Interview Questions

They are cousins, but NOI is property-level and EBITDA is company-level. Both strip out interest, taxes, and depreciation to isolate operating performance, but NOI measures a single asset's income after property operating expenses, while EBITDA measures an operating company's earnings after all its costs including corporate overhead. NOI excludes corporate G&A and capital items and is what you capitalize at a cap rate to value a building; EBITDA is what you apply a multiple to in order to value a business. In REIT land you see both: the assets are valued on NOI and cap rates, while the company is looked at on EV/EBITDA or, more often, P/FFO.

Because a single property throws off NOI, not corporate EBITDA, and the market prices real estate in its own native conventions. Cap rate (NOI over value) and price per square foot, per unit, or per key let you compare one building directly against another and against recent sales. EV/EBITDA is built for operating companies with corporate earnings and overhead; for a standalone asset it adds nothing you cannot read more cleanly from the cap rate. At the entity level it flips: you do value a whole REIT on multiples like P/FFO or EV/EBITDA.

A real estate model is cash-flow-first and asset-level, whereas a corporate three-statement model integrates an income statement, balance sheet, and cash flow statement for a whole company. In real estate you build a property-level cash flow, NOI, then debt service, capex, and the equity distributions, plus a sale at exit, often with an equity waterfall, over a defined hold period. You generally do not build a full balance sheet, working-capital schedules, or corporate accruals; there is no inventory, receivables cycle, or consolidated entity to tie out. The focus is the cash a single asset throws off to its debt and equity over time, which is why the waterfall and the reversion, not a balance sheet, are the heart of the model.

The core types and what drives each: multifamily, driven by jobs and household formation, with short leases that reprice quickly; office, driven by employment and now work-from-home, with long, TI-heavy leases; industrial and logistics, driven by e-commerce and supply chains, with low capex; retail, driven by consumer spending and location and pressured by e-commerce; and hospitality, driven by travel demand with nightly repricing, the most cyclical. Beyond the traditional five sit data centers, driven by cloud and AI power demand, and net lease, driven by tenant credit on long leases. Each has its own demand drivers, lease length, and capex profile, which is what shapes its cap rate and risk.

You build it from the top line down to equity cash flow. Start with potential gross income, every unit at market rent, then subtract vacancy and credit loss to get effective gross income. Subtract operating expenses to get NOI. Below NOI, subtract capex, tenant improvements, and leasing commissions, and subtract debt service, to get levered cash flow to equity. At the end of the hold, add a reversion: the sale priced at an exit cap, net of selling costs and the loan payoff. Then you run IRR, equity multiple, and cash-on-cash off those equity cash flows over the hold.

All else equal, rising rates push cap rates up and values down, because higher rates raise the cost of debt and lift the returns investors require, so they pay less per dollar of NOI. But the link is not mechanical: a cap rate is really a spread over Treasuries, so if that spread compresses, if NOI is growing quickly, or if there is heavy capital chasing deals, cap rates can hold or even fall while rates rise. That gap between rates and cap rates is one of the first things I would look at in the current market.

The going-in cap is year-one NOI over the purchase price, your entry yield. The exit (going-out) cap is the forward NOI over the sale price you assume at disposition, used to set the terminal value. Underwriters almost always assume the exit cap is equal to or higher than the going-in cap, usually by 25 to 50 bps, to stay conservative: the asset is older at sale and you cannot control where rates and cap rates sit years out. Assuming cap-rate compression at exit is how you flatter a deal, so reviewers push on that assumption hard.

Yield on cost is stabilized NOI divided by total project cost, so a development that costs $80M all-in and stabilizes at $5.6M of NOI has a 7% yield on cost. It is the unlevered return a developer earns on what it actually cost to build, as opposed to the cap rate, which is the yield the market charges on price. The reason to build rather than buy is to earn a yield on cost above the cap rate you could buy at, and that gap is the development spread.

The development spread is the yield on cost minus the market (exit) cap rate, and developers typically want roughly 150 to 200 bps. It is the compensation for taking construction, lease-up, and timing risk that a buyer of a finished, stabilized asset does not take. If you build to a 7% yield on cost and the finished product trades at a 5% cap, that 200 bps spread is your margin; when the spread compresses toward zero, the risk is no longer worth it and the deal dies.

Yield on cost is 7%: $5.6M of stabilized NOI divided by $80M of total cost. Against a 5% exit cap, the development spread is 7% minus 5%, or 200 bps, a healthy margin. The stabilized value is $5.6M / 0.05 = $112M, so the value created is $112M minus the $80M cost, about $32M before financing. That spread and the $32M of value are the developer's compensation for construction and lease-up risk; if costs rose or the exit cap widened, the margin would compress fast.

Replacement cost is what it would cost to build an equivalent asset today, including land, hard costs, and soft costs. It matters because buying well below replacement cost is a margin of safety: if you own below what a competitor must spend to build new, no rational developer can add competing supply profitably until rents rise, which protects your occupancy and pricing power. It is also why, when construction costs are high, existing assets can be worth more than their in-place income alone implies.

NOI has to rise 20%. Value equals NOI over the cap rate, so to hold value flat when the cap rate rises, NOI has to rise by the same percentage the cap rate did. Going from 5.0% to 6.0% is a 20% relative increase (1.0 over 5.0), so NOI must climb 20% just to offset it. It is a sharp illustration of why cap-rate expansion is so punishing: even solid NOI growth can be entirely eaten by a modest move in rates.

