How to Build a Full Capital Stack Model for Commercial Real Estate

Key Takeaways

What a Capital Stack Model Does

A capital stack model is the quantitative framework that determines how a commercial real estate acquisition or development is financed. It answers three questions: How much debt can the deal support? What types and amounts of subordinate capital fill the gap between maximum debt and total cost? And what is the residual equity requirement for the sponsor and investors?

The term "capital stack" describes the layered structure of debt and equity that funds a real estate transaction. Each layer has a different priority of repayment, a different risk profile, and a different cost. Senior debt sits at the bottom of the stack with the lowest risk and lowest cost. Common equity sits at the top with the highest risk and highest expected return. Everything in between, including mezzanine debt, preferred equity, and bridge financing, fills the structural gap between what the senior lender will provide and what the deal costs.

The model is not just an allocation exercise. It is a constraint-satisfaction problem. The senior lender's underwriting criteria (LTV, DSCR, debt yield) set the maximum first mortgage. The subordinate lender's attachment-point requirements and total-leverage ceiling set the boundaries of the middle layers. And the residual, whatever is left after all debt and preferred positions are sized, is the common equity check that the sponsor and limited partners must fund.

Building this model correctly matters because every downstream calculation in the pro forma depends on it. Levered cash flows, equity returns, promote waterfalls, and refinancing scenarios all flow from the capital stack. An error in how the stack is constructed cascades through the entire investment analysis. The model also drives the most important early-stage decision in any deal: is this transaction financeable at the return threshold the sponsor needs, or does the capital structure kill it before diligence begins?

The approach mirrors institutional practice: start from the deal, size the senior tranche by its binding constraint, fill the subordinate gap, calculate blended cost of capital, and show how the stack feeds into the levered pro forma. The worked example is a $50M stabilized multifamily acquisition.

Step 1: Define the Deal Inputs

Every capital stack model begins with three deal-level inputs: the purchase price (or total development cost), the net operating income, and the going-in capitalization rate. These inputs anchor the entire model and determine the maximum leverage the deal can support.

Purchase Price

The purchase price is the total acquisition cost, typically inclusive of closing costs, transfer taxes, and any immediate capital reserves the lender requires at closing. For our worked example, the purchase price is $50,000,000 for a 250-unit stabilized multifamily property in a Southeast MSA.

In a development deal, the equivalent input is total development cost (TDC), which includes land, hard costs, soft costs, and reserves. The capital stack construction logic is identical. The only difference is that development deals use projected stabilized NOI rather than in-place NOI for debt sizing, and lenders apply more conservative constraints (lower LTV, higher DSCR thresholds) to reflect the construction and lease-up risk.

Net Operating Income

The NOI is the property's annual net operating income, calculated as effective gross income minus operating expenses. For an acquisition, this is typically the trailing twelve-month (T-12) NOI, adjusted for any one-time items, normalized vacancy, and market-rate expense assumptions. Our worked example uses in-place NOI of $3,000,000, implying a going-in cap rate of 6.00%.

The relationship between the NOI and the debt sizing constraints is direct. Higher NOI supports more debt at any given DSCR or debt yield threshold. Sponsors focus on NOI stabilization before seeking permanent financing because every dollar of incremental NOI supports roughly $9 to $12 of additional senior debt when DSCR is the binding constraint (about $11 in our example), depending on the interest rate, amortization, and coverage threshold. When LTV binds, extra NOI adds debt capacity only to the extent it raises the lender's valuation.

Capitalization Rate

The going-in cap rate is the ratio of NOI to purchase price. It serves as the market's pricing signal for the asset and implicitly defines the relationship between the property's income and its value. In our example, $3,000,000 NOI divided by $50,000,000 purchase price equals a 6.00% cap rate.

The cap rate matters for the capital stack because it directly affects the LTV calculation. A lower cap rate (higher price per dollar of NOI) means the same NOI supports less debt relative to the purchase price. A property trading at a 4.5% cap rate has less debt capacity per dollar of value than the same property at a 6.5% cap rate, all else equal, because the DSCR and debt yield constraints bind sooner relative to the higher price.

