IRR, Cap Rate & DSCR: Modeling UAE Real Estate Returns at Institutional Scale

A 7% gross rental yield in Dubai looks great until you institutional-grade the model. Once leverage, vacancy, service-charge drift, exit cap rate, and DSCR constraints enter the spreadsheet, the headline number collapses to something more honest. This is how funds and family offices actually underwrite UAE real estate assets — and where retail-style ROI calculations consistently mislead.

Cap rate: going-in vs stabilised

The going-in cap rate is the net operating income at acquisition divided by purchase price. The stabilised cap rate uses NOI at full occupancy and normalised expenses. For UAE residential assets in transition (recently delivered, lease-up phase, or post-refurbishment), the gap can be 80–150 bps.

Institutional convention: underwrite to stabilised NOI, then haircut for the first 12–24 months of lease-up.

Typical 2026 cap rates by asset class in Dubai prime areas:

• Residential — value (Discovery Gardens, International City, JVC): 6.5–7.5% gross, 5–6% net.
• Residential — prime (Marina, Downtown, JBR, Palm): 4.5–6.0% gross, 3.5–4.5% net.
• Commercial — Grade A office (DIFC, Business Bay): 7.5–9.0% net.
• Hospitality — branded residences with leaseback: 6.0–7.5% net, with sponsor-credit risk overlay.

Unlevered vs levered IRR

Unlevered IRR ignores debt — useful for asset comparison. Levered IRR includes mortgage cash flows — useful for equity return modelling.

A typical Dubai residential underwriting at 60% LTV with a 5-year UAE variable-rate mortgage shows levered IRR 200–400 bps above unlevered, depending on rate path and exit cap. The catch: leverage amplifies downside as much as upside, and UAE variable-rate exposure means a 200 bps rate shock can move levered IRR by similar magnitude.

Institutional practice: present both unlevered and levered IRR. Stress-test levered IRR at +200 bps rate move. Run a downside case at -10% exit price.

DSCR: the UAE bank threshold

Debt Service Coverage Ratio is the NOI divided by annual debt service. UAE banks typically require DSCR > 1.25x for commercial real estate financing; 1.15x for residential investment. Major UAE lenders (Emirates NBD, ADCB, Mashreq, FAB) publish broadly similar standards.

Practical implication: at 60% LTV on a 25-year UAE mortgage at ~5.5% all-in rate, a Dubai residential asset needs net yield above approximately 4.2% just to hit DSCR. Below that, financing is unavailable on conventional terms.

The exit cap-rate question

Underwriting any 5–7 year hold requires an exit cap-rate assumption. UAE residential historical cap-rate volatility is high — going-in cap rates have ranged 5–9% across cycles. Institutional convention: assume exit cap = going-in cap + 50 bps (cap-rate widening, conservative). Stress-test at +100 bps.

Retail-style ROI calculations almost always assume flat cap rates — this is the single biggest source of institutional-vs-retail underwriting divergence.

Where retail-style ROI breaks down

• Service-charge inflation ignored — retail models hold service charges flat. Institutional models inflate 4–6% per annum.
• Vacancy underestimated — retail models often assume 100% occupancy. Institutional models build in 8% vacancy in prime, 12%+ in value areas.
• Maintenance under-reserved — retail models use AED 1–2/sqft. Institutional models reserve 8–12% of gross rent for maintenance and capex.
• No exit cap-rate widening — retail models use flat cap. Institutional adds 50–100 bps for exit.
• No DSCR check — retail models assume the loan is available. Institutional models confirm DSCR > 1.25x at base case and 1.05x at downside.
• No transaction-cost amortisation — retail models ignore the 6–8% in/out friction. Institutional models include both entry (DLD 4% + agency 2% + closing) and exit (agency 2%, possibly DLD on buyer side).

The institutional underwriting template

A complete institutional underwriting model contains:

• Year-by-year NOI projection (10-year window typical), with explicit assumptions on rent inflation, vacancy, service-charge inflation, and reserve.
• Debt schedule with rate-reset path (UAE variable rates reset annually).
• Capex schedule for major works.
• Exit pro-forma: stabilised NOI at year of exit / exit cap rate, less transaction costs.
• Three scenarios: base, upside (-25 bps exit cap, +10% rent inflation), downside (+100 bps exit cap, +200 bps rate shock).
• DSCR check at base and downside.
• Levered and unlevered IRR for each scenario.

Where REMAP fits

Run the deal-level math on REMAP's Dubai ROI calculator for a starting point, then institutional-grade it offline against the template above. REMAP's institutional workspace tracks the full underwriting pack against the realised cash flows over the hold period — quarterly variance reports show where the model assumptions held and where they slipped.

Common modelling errors

• Cap rate confused with yield — they are the same thing in the simple case (NOI/price) but cap rate is a market-implied discount rate while yield is a buyer-specific return.
• Stabilised NOI used at acquisition — overstates Year 1 cash. Use going-in NOI with stabilisation modelled separately.
• DSCR computed at going-in rather than after rate reset — variable-rate mortgages reset annually; DSCR at Year 2 may be tighter than at Year 1.
• Exit cap held flat — historically the single largest model-vs-actual divergence