It depends, and I would say what on. A 6% cap is attractive or not relative to a few things: the spread between that 6% and my cost of debt (is there positive leverage?), the NOI growth outlook, my basis versus replacement cost, the lease and rollover risk, and where the market is heading. If debt costs 5% and the asset has real rent upside and I am below replacement cost, a 6% going-in looks good; if debt is 7% and rents are flat, it does not. I would frame those tradeoffs and name the one assumption, usually rent growth or the exit cap, that would flip my answer.

I would walk it top to bottom. Start with the asset and market: property type, location, size, and the supply-demand picture. Then the in-place situation, current rents and occupancy versus market, and the business plan, whether it is core, value-add, lease-up, or development. Next the capital structure, the debt terms and the equity, and the underwriting, NOI growth, exit cap, and hold period. Then the returns, IRR, equity multiple, and cash-on-cash, and the key risks. I would close with my specific role on the deal and the one or two assumptions the outcome really hinges on.

It is usually a timed or take-home case on a sample acquisition or development. You build a property cash flow from the rent roll, vacancy and operating expenses down to NOI, then capex, tenant improvements, and leasing commissions; a sources and uses table; debt sizing and an amortization schedule; levered cash flows to equity; and a reversion at an exit cap. From there you compute the return metrics, IRR, equity multiple, and cash-on-cash, and run sensitivities, often with a JV equity waterfall if there are multiple partners. It typically ends with a short written recommendation and the key risks. Speed, accuracy, and the judgment to build it cleanly and then explain what it means matter as much as the raw mechanics.

It is a spectrum of who pays the operating expenses. In a full-service (gross) lease the landlord pays all operating costs, taxes, insurance, and maintenance, and the tenant pays one flat rent. In a modified gross lease the two share expenses, often with the tenant covering its own utilities and janitorial above a base-year stop. In a triple net (NNN) lease the tenant pays rent plus taxes, insurance, and maintenance, directly or by reimbursement. The more the tenant covers, the more predictable and bond-like the landlord's NOI, which is why net lease assets are valued so heavily on tenant credit and lease term.

Real estate hedges inflation through three channels. First, leases carry rent escalators, fixed annual bumps or CPI-linked increases, so rental income rises with prices, and short-lease sectors like multifamily and hotels reprice fastest. Second, replacement cost rises with inflation, which supports the value of existing buildings. Third, if you financed with fixed-rate debt, you repay it in cheaper future dollars. Net lease with CPI escalators is the cleanest example, because the structure passes rising costs straight to tenants while the landlord's income keeps climbing.

A stabilized property is leased to market occupancy with reliable in-place NOI, so you value it directly on a cap rate. An unstabilized asset is mid-lease-up, partly vacant, or under renovation, so its current NOI understates its potential; you value it on forward stabilized NOI and then subtract the cost, time, and risk to get there. That gap is exactly where value-add and development investors play: buy unstabilized at a discount, execute the plan, and sell or refinance once it is stabilized and commands a lower cap rate.

$15M. At 70% LTV the loan is 0.70 times $50M, or $35M, so the equity is the remaining $50M minus $35M, which is $15M. Quick and clean: equity is just the purchase price times one minus the LTV, here 30% of $50M.

Debt yield is NOI divided by the loan amount, so $5M of NOI on a $50M loan is a 10% debt yield. It is the lender's unlevered return if it had to foreclose and own the asset on day one. Lenders like it because, unlike LTV and DSCR, it ignores the cap rate and the interest rate, so it cannot be gamed by a low rate or an aggressive valuation; it is a pure income-to-loan check. CMBS lenders often want a minimum around 8 to 10%.

Debt yield is 10%: NOI divided by the loan, $5M / $50M. That sits at or above a typical CMBS minimum of around 8 to 10%, so on this loan debt yield would not be the binding constraint, the loan size would more likely be capped by LTV or DSCR. If the minimum debt yield were 10% and you wanted a larger loan, you could not get there on this NOI without the lender's debt-yield test pulling the loan size back down.

Lenders run three tests and take the most restrictive, the one that produces the smallest loan: LTV, DSCR, and debt yield. LTV caps the loan as a share of value, DSCR caps it so income comfortably covers debt service, and debt yield caps it as a share of NOI. When values are high and cap rates low, DSCR or debt yield usually binds, because the income does not support as much debt as the value implies; when values are depressed, LTV binds. Sizing a loan means solving all three and quoting the lowest.

Once NOI falls enough that DSCR drops below the 1.25x covenant, the lender's protections kick in. Depending on the loan, that can trigger a cash sweep or cash trap (excess cash flow is held by the lender instead of distributed to equity), a technical default, or a required paydown to restore coverage. Practically, equity stops receiving distributions and may have to inject capital, while the lender gains control over the property's cash. It is a clean example of how leverage turns a moderate NOI dip into an equity and control problem, which is why DSCR cushion matters so much at underwriting.

Unlevered IRR is the return on the property's own cash flows with no debt, so it measures the quality of the asset itself. Levered IRR runs the cash flows to equity after debt service, so it measures the return to the investor after financing. Leverage is accretive when the unlevered return (or yield on cost) exceeds the cost of debt, which is positive leverage and lifts the levered IRR above the unlevered one; when borrowing costs exceed the asset's yield, leverage drags the IRR down. Comparing the two tells you how much of the return comes from the asset versus the financing.

The deal that pays cash flow earlier. Both return the same total dollars, but IRR is time-weighted, so a dollar received in year one is worth more than the same dollar at sale in year five. Front-loaded cash flow compounds the return, so its IRR is higher even though the equity multiple is identical. It is a clean reminder that timing, not just total profit, drives IRR.