As an illustrative assumption, stabilized multifamily cap rates fall roughly in a 5.0% to 6.5% range depending on market tier and vintage; the CBRE Cap Rate Survey H1 2026 tracks current estimates by market and subtype. Our 6.00% example sits in the middle of that range for a secondary-market, value-oriented acquisition.

Step 2: Size the Senior Debt

Senior debt sizing is the most mechanically precise step in the model. The senior lender applies three constraints simultaneously: loan-to-value ratio (LTV), debt service coverage ratio (DSCR), and debt yield (DY). Each constraint produces a maximum loan amount. The actual loan amount is the smallest of the three, because the lender will not exceed any single constraint. This is the MIN() function approach, and it is how every institutional lender sizes a first mortgage.

Constraint 1: Loan-to-Value Ratio (LTV)

The LTV ratio is the loan amount divided by the appraised value (or purchase price, if lower). The lender sets a maximum LTV, typically 60% to 70% for stabilized acquisitions and 55% to 65% for value-add or transitional deals. The maximum loan amount under the LTV constraint is simply the purchase price multiplied by the maximum LTV.

Formula: Max Loan (LTV) = Purchase Price x Maximum LTV

For our $50M example at a 65% LTV cap: Max Loan (LTV) = $50,000,000 x 0.65 = $32,500,000.

The LTV constraint gets complicated when the lender uses a different valuation methodology than the purchase price. Agency lenders (Fannie Mae, Freddie Mac) underwrite to their own appraised value, which may differ from the contract price. CMBS rating agencies apply their own stressed valuations, which can constrain conduit loan size. Life companies typically underwrite to the lower of appraised value or purchase price. The capital stack model should use the lender's underwriting value, not the sponsor's purchase price, as the LTV denominator.

Constraint 2: Debt Service Coverage Ratio (DSCR)

The DSCR is the property's NOI divided by the annual debt service (principal plus interest). The lender sets a minimum DSCR, typically 1.20x to 1.35x for stabilized deals. The maximum loan amount under the DSCR constraint requires working backward from the minimum DSCR to the maximum allowable debt service, then converting that debt service to a loan amount using the loan's interest rate and amortization schedule.

Formula: Max Annual Debt Service = NOI / Minimum DSCR

Formula: Max Loan (DSCR) = PV of annuity at the loan rate, with the max annual debt service as the payment

For our example with $3,000,000 NOI, a 1.25x minimum DSCR, 6.00% interest rate, and 30-year amortization:

Max Annual Debt Service = $3,000,000 / 1.25 = $2,400,000.

Max Monthly Payment = $2,400,000 / 12 = $200,000.

Max Loan (DSCR) = PV(6.00%/12, 360, -$200,000) = $33,358,323. In Excel, this is =PV(0.06/12, 360, -200000), which returns approximately $33.36M.

The DSCR constraint is more mechanically involved than LTV because it depends on the loan's terms (rate and amortization), not just the property's value. A higher interest rate tightens the DSCR constraint by increasing debt service per dollar of loan amount. A shorter amortization (25 years instead of 30) also tightens it by increasing the principal component of each payment. Sponsors care about rate and amortization, not just loan amount, because the terms affect how much debt the NOI can service.

Constraint 3: Debt Yield

Debt yield is the property's NOI divided by the loan amount. It is an unlevered return metric from the lender's perspective, measuring the income the property generates as a percentage of the lender's exposure. As illustrative ranges, minimum debt yield thresholds often fall around 8.0% to 10.0% for CMBS and 7.0% to 9.0% for bank and life company lenders.

Formula: Max Loan (DY) = NOI / Minimum Debt Yield

For our example with $3,000,000 NOI and a 9.0% minimum debt yield: Max Loan (DY) = $3,000,000 / 0.09 = $33,333,333.

