disputes in the next market downturn: When lease-up slows or construction costs spike, CapEx reserve release variances become leverage points. Sponsors with detailed cost reconciliation will have more optionality than those with loose contingency budgets.
• The reserve stack in floating-rate refi scenarios: If SOFR rises during the hold, interest reserves grow, and refinance proceeds shrink. Sponsors should model reserve scenarios, not just base-case proceeds.