Neither, both are a 0% IRR. If you buy for $1M and sell for $1M with no cash flow in between, you have only gotten your capital back, a return *of* capital, not a return *on* it. With no interim cash flow and no gain there is nothing to annualize, so the hold period is irrelevant: 0% over five years and 0% over ten.

About 18%. Doubling your money is a 2.0x equity multiple, and the rule of 72 says the annual return is roughly 72 divided by the number of years, so 72 over 4 is about 18%. If the same double took five years, it is 72 over 5, or about 14 to 15%. The rule of 72 is the fast way to get close on IRR in your head, without a calculator.

The longer hold has the higher equity multiple. Same purchase, same annual cash flow, same sale price, but holding longer means you collect more years of interim cash flow, which lifts total distributions and therefore the multiple. The catch is that the longer hold usually has the lower IRR, because the cash and the sale are spread over more years. The equity multiple rewards total dollars; IRR rewards speed.

About 25%. The clean way: a 3.0x over 5 years is 3 to the power of one-fifth, roughly 1.25, so about 25% per year. As a rule-of-72 sanity check, a 3x is about 1.5 doublings, so you are doubling every roughly 3.3 years, and 72 divided by 3.3 is about 22%, in the same ballpark. Either way you land in the low-to-mid 20s, which is the fast estimate an interviewer wants without a calculator.

You need $40M of total distributions, since 2.0x on $20M of equity is $40M. To back into the exit price, subtract the cumulative interim cash flow over the hold from that $40M to get the net sale proceeds equity needs at exit, then gross that up for the loan payoff and selling costs to reach the gross sale price. So if the deal threw off $8M of cash along the way, equity needs $32M of net proceeds, and you add back debt and transaction costs to get the price the building has to sell for.

A 1031 exchange (like-kind exchange) lets an investor defer the capital gains tax on selling a property as long as the proceeds are reinvested into replacement real estate within set windows: 45 days to identify the replacement and 180 days to close. It is foundational to how much real estate trades, because it lets owners roll gains forward rather than pay tax on every sale, and it is why sellers often prefer property swaps or OP-unit deals over a taxable cash sale. The constraints are the tight clock and the like-kind reinvestment requirement.

Because the requirement to distribute at least 90% of taxable income leaves a REIT with very little retained cash to reinvest. Unlike a normal company that can fund growth from retained earnings, a REIT has to keep tapping the capital markets, issuing equity and debt, to fund acquisitions and development. That dependence is why a REIT's cost of capital and access to capital are so central to its growth and competitiveness: a REIT that can raise cheap equity and debt can grow accretively, while one with a high cost of capital is effectively shut out of new deals. It ties directly back to spread investing.

An UPREIT (umbrella partnership REIT) holds its properties through an operating partnership rather than directly, and that partnership can issue OP units. The reason the structure exists is tax: a property owner can contribute appreciated real estate into the operating partnership in exchange for OP units without triggering a taxable sale, then convert those units into REIT shares or cash later. That tax-deferral makes a REIT a far more attractive buyer to low-basis sellers, which is why the UPREIT has been the dominant US REIT structure since it appeared in 1992.

OP units are units in the REIT's operating partnership that are economically equivalent to REIT shares: they receive the same distributions and convert roughly one-for-one into shares after a holding period. A seller takes OP units instead of cash to defer the capital gains tax on a low-basis property while still gaining liquidity and diversification, since a cash sale would trigger the gain immediately. This tax-deferred contribution of property for OP units is a Section 721 (UPREIT) exchange: unlike a 1031 it has no 45- or 180-day clock, though the OP units themselves are not later eligible for a 1031.

Because GAAP net income is misleading for a REIT. Depreciation assumes buildings steadily lose value, but well-maintained real estate often holds or appreciates, so that large non-cash charge understates a REIT's true economic earnings and makes P/E close to meaningless. FFO and AFFO add the real estate depreciation back, giving a cash-based measure that actually reflects how the portfolio performs, which is why analysts quote P/FFO instead of P/E for REITs.

AFFO (adjusted FFO) takes FFO and subtracts the recurring capital the business actually needs, mainly maintenance capex, plus non-cash rent adjustments like straight-lining, and often leasing costs such as tenant improvements and commissions. FFO adds back all depreciation but ignores the real, recurring capital a property consumes; AFFO pulls that back out, so it is a closer measure of the cash truly available to pay dividends. The catch is that FFO has a standard NAREIT definition while AFFO does not, so you have to check how each company calculates it before comparing.

For a REIT you measure the payout ratio against FFO or AFFO, not EPS, since net income is distorted by depreciation. A lower payout means more retained cash, a bigger cushion under the dividend, and some internal funding for growth; a payout near or above 100% of AFFO is a red flag, because the REIT is distributing essentially all of its real cash flow and has no margin if income dips. So the payout ratio is really a dividend-safety and growth-flexibility gauge.

You value the assets, net off the liabilities, and divide by shares. Capitalize forward cash NOI at a market cap rate to value the operating portfolio, add cash, receivables, and the value of any non-NOI assets (development, management businesses), then subtract debt and preferred to reach net asset value; divide by shares plus OP units for NAV per share. For example, $90M of NOI at a 6% cap is a $1.5B portfolio; less $600M of debt leaves $900M; over 30M shares that is $30 per share. The art is in the cap rate and in valuing the non-NOI pieces.