Debt yield has become increasingly important since the 2008 financial crisis. Before the crisis, many lenders relied primarily on LTV and DSCR. The problem was that both metrics could be manipulated: LTV depended on the appraisal (which could be inflated), and DSCR depended on the interest rate (which was artificially low in a declining-rate environment). Debt yield removes valuation and interest rate from the calculation, which is why CMBS lenders and rating agencies give it significant weight in sizing conduit loans. It measures the raw income yield on the lender's exposure, independent of property valuation or financing terms, though it still depends on the underwritten NOI.

In the 2026 lending environment, the Federal Reserve's July 2026 Senior Loan Officer Opinion Survey (covering Q2 2026) shows large banks easing standards for all CRE loan types, while other banks left multifamily and construction standards basically unchanged. This bifurcation means the debt yield constraint can bind differently depending on the lender. As an illustrative assumption, a large bank might underwrite to an 8.0% debt yield while a smaller bank holds at 9.5% or higher.

The Binding Constraint: MIN() of Three

The actual senior loan amount is the minimum of the three calculations:

Senior Loan = MIN(Max Loan LTV, Max Loan DSCR, Max Loan DY)

For our example:

• Max Loan (LTV at 65%) = $32,500,000
• Max Loan (DSCR at 1.25x) = $33,358,323
• Max Loan (DY at 9.0%) = $33,333,333

The binding constraint is LTV, producing a maximum senior loan of $32,500,000. The DSCR and debt yield constraints are slack at this loan amount: the actual DSCR at $32.5M would be approximately 1.28x (above the 1.25x minimum), and the actual debt yield would be 9.23% (above the 9.0% minimum).

Identifying which constraint binds matters for two reasons. First, it tells the sponsor what drives their debt capacity. If LTV binds, more debt requires a higher valuation (or a lender willing to go higher on LTV). If DSCR binds, more debt requires higher NOI or a lower interest rate. If debt yield binds, more debt requires higher NOI, period, and rate changes do not help. Second, it tells the model builder where the sensitivity analysis should focus. A 50 bps change in cap rate shifts the LTV constraint significantly but does not touch the DSCR or DY constraints. A 50 bps change in interest rate shifts the DSCR constraint but not the LTV or DY constraints. The binding constraint determines which variable the model is most sensitive to.

Step 3: Fill the Gap with Subordinate Capital

The subordinate capital gap is the difference between the total capitalization (purchase price plus closing costs and reserves) and the maximum senior loan. In our example, the gap is $50,000,000 minus $32,500,000 = $17,500,000. This $17.5M gap can be filled with common equity alone, or it can be split between one or more subordinate capital layers and a smaller equity check.

The decision to add subordinate leverage is a risk/return tradeoff. More leverage amplifies equity returns when the asset earns more than the added capital costs, and amplifies losses in downside scenarios. The capital stack model does not make this decision. It prices the options and shows the sponsor what each structure costs in terms of blended capital cost and equity return impact.

Mezzanine Debt

Mezzanine debt is a loan secured by a pledge of ownership interests in the property-owning entity. It sits between senior debt and equity in the capital stack. The mezz lender's collateral is not the real property itself (that secures the senior mortgage) but the membership interests in the LLC that owns the property. Foreclosure on a mezz default proceeds under UCC Article 9, not through real property foreclosure. Article 9 requires a commercially reasonable sale after reasonable advance notice (notice sent 10 or more days before the sale is timely under section 9-612(b) in non-consumer transactions); in practice, mezz UCC sales commonly close in 30 to 60 days.

Typical mezz sizing fills the capital stack from the top of the senior tranche (65% LTV in our example) to 75–80% of total capitalization. The mezz lender has its own leverage limits: as an illustrative range, mezz lenders commonly cap combined loan-to-value (the mezz detachment point) at 75% to 80%, with tighter caps more likely for non-multifamily assets.

In 2026, mezz pricing on stabilized multifamily runs 11% to 14% for positions attaching at 65% and detaching at 75–80% LTV. Terms are typically 2 to 5 years, interest-only, with a lockout period followed by declining prepayment penalties. Origination fees run 1.0% to 2.0%. The intercreditor agreement between the senior and mezz lenders governs cure rights, standstill periods, and UCC foreclosure timing.