Trading at a premium to NAV means the share price is above the per-share value of the underlying real estate, so the market is paying up for expected NOI growth, a team that can invest accretively, or a platform worth more than its bricks. Trading at a discount means the market values the company below the private-market value of its properties, which can signal distress, a poor cost of capital, or a take-private and M&A opportunity, since a buyer could acquire the whole company for less than the assets are worth. NAV premium/discount is a quick read on whether the public market agrees with private real estate values.

You compute FFO per share, then apply a peer or sector P/FFO multiple, exactly like a P/E for an operating company. For example, $100M of FFO over 50M shares is $2.00 of FFO per share; at a 15x P/FFO that is a $30 share price. The multiple comes from where comparable REITs in the same property type trade, adjusted for growth and balance-sheet quality. It is fast and market-based, which is why P/FFO is the most-quoted REIT valuation shorthand.

You reach for P/AFFO when comparing REITs with very different capital intensity. Office and hotels burn heavy TI, leasing commissions, and capex, while net lease burns almost none, so FFO flatters the capital-hungry ones; AFFO subtracts that recurring capital and normalizes for the cash actually available to distribute. For example, $90M of AFFO over 50M shares is $1.80 per share; at a 16x P/AFFO that is about $28.80. The downside is that AFFO is not standardized, so you confirm each company computes it the same way before comparing.

Same-store (or same-property) NOI measures NOI growth only from properties the REIT owned in both periods, stripping out anything acquired, sold, or developed in between. The point is to isolate organic performance, rent and occupancy growth on the existing portfolio, from growth that was simply bought with capital. Analysts focus on it because a REIT can always grow total NOI by issuing equity and buying buildings; same-store tells you whether the underlying portfolio is actually getting healthier.

Work-from-home lowered structural demand for office space and pushed vacancy up, which raises cap rates and re-leasing risk. In underwriting, that shows up as lower assumed renewal probabilities, more conservative market rents and occupancy, longer downtime, and wider exit caps. The market has also bifurcated: trophy and well-located Class A space with strong tenants has held up, while commodity Class B and C has repriced sharply and in some cases faces conversion or write-down. So the WFH effect is less about office as a monolith and more about a widening gap between the best assets and the rest.

WALT is the weighted average lease term, the average remaining lease length across all tenants, weighted by rent or by leased area. A long WALT means stable, bond-like income and little near-term rollover, which the market rewards with a lower cap rate; a short WALT means leases are rolling soon, bringing re-leasing risk, downtime, and tenant-improvement and leasing-commission costs. It is central to office and net-lease underwriting because so much of the value, and the risk, sits in when the leases expire and whether they renew.

Even though office leases are long, every new or renewed lease comes with heavy capital costs: tenant improvement allowances to build out the space, leasing commissions to brokers, free-rent periods, and downtime while the space sits empty between tenants. All of that sits below the NOI line, so reported NOI meaningfully overstates the cash an office building actually generates, which is why office AFFO and cash-on-cash run well below NOI. It is the core reason office is more capital-hungry, and often lower-quality cash flow, than a sector like net lease or industrial.

A few reasons stack up. Leases are short, roughly 12 months, so rents reprice quickly and income tracks inflation, unlike a 10-year office lease. The tenant base is granular, hundreds of units, so no single default sinks the asset, which diversifies credit risk. Housing demand is need-based and resilient through cycles. And critically, multifamily has deep, low-cost agency financing from Fannie Mae and Freddie Mac plus the broadest buyer pool of any property type, which gives it the cheapest debt and the most liquidity. Put together, that is why multifamily is the bread-and-butter, most-financeable sector.

Loss to lease is the gap between in-place contractual rents and what the same occupied units would command at today's market rents, typically because legacy tenants sit on older, below-market leases. A value-add multifamily investor attacks it by renovating units and rolling leases up to market as they expire, capturing that embedded upside. The appeal is that it lifts NOI without relying on aggressive assumptions about future market-rent growth, you are just closing a gap that already exists, which makes it the most direct and controllable NOI lever in multifamily.

The loss to lease is $150 per unit per month ($1,650 market minus $1,500 in-place). Rolling all 200 units to market adds $150 x 200 x 12 = $360k of annual revenue, which flows straight to NOI. At a 5% cap, that is worth $360k / 0.05 = $7.2M of value created. The lesson is that a seemingly small $150 per-unit mark-to-market becomes a large number at the property level, which is exactly why closing loss to lease is the core multifamily value-add play.

Physical occupancy is occupied units divided by total units, how full the building is. Economic occupancy is rent actually collected divided by gross potential rent, how full it is in dollars. The gap, usually 2 to 5 points (a stabilized property might be 93 to 96% physical but 90 to 93% economic), comes from concessions and free rent, bad debt and delinquency, non-revenue model and employee units, and below-market leases. It matters because economic occupancy is what actually reaches NOI: a building can be physically full yet underperform on income, so you always underwrite the economic number, not the headline occupancy.

The price is $250k per unit x 200 units = $50M. Total NOI is $12.5k per unit x 200 = $2.5M. So the implied cap rate is $2.5M / $50M = 5%. Whether that makes sense depends on the market: a 5% cap on stabilized multifamily is reasonable in many markets, and I would sanity-check it against where comparable assets trade and against my cost of debt. Moving fluidly between per-unit pricing and a cap rate is the core skill here.

Demand is driven mainly by e-commerce, which needs roughly three times the warehouse space of traditional brick-and-mortar because inventory has to be staged closer to customers, plus inventory restocking and supply-chain reshoring. That demand drove years of strong rent growth, and because warehouses are cheap to operate and maintain (low capex, low management intensity, often NNN leases), investors paid up, pushing industrial cap rates to among the lowest of any property type. The underwriting question is usually how far in-place rents lag the market, since mark-to-market on rollover has been a big source of NOI growth.

Last-mile facilities are smaller, infill warehouses close to dense population centers that enable fast or same-day delivery, the final leg from warehouse to customer. They command premium pricing because they are scarce and hard to replicate near cities, where land is expensive and entitlement is difficult, yet they are critical to e-commerce fulfillment. That scarcity plus essential function pushes them to the tightest cap rates in the sector, low 4s to 5s in top markets, well through big-box distribution in cheaper locations.

Cold storage is temperature-controlled warehousing for food and pharmaceuticals. It costs far more to build and run than dry warehouse because of the refrigeration infrastructure, insulation, and heavy power use, but it earns higher rents, has stickier tenants (the specialized buildout and process integration make them hard to move), and high barriers to entry that limit new supply. So it is a niche with better economics and a defensive demand profile, which is why it has drawn institutional capital despite the operating intensity.

Because the grocery anchor generates necessity-based, recurring traffic that holds up through cycles and has so far resisted e-commerce, since most grocery is still bought in person. That steady foot traffic supports the inline tenants (the coffee shop, the salon, the dry cleaner), who benefit from the customers the grocer pulls in. Stable anchor sales and rents make the cash flow durable in a downturn, which is why grocery-anchored centers are seen as the defensive corner of retail and trade at tighter cap rates than malls or power centers.

Percentage rent is additional rent a retail tenant pays as a percentage of its sales above a contractual breakpoint, on top of base rent. The natural breakpoint is base rent divided by the percentage rate, so a tenant with $300k of base rent and a 6% rate hits its breakpoint at $5M of sales; above that it pays 6% of every additional sales dollar. It lets the landlord share in a tenant's upside and aligns the two parties, which is why it is common in malls and anchored retail, where the landlord's leasing and marketing help drive traffic and sales.

The natural breakpoint is base rent divided by the percentage rate: $300k / 0.06 = $5M of sales. At $6M of sales the tenant is $1M above the breakpoint, so percentage rent is 6% x $1M = $60k. Total rent is base plus percentage, $300k + $60k = $360k. The structure means the landlord shares in upside only once the store is doing well enough to clear the breakpoint.

A co-tenancy clause lets an inline tenant cut its rent or even terminate its lease if a named anchor goes dark or center occupancy falls below a threshold, often 70 to 80%. It exists because the small tenants are paying for the traffic the anchor and the overall center generate. The risk for a landlord is that it is contagious: one anchor closure can trip co-tenancy across many inline leases at once, cascading into rent cuts and move-outs, so anchor health and occupancy effectively drive the value of the whole center, not just the anchor's own rent.

The occupancy cost ratio is a tenant's total occupancy cost, base rent plus CAM plus percentage rent, divided by its sales at that location. It is the core health check in retail underwriting: a low ratio means the store is comfortably profitable and the landlord has room to push rent at renewal, while a high ratio means the tenant is overextended and at risk of not renewing or closing. So a tenant doing $2M of sales that pays $200k all-in is at a 10% occupancy cost ratio, healthy for many categories; well into the high teens or twenties starts to flash renewal and closure risk.

Because a hotel effectively re-leases every room every night, so its revenue reprices instantly with demand and there is no contractual income backlog to cushion a downturn. Add high fixed operating and labor costs, and the operating leverage is severe: when occupancy and rate fall together, profit drops fast. That is why hotels are the most cyclical and recession-exposed property type, are run as operating businesses rather than passive leases, and are valued on RevPAR and EBITDA multiples (and per key) as much as on a cap rate.

Three linked metrics. ADR (average daily rate) is room revenue divided by rooms sold, the average rate per occupied room. Occupancy is rooms sold divided by rooms available. RevPAR (revenue per available room) is ADR times occupancy, equivalently room revenue divided by all available rooms. So a hotel at a $200 ADR and 75% occupancy has a $150 RevPAR. RevPAR is the headline hotel KPI because it captures both how much you charge and how full you are: you can lift it by raising rate, raising occupancy, or both, and revenue management is the art of finding the mix that maximizes it.

RevPAR is $150: ADR times occupancy, $200 x 0.75. For annual room revenue, multiply RevPAR by rooms and by nights: $150 x 150 rooms = $22,500 per night, and times roughly 365 nights is about $8.2M a year. In an interview I would round 365 to 360 to keep it clean, $22,500 x 360 is about $8.1M, close enough to show the method.

Because a hotel is an operating business, not a passive lease, so its P&L is built like an operating company's. You have departmental revenues (rooms, food and beverage, other) and departmental expenses, then undistributed operating expenses (admin, sales and marketing, utilities), building down to GOP (gross operating profit), and then to EBITDA after management fees, an FF&E reserve, property taxes, and insurance. The key feature is operating leverage and flow-through: with so many costs fixed, a large share of each incremental revenue dollar drops to profit, so small RevPAR moves swing EBITDA hard in both directions, which is why hotels are valued on EBITDA multiples and per key as well as on cap rates.

A franchise agreement licenses the brand (the flag) to the owner, who pays royalty, marketing, and reservation-system fees but runs the hotel itself or hires a separate manager. A management contract is with a third-party operator who actually runs the hotel day to day for a base management fee plus an incentive fee tied to profit. They are not mutually exclusive: many hotels carry a brand under a franchise agreement and a separate operator under a management contract. For an owner, the choice drives how much control, brand benefit, and operating risk it keeps versus hands off.