For detailed coverage of mezz mechanics, see Mezzanine Debt & Intercreditor Agreements.

Preferred Equity

Preferred equity is an equity investment that receives a priority return ahead of common equity. Unlike mezzanine debt, preferred equity is not a loan. It is an equity position in the property-owning entity (or a holding company above it) with contractual rights to a preferred return, typically structured as a cumulative, compounding return that must be paid before any distributions to common equity holders.

Preferred equity pricing in 2026 runs 12% to 16%, slightly wider than mezz because the remedies are weaker. A preferred equity holder has no foreclosure right. Its remedies upon default are contractual: removal of the sponsor as manager, forced redemption, litigation. In our experience these remedies can take 6 to 18 months to execute (an illustrative range), compared to the typical 30-to-60-day UCC timeline for mezz. The tradeoff is that preferred equity does not require an intercreditor agreement with the senior lender (because it is equity, not debt), and it is often treated differently from debt in senior loan covenants, though many senior lenders still require approval of preferred equity terms.

Some senior lenders restrict subordinate debt more tightly than preferred equity. In these cases, preferred equity may be the more practical option for subordinate leverage. Agency lenders permit both only under their own rules: the Fannie Mae Multifamily Guide, for example, allows mezzanine financing from approved DUS lender affiliates or with case-by-case approval, subject to a Fannie Mae-approved intercreditor agreement.

Bridge Debt

Bridge loans are short-term, floating-rate loans used to finance transitional assets or acquisitions where the business plan involves significant NOI improvement. Bridge lenders typically provide higher leverage than permanent senior lenders (70–80% of cost or value, depending on the deal profile) but at a higher cost (illustratively SOFR + 250 to 450 bps) and shorter term (2 to 3 years plus extensions).

In a capital stack model, a bridge loan can replace both the senior and subordinate layers with a single, higher-leverage loan. A bridge lender providing 75% of cost on a $50M deal provides $37,500,000 in a single tranche, eliminating the need for a separate mezz or preferred equity layer. The tradeoff is floating-rate exposure and a short maturity that requires refinancing into permanent debt upon stabilization.

Bridge debt is not always a substitute for the senior-plus-subordinate structure. Many deals use a bridge loan as the senior piece and add mezz or preferred equity on top. This "bridge plus mezz" structure can push total leverage to 85% or higher, though it carries significant execution risk if the business plan does not deliver the NOI growth needed to refinance into permanent debt at maturity.

Common Equity

Common equity is the residual. It is whatever is left after all debt and preferred layers are sized. The common equity holders bear the first loss and receive the highest expected return. In a GP/LP structure, the common equity is split between the general partner (sponsor) and limited partners, with the GP earning a promoted interest (carry) above a preferred return hurdle.

The capital stack model calculates the common equity requirement as: Total Capitalization minus Senior Debt minus Subordinate Capital. In our example, if we size $32.5M of senior debt and $5M of mezz at 13%, the common equity requirement is $50M minus $32.5M minus $5M = $12.5M, or 25% of the stack.

The sponsor's goal in constructing the capital stack is to minimize the equity check (to improve return on equity) without creating unacceptable risk in the form of negative cash flow, tight refinancing timelines, or intercreditor complexity. The model quantifies both sides of this tradeoff.

Capital Source Comparison: 2026 Terms

The table below summarizes typical terms for each layer of the capital stack in mid-2026, based on stabilized multifamily transactions. Pricing for other property types (industrial, retail, office, hospitality) varies by risk premium and lender appetite. The Wall Street Prep capital stack framework provides additional context on how institutional lenders evaluate each layer.

Several patterns emerge from this comparison. First, there is a clear monotonic relationship between position in the stack and cost of capital. Senior debt costs 5.50% to 6.75%. Subordinate debt costs 11% to 16%. Common equity targets 15%+ IRR. Each step up the stack (closer to first-loss position) adds cost, with the largest jump between senior debt and the first subordinate layer (about 500 to 700 bps). Second, terms shorten as you move up the stack. Senior permanent debt runs 5 to 12 years. Subordinate capital runs 2 to 5 years. This maturity mismatch creates refinancing risk: the subordinate capital matures before the senior, forcing the sponsor to refinance or repay the subordinate tranche while the senior loan is still outstanding.

Third, amortization disappears in the subordinate layers. Senior debt amortizes over 25 to 30 years, building equity in the property over time. Mezzanine debt is interest-only. Preferred equity accrues and pays at exit. This means the subordinate capital provides leverage without forcing ongoing principal reduction, but it also means the subordinate balance does not decline over time, concentrating refinancing risk at maturity.

Step 4: Calculate Blended Cost of Capital

The blended cost of capital is the weighted average cost across all layers of the stack. It is the CRE equivalent of the corporate finance concept of weighted average cost of capital (WACC). The blended cost tells the sponsor what the deal needs to yield, in aggregate, to satisfy all capital providers.

Formula: Blended Cost = Sum of (Layer Amount x Layer Cost) / Total Capitalization

There is an important nuance here. For debt layers, the "cost" is the coupon rate (the contractual interest rate). For preferred equity, the "cost" is the preferred return rate. For common equity, the "cost" is the target return (IRR or equity multiple), which is not contractual but is the threshold below which the sponsor would not pursue the deal. The blended cost of capital calculation is only fully meaningful when you include the common equity layer's target return.

A simpler and more commonly used version of the calculation includes only the debt and preferred layers (excluding common equity), producing a "blended debt cost" rather than a true WACC. This version answers a more practical question: what is the minimum yield the property must generate to cover all contractual obligations (debt service plus preferred returns)?

Blended Cost Calculation for Our Example

Using the two-layer debt structure from our worked example ($32.5M senior at 6.00% and $5M mezz at 13.0%):

Blended Debt Cost = ($32,500,000 x 0.0600 + $5,000,000 x 0.1300) / $37,500,000

= ($1,950,000 + $650,000) / $37,500,000

= $2,600,000 / $37,500,000

= 6.93%

Compare this to the 6.00% cost with senior debt only. The $5M mezz tranche adds 93 bps to the blended debt cost. That 93 bps is the price of reducing the equity check from $17.5M to $12.5M, a $5M reduction that the sponsor can redeploy into other deals or use to meet the GP's co-investment requirement with less capital at risk.

If we extend the calculation to include the common equity layer's target return (say, 18% IRR on the $12.5M equity check):

Full Blended Cost = ($32,500,000 x 0.0600 + $5,000,000 x 0.1300 + $12,500,000 x 0.1800) / $50,000,000

= ($1,950,000 + $650,000 + $2,250,000) / $50,000,000

= $4,850,000 / $50,000,000

= 9.70%

This means the deal must generate a 9.70% unlevered yield to satisfy all capital providers at their target returns. With a 6.00% going-in cap rate, the deal is "under water" on a static basis. The 370 bps gap between the unlevered yield (6.00%) and the blended cost of capital (9.70%) must be closed by NOI growth and/or cap rate compression at exit. This is a standard dynamic in institutional CRE: deals are financed on the expectation of value creation, not on static current yield.

Worked Example: $50M Multifamily Acquisition

Consider a 250-unit stabilized multifamily property in a Southeast MSA with strong population growth and employment fundamentals. The sponsor is a regional operator with a 15-year track record in the market.

Deal Inputs

Scenario A: Senior Debt Only (65% LTV)

Senior annual debt service: $32,500,000 at 6.00%, 30-year amortization = approximately $2,338,247. DSCR: $3,000,000 / $2,338,247 = 1.28x (above the 1.25x minimum). Cash flow to equity after debt service: $661,753. Cash-on-cash yield: $661,753 / $17,500,000 = 3.8%.

With a 5-year hold, 3% annual NOI growth, and a 6.25% exit cap applied to forward (Year 6) NOI of approximately $3,478,000, the exit value is approximately $55,645,000 (no sale costs assumed). After repaying the remaining senior balance of approximately $30,243,000, equity proceeds are approximately $25,402,000 on an initial equity investment of $17,500,000, plus about $4,236,000 of cumulative cash flow over the five years. The levered equity IRR is approximately 11.9%, and the equity multiple is approximately 1.69x. For reference, the unlevered IRR on the same assumptions is approximately 8.3%.