Medical office buildings are defensive because their tenants, physician groups and health systems, are sticky: they invest heavily in built-out exam and procedure space, build a local patient base, and renew at high rates, and healthcare demand holds up through recessions. Within MOBs, on-campus buildings (adjacent to a hospital) command premium rents and carry lower risk thanks to the health-system affiliation and patient flow, while off-campus buildings are more retail-like, competing on convenience and location. That combination of sticky tenancy and recession-resistant demand is why MOBs trade at relatively low cap rates for healthcare.

Life sciences (lab) real estate is distinct from regular office because of the specialized buildout: heavy ventilation, redundant power, lab benching, and vibration control. That means very high tenant-improvement and capex costs, but also longer leases and stickier tenants, since a fitted-out lab is expensive and disruptive to relocate. Demand clusters in a few innovation hubs near universities and talent and is driven by biotech funding cycles, when venture and public-market capital is flowing, lab demand is strong, and it cools when funding dries up. High barriers to conversion and concentration in a handful of markets make it a specialized, higher-beta corner of the office world.

The dominant driver is demographics: the aging of the baby boomers, with the 80-plus population set to grow sharply over the next decade. That creates need-based demand for assisted living, memory care, and skilled nursing that is largely independent of the economic cycle, since people move into senior housing out of health necessity, not because the economy is strong. Supply matters too, periods of overbuilding have hurt occupancy, but the long-run demographic wave is the structural tailwind that makes the sector attractive.

Data centers have become a focus sector because cloud and AI compute demand is driving enormous leasing while the supply that can meet it is constrained. Value hinges on secured power capacity (in megawatts), location and network connectivity, and creditworthy hyperscaler tenants on long leases. Because the deals involve huge capital outlays, long development timelines, and power and interconnection limits, they are often underwritten more like infrastructure than traditional real estate, with the binding question being how much power you can actually get to the site.

Because a data center needs enormous, reliable power and a grid interconnection, and securing that, not finding land or pouring concrete, is what actually gates new supply. Utilities have limited available capacity, and interconnection queues to connect a new large load can run years, so even with demand and capital in hand you cannot bring a site online until the power is committed. That scarcity is why well-located capacity with power already secured is so valuable and why supply responds slowly to surging demand, which supports rents and pricing for existing, powered assets.

Data center leases are typically triple-net or modified-NNN, but the economics are driven by power, not floor space. The key term is committed critical IT load, measured in kilowatts or megawatts, with base rent on the physical space layered on top, and power itself is metered and passed through to the tenant. Rent is quoted per kilowatt of critical capacity rather than per square foot because the binding constraint and the real value are electricity and the infrastructure to deliver it reliably, not the size of the building. So underwriting a data center lease really means underwriting committed power, the tenant's credit, and the lease term.

The main tenants are hyperscalers, the large cloud and AI operators, plus enterprises taking colocation space. Hyperscaler credit matters because these are among the most creditworthy companies in the world, and on long leases that makes the income stream bond-like, which supports low cap rates. The flip side is concentration risk: a single asset may lean on one hyperscaler, and if demand slows, hyperscalers tend to trim their leased third-party footprint before their owned facilities, so a landlord can face non-renewal exactly when the market softens. You weigh the strong credit against that tenant and lease concentration.

Because a net lease asset is essentially a credit instrument wrapped around a building. A single, creditworthy tenant signs a long triple-net lease with fixed escalations, so the landlord collects predictable, passive income with almost no operating responsibility, taxes, insurance, and maintenance all sit with the tenant. That makes value a function of the tenant's credit, the remaining lease term (WALT), and the escalation structure, priced as a spread over interest rates, much like a corporate bond. When rates rise, net lease cap rates tend to track them closely, precisely because investors evaluate the income stream the way they would a bond.

You start by capitalizing the in-place NOI, but the cap rate is driven less by the building and more by credit-style factors: the tenant's credit quality, the remaining lease term (WALT), the escalation structure, and the residual or re-leasing risk when the lease ends. A long lease to an investment-grade tenant with healthy bumps prices at a tight cap rate; a shorter lease to a weaker tenant, or one with risky residual (a special-purpose building that is hard to re-lease), prices wider. So valuing net lease is closer to a credit-spread exercise than to underwriting an operating property, and the big swing factor is what happens at lease expiry.

In a sale-leaseback a company that owns and occupies its real estate sells the property to an investor and simultaneously signs a long-term lease to keep using it. The seller does it to unlock 100% of the asset's value as capital, often cheaper than equity and without giving up operational control, and historically to move the asset off balance sheet. The buyer gets a bond-like, long-term net lease backed by the operating company's credit. It is a common way for retailers, industrial firms, and other corporates to monetize owned real estate, and it is the supply engine for net lease REITs.

In an entity (stock) deal you buy the company or partnership that holds the real estate, taking on all of its assets and liabilities. Sellers often prefer it for a clean full exit and better tax treatment, and it carries over in-place financing, the management platform, and the existing (often low) tax basis. In an asset deal you buy the property itself: the buyer gets a stepped-up depreciable basis and can leave behind unwanted liabilities, but it may trigger transfer taxes and force a payoff of the existing loan. So the choice turns on three things, transfer taxes, whether attractive in-place debt can be assumed, and whether the buyer gets to reset depreciation, and buyer and seller often want opposite structures.

A transaction goes to the entity or portfolio level when a buyer wants scale quickly, an operating platform and team, or attractive in-place financing it can assume, or when a seller wants a single clean exit and the tax efficiency of selling the company rather than asset-by-asset. Portfolio and entity deals can also capture a portfolio premium (buyers pay up for size and rarity) or a discount (a grab-bag of mixed assets), plus G&A and scale synergies a single building cannot offer. The trade-off is complexity and pricing precision: a single-asset trade is cleaner and easier to price, while an entity deal bundles good and bad assets, liabilities, and overhead together.