Scenario B: Senior + Mezzanine (75% Total Leverage)

Combined debt service: Senior $2,338,247 + Mezz $650,000 (IO at 13%) = $2,988,247. DSCR on senior only: 1.28x (unchanged). DSCR on total debt: $3,000,000 / $2,988,247 = 1.00x. Cash flow to equity after all debt service: $11,753 in Year 1. Cash-on-cash yield: 0.1%.

The near-zero Year 1 cash flow is typical for leveraged acquisitions with subordinate debt. The equity return comes primarily from NOI growth and the leveraged exit, not from ongoing cash distributions. With the same 5-year hold, 3% annual NOI growth, and 6.25% exit cap, the mezz must be refinanced or repaid at Year 3 (per its 3-year maturity). Year 1 to 3 cash flow to equity totals only about $0.3M against the $5M balance, so in this example repayment requires new mezz or an equity capital call. If the mezz is refinanced on the same terms and carried to exit, the levered equity IRR is approximately 11.5% (11.6% if equity repays it at Year 3), versus 11.9% in Scenario A. The equity multiple edges up from 1.69x to approximately 1.71x, but the IRR falls.

The reason is negative leverage. The sponsor replaced $5M of equity with $5M of mezz costing 13%, but the asset itself earns an unlevered return of only about 8.3% on these assumptions. Every dollar of mezz costs more than the asset produces, so the shortfall is borne by a smaller equity base. Subordinate debt adds return only when its cost is below the return the asset earns; in a higher-growth or lower-exit-cap case the same tranche could push the equity IRR up. The comparison that matters is the mezz cost versus the asset's return, not the mezz cost versus the sponsor's equity target.

Scenario C: Senior + Preferred Equity (75% Total Leverage)

The preferred equity structure produces a lower equity IRR (approximately 10.8% vs 11.5%) because the pref costs 100 bps more than the mezz (14.0% vs 13.0%) and compounds unpaid for the full hold. But it eliminates the intercreditor agreement entirely, avoids the 3-year mezz maturity refinancing event, and may avoid total leverage restrictions in the senior loan, depending on the loan documents. For deals where the senior lender restricts mezzanine debt, preferred equity may be the more practical structure.

The preferred equity also accrues rather than paying current interest. This means Year 1 cash flow is higher in Scenario C than Scenario B: the equity holders receive $661,753 in cash flow (same as the senior-only scenario) because the preferred return is not paid currently but compounds and is paid at exit. The tradeoff is a larger preferred equity balance at exit (approximately $9,627,000 after 5 years of annual compounding at 14%), which reduces the proceeds available to common equity. The resulting common equity IRR is approximately 10.8% (1.60x multiple), below both the senior-only and mezzanine scenarios.

Excel Mechanics: The MIN() Function Approach

The Excel implementation of the capital stack model follows a systematic structure. The model starts with the deal inputs at the top of the worksheet, runs the three sizing constraints in parallel, selects the binding constraint via the MIN() function, and then allocates the remaining capital to subordinate and equity layers.

Senior Debt Sizing Block

The senior debt sizing block contains three parallel calculations, each producing a maximum loan amount. The actual loan amount is the minimum of the three. Here is the structure, assuming deal inputs are in cells B3 through B9:

The MIN() function is the critical formula. It ensures the model automatically selects whichever constraint produces the smallest (most conservative) loan amount. As you change the deal inputs, the binding constraint may shift. For example, increasing the interest rate to 7.00% would tighten the DSCR constraint enough to make it bind instead of LTV. The MIN() function handles this transition automatically without requiring the model builder to manually identify which constraint is active.

Identifying the Binding Constraint

It is helpful to flag which constraint is binding in the model. A simple conditional formula returns the name of the binding constraint:

=IF(B17=B12,"LTV",IF(B17=B14,"DSCR","Debt Yield"))

This formula checks which of the three maximum loan amounts equals the actual loan amount and returns the corresponding label. In our example, it returns "LTV" because $32,500,000 equals the LTV-constrained maximum.

Subordinate Capital Allocation

The subordinate capital block sizes any mezzanine, preferred equity, or bridge layer. The typical approach sizes the subordinate tranche as the difference between the total leverage target and the senior loan amount, subject to the subordinate lender's maximum combined LTV:

=MIN(B3*B20-B17, B3*B21-B17)

Where B20 is the sponsor's target total leverage (e.g., 75%) and B21 is the subordinate lender's maximum combined LTV (detachment point, e.g., 80%). This formula takes the smaller of: (a) the amount needed to reach the sponsor's target leverage, and (b) the maximum the subordinate lender would provide. If the sponsor targets 75% total leverage on a $50M deal with $32.5M of senior debt, the subordinate tranche is MIN($50M x 0.75 - $32.5M, $50M x 0.80 - $32.5M) = MIN($5.0M, $7.5M) = $5.0M.

Equity Residual

The common equity is the residual:

=B3-B17-B22

Where B22 is the subordinate tranche amount. This formula subtracts the senior loan and subordinate capital from the total capitalization. The result is the equity check: $50,000,000 - $32,500,000 - $5,000,000 = $12,500,000.

Actual DSCR and Debt Yield Checks

After sizing the senior loan by the binding constraint, the model should calculate the actual DSCR and debt yield at the actual loan amount. These serve as check figures to confirm the model is working correctly and to show how much "cushion" exists on the non-binding constraints:

Actual DSCR = B4 / PMT(B5/12, B6*12, -B17) * 12

Actual Debt Yield = B4 / B17

In our example, the actual DSCR is 1.28x (above the 1.25x threshold) and the actual debt yield is 9.23% (above the 9.00% threshold). Both non-binding constraints have positive cushion, confirming that LTV is indeed the binding constraint.

Blended Cost of Capital Formula

The blended cost formula is a SUMPRODUCT divided by the total capitalization:

=SUMPRODUCT(B27:B30, C27:C30) / SUM(B27:B30)

Where B27:B30 contains the layer amounts and C27:C30 contains the corresponding costs (coupon rates for debt, target return for equity). This formula handles any number of layers and updates automatically as you add or remove tranches.

Sensitivity Tables

The capital stack model should include a two-way sensitivity table showing how the equity IRR changes across different combinations of leverage and mezz cost. The standard approach is an Excel data table with total leverage on one axis (70%, 75%, 80%) and mezz coupon on the other (11%, 12%, 13%, 14%, 15%). The output cell is the equity IRR formula. This table reveals the leverage-cost tradeoff in a single view and is the standard deliverable for IC presentations.

A second useful sensitivity table shows how the binding constraint shifts across interest rate and cap rate combinations. With the interest rate on one axis and the cap rate on the other, each cell returns the binding constraint label ("LTV," "DSCR," or "Debt Yield"). This table reveals the regions where different constraints dominate and helps the model builder understand which market moves matter most for debt capacity.

Six Modeling Errors That Break the Stack

Capital stack models fail in predictable ways. These six errors appear repeatedly in deal submissions and analyst work product, and each one distorts the economics enough to change the investment decision.

1. Using purchase price instead of lender's underwriting value for LTV. Agency lenders, CMBS conduits, and life companies underwrite to their own valuation, not the contract price. If the lender's appraised value is $47M but the purchase price is $50M, the LTV constraint should use $47M as the denominator. Using $50M overstates maximum debt by the ratio of the difference. On a 65% LTV cap, the error is ($50M - $47M) x 0.65 = $1.95M of phantom debt capacity. This error is especially common in off-market acquisitions where the purchase price reflects a relationship discount or a motivated seller, and the appraisal comes in below contract.