A follow-on offering (often an overnight or marketed deal) raises a large block of equity all at once, at a set price that is usually a discount to the market to clear the supply. An ATM sells smaller amounts of stock continuously into the market at prevailing prices. The trade-off is size versus impact: a follow-on is the tool for a big, known, lumpy need, like funding a large acquisition, where you want certainty of proceeds now, while an ATM funds ongoing or pipeline needs with less price impact and less dilution drag. Many REITs use both, an ATM for routine funding and follow-ons for large transactions.

An ATM (at-the-market) program lets a REIT sell new shares gradually into the open market over time at prevailing prices, through its brokers, rather than raising a single large block at a set, discounted price. REITs favor it because it has lower issuance costs, minimal price impact (you are feeding shares into normal trading volume), and lets them fund acquisitions and development just-in-time, matching equity raised to capital needed rather than raising a big slug and sitting on cash. It is now an extremely common funding tool, with most large REITs running active ATM programs.

An investment-grade rating lets a REIT borrow in the unsecured bond market: cheap, long-dated debt that does not require pledging specific properties. That matters for two reasons. It lowers the blended cost of capital, and it keeps the asset base unencumbered, which preserves financial flexibility, since unpledged assets can be sold, refinanced, or borrowed against later. Losing investment grade forces a REIT back onto secured, asset-by-asset mortgage debt, which is more expensive, ties up properties as collateral, and constrains the balance sheet. So defending the IG rating, mainly by managing net debt to EBITDA and coverage, is central to a REIT's capital strategy.

The main gauges are net debt to EBITDA, debt to gross asset value (or to total enterprise value), and fixed-charge coverage (how comfortably EBITDA covers interest and preferred dividends). REITs generally manage to investment-grade levels, often net debt around 5 to 6x EBITDA, because an IG rating gives access to cheap unsecured bonds and keeps assets unencumbered. Rating agencies focus on net debt to EBITDA and coverage, so those are the metrics a REIT protects to defend its cost of capital.

The main lenders, roughly from most conservative to most aggressive: banks (balance-sheet loans, construction, and bridge, often with recourse); agencies, Fannie Mae and Freddie Mac, which dominate multifamily with cheap, high-leverage, non-recourse debt; CMBS / conduit lenders, who make non-recourse fixed-rate loans and securitize them; life insurance companies, who fund low-leverage, low-rate loans on core assets; and debt funds and private credit, who do higher-rate bridge and transitional lending on assets the others will not touch yet. Which fits depends on the asset's stabilization, the leverage needed, and the borrower's appetite for recourse.

With non-recourse debt the lender's claim is limited to the property and its cash flow: on default the lender takes the asset but cannot pursue the sponsor's other assets, subject to standard "bad-boy" carve-outs for fraud or misconduct. With recourse debt the borrower or a guarantor is personally liable for any shortfall. Stabilized CRE and CMBS loans are typically non-recourse, while construction and bridge loans are often recourse, because the lender wants a guarantee while the asset is still risky. Recourse shifts risk to the borrower, so it usually prices a touch cheaper.

The capital stack is the order of who gets paid and who bears loss, from safest to riskiest. At the bottom (most senior) is the senior mortgage, lowest cost and lowest risk, paid first from cash flow and on a sale. Above it sits mezzanine debt and/or preferred equity, which fill the gap between the senior loan and the equity at a higher cost. At the top is common equity, paid last but keeping all the residual upside. The rule is simple: the higher you sit in the stack, the lower your risk and return and the earlier you are paid, while common equity at the top takes the first losses and the last dollars.

A CMBS loan is a commercial mortgage that gets pooled with many others in a trust, which issues bonds backed by that pool. The bonds are sliced into tranches by risk: from senior AAA tranches, paid first, down through mezzanine tranches to the unrated first-loss B-piece. Losses flow bottom-up, so the B-piece absorbs the first dollars of loss and earns the highest yield for it, and the B-piece buyer usually has the right to name the special servicer. CMBS loans are generally non-recourse and fixed-rate, with prepayment handled through defeasance or yield maintenance rather than a simple payoff. The model lets lenders originate and distribute rather than hold the loans on balance sheet.

Defeasance is a way to effectively prepay a CMBS loan that cannot simply be paid off. Instead of repaying the lender, the borrower buys a portfolio of government securities whose cash flows exactly replicate the loan's remaining principal and interest, and substitutes those securities as collateral, which releases the property so it can be sold or refinanced. It exists because CMBS loans are locked into a securitization and the bondholders need their payment stream preserved. It can be expensive, especially when rates have fallen since origination, because the Treasuries needed to replicate the payments cost more.

Yield maintenance is a prepayment penalty that makes the lender whole for the interest it loses when a borrower pays off early, typically the present value of the difference between the loan rate and current Treasury yields over the remaining term. The difference from defeasance is mechanical: with yield maintenance the borrower actually repays the loan and pays a make-whole fee, while with defeasance the borrower swaps in replacement securities and the loan stays outstanding. Both exist to protect the lender's and bondholders' yield against early payoff, and both are most costly when rates have fallen.

In a CMBS deal, performing loans are handled by the master servicer, but when a loan defaults or is at imminent risk it moves to the special servicer, who handles workouts, modifications, and if needed foreclosure and disposition. The B-piece buyer, the investor in the first-loss tranche, typically has the right to appoint and replace the special servicer. That alignment is deliberate: the party taking the first losses gets to control how troubled loans are worked out, since it has the most at stake in maximizing recovery. So the B-piece buyer is both the risk-taker and, effectively, the credit decision-maker on the pool.

A bridge loan is short-term, usually floating-rate financing for a transitional asset, one in lease-up, under renovation, or being repositioned, that does not yet have the stabilized cash flow to qualify for permanent (agency or CMBS) financing. It bridges the gap until the business plan is executed, after which the borrower refinances into cheaper permanent debt or sells. Because the asset is riskier and the loan often funds future capital, bridge debt carries a higher rate and fees and frequently includes some recourse or completion guarantees. It is the financing of the value-add and development world.

A construction loan funds in draws against verified construction progress rather than all at once, so the borrower pays interest only on what has been drawn, and it usually includes an interest reserve to cover interest during the build before there is any income. It is sized on loan-to-cost and on a stabilized DSCR or debt-yield test, so the lender is protected on both what it costs to build and what it will be worth stabilized. It is repaid by a takeout, a permanent loan once the project leases up and stabilizes, or by a sale. Because construction carries completion and lease-up risk, these loans are often recourse to the developer.

Both sit between the senior mortgage and common equity and raise total leverage at a higher cost, but they are legally different. Mezzanine debt is a loan, usually secured by a pledge of the equity interests in the property-owning entity rather than a mortgage lien, so on default the mezz lender can foreclose on that equity and take control relatively quickly. Preferred equity is an equity interest with a fixed preferred return that sits ahead of common; it has no mortgage lien, and its remedies come through control and redemption rights in the operating agreement, such as stepping into management on a default. Mezz is generally cheaper and faster to enforce; pref is more flexible and often used when senior lenders prohibit additional debt.

A closed-end fund has a fixed life, usually 7 to 10-plus years: investors commit capital up front, it is drawn down over an investment period, assets are bought, improved, and sold, and capital plus profit is returned by the end. It suits value-add and opportunistic strategies that need a defined hold to execute. An open-end fund (the ODCE-style core vehicle) is perpetual, with no set end date, and investors can subscribe or redeem during periodic windows, so it holds stabilized, income-producing core assets meant to be owned indefinitely. The fundamental differences are liquidity and strategy: closed-end is illiquid and driven by value creation and gains; open-end is more liquid and income-driven.

The classic shape is "2 and 20." The GP earns a management fee, often 1 to 2% of committed or invested capital, to run the fund, plus carried interest (the promote), commonly 20% of profits, but only above a preferred return to LPs, typically 8 to 9%. There is usually a GP catch-up after the pref so the GP reaches its full carry share, and often a clawback to true things up at the end. The point of the structure is alignment: the GP makes real money only after LPs get their capital back and a minimum return, so the promote rewards genuine outperformance.

There is no single right answer; cap rates are loosely anchored to interest rates through the spread over Treasuries, since real estate competes with bonds for capital, so I would read the environment off three things: the level and direction of benchmark rates, the size of the cap-rate spread over those rates (a fat spread gives cushion, a thin spread leaves values exposed if rates rise), and whether NOI growth is offsetting any rate pressure. When rates rise, cap rates tend to follow and values fall, unless the spread compresses or income growth picks up the slack. Applying that lens, as of mid-2026 rates are settling into a higher-for-longer range, roughly a 4 to 5% policy rate, rather than returning to the near-zero era; cap rates are broadly stable, with modest compression in durable-income sectors (industrial, data centers, necessity retail, senior housing) and continued pressure on challenged ones like commodity office; and spreads over Treasuries are relatively tight. The key point is that cap rates are not mechanically tied to rates, the spread and income growth can move them independently, so right now values are driven much more by income growth than by further cap-rate compression. (Time-sensitive; refresh annually.)

There is no single right answer; the cycle turns on capital flows, supply, and the cost of debt, so I would read it off a few signals: the direction of cap rates and their spread over Treasuries (rising cap rates and widening spreads signal a downturn or repricing, compression signals recovery and competition for assets), transaction volume and credit availability, the supply pipeline by sector, and the trend in rents and occupancy. I would synthesize those into a phase, recovery, expansion, oversupply, or recession, noting that sectors rarely move together, so some can be early-cycle while others are late. Applying that lens, as of mid-2026 real estate looks to be at a recovery or inflection point: after the sharp, rate-driven repricing of the prior couple of years, transaction activity and values are stabilizing, and returns have shifted toward income and NOI growth rather than cap-rate compression or financial engineering. The honest takeaway is that the easy repricing is largely done, so from here you earn returns through operations and selection, not by riding falling rates. (Time-sensitive; refresh annually.)

There is no single right answer; what matters is a defensible view, and I would build it by judging every sector on the same factors: the demand drivers and whether they are structural or cyclical, the supply pipeline (the fastest way a strong demand story gets spoiled), lease length and pricing power (short leases reprice with inflation while long leases lock it out), capital intensity, and where institutional capital is currently flowing. Applying that lens, as of mid-2026 I would put the tailwind sectors as industrial and logistics (e-commerce, low capex), data centers (cloud and AI power demand), senior housing (the demographic wave), and necessity and grocery-anchored retail (defensive, e-commerce-resistant). The most challenged is commodity office: the bifurcation thesis holds, trophy and well-leased Class A is stabilizing while commodity Class B and C faces structurally lower demand, refinancing stress, and conversions, tied to work-from-home, lease rollover, and the wall of maturing debt. I would pick a side on a sector and defend it with the drivers. (Time-sensitive; refresh annually.)