2. Calculating DSCR on an interest-only basis when the lender underwriting assumes amortization. Many bridge and transitional loans are interest-only during the initial term. But permanent senior lenders (banks, agencies, CMBS) underwrite DSCR on a fully amortizing basis even if the loan includes an IO period. If you calculate the DSCR using the IO payment ($32.5M x 6.00% = $1,950,000) instead of the amortizing payment ($2,338,247), you overstate coverage by 20% and oversize the loan accordingly. The model should always use the amortizing payment for DSCR sizing, even if the loan includes IO years. The IO period is a cash flow benefit, not a coverage benefit from the lender's perspective.

3. Ignoring the mezz maturity mismatch. Mezzanine terms are often shorter than the senior loan. If the senior loan has a 10-year term and the mezz has a 3-year term, the mezz matures in Year 3, forcing a refinancing or repayment event. Many models treat the mezz as if it runs through the full hold period, which understates the refinancing risk and overstates the equity cash flows in Year 3. The correct approach is to model the mezz payoff at its contractual maturity (either through a refinancing into new mezz, conversion to equity, or payoff from cash flow or a capital call). Each exit path has different economics and should be modeled explicitly.

4. Double-counting the total DSCR as a constraint. The total DSCR (NOI divided by total debt service including mezz) often drops below 1.0x in leveraged deals. This is expected. The senior lender cares about the senior DSCR (NOI divided by senior debt service only), not the total DSCR. The mezz lender accepts the subordinate position precisely because it expects the total debt service to consume most or all of the current NOI, with the equity return coming from NOI growth and the exit. Treating total DSCR below 1.0x as a "problem" or using it as a constraint will kill every deal that includes subordinate leverage.

5. Omitting origination fees and closing costs from the equity requirement. The capital stack model should size the equity requirement as: Total Capitalization (including closing costs) minus Total Debt Proceeds (net of origination fees). A $5M mezz loan with a 1.5% origination fee delivers $4,925,000 in net proceeds, not $5,000,000. The $75,000 fee comes out of the equity check. Similarly, the senior loan's origination fee (typically 0.50% to 1.00%) reduces net proceeds and increases the equity requirement. These fees add up: on a $50M deal with $32.5M of senior debt and $5M of mezz, total origination fees can approach $425,000, a material increase to the equity check.

6. Modeling the exit without adjusting for remaining loan balances at the hold-period end. The exit proceeds calculation is: Sale Price minus Remaining Senior Loan Balance minus Remaining Subordinate Balance minus Disposition Costs. A common error is using the original loan amount rather than the remaining balance after amortization. On a $32.5M loan amortizing over 30 years at 6.00%, the remaining balance at Year 5 is approximately $30.2M, not $32.5M. The $2.3M of amortization that has been paid down over 5 years is equity that has been built into the property and should flow to the equity holders at exit. Using the original loan amount understates equity proceeds by this amortization amount and, in this example, cuts the calculated IRR by roughly 180 bps.

Model It in Apers

Related Articles

• Mezzanine Debt & Intercreditor Agreements. The institutional guide to mezz structure, UCC foreclosure, and the intercreditor provisions that govern the relationship between senior and subordinate lenders. Covers 2026 pricing and a worked $40M example.
• Preferred Equity: Priority of Payments and Redemption. How preferred equity structures its payment waterfall, redemption mechanics, and sponsor governance rights. The equity-side alternative to mezzanine debt in the subordinate layer.
• Mezzanine Debt vs Preferred Equity: Structural Differences. The full side-by-side comparison across eight dimensions, including legal form, collateral, remedies, timeline, tax treatment, and lender consent requirements.
• Bridge Loans, Floating Rate Risk, and Exit Assumptions. How bridge debt fits into the capital stack as either a standalone senior tranche or a complement to subordinate leverage. Covers SOFR-based pricing and extension mechanics.
• LP/GP Structures, Promote, Catch-Up, and Clawback. The equity waterfall that sits above the debt layers. How promote economics change when subordinate leverage compresses the equity check.
• CMBS: Conduit vs SASB, Defeasance, and Yield Maintenance. How CMBS sizing constraints (particularly debt yield) differ from bank and agency approaches, and how these differences affect capital stack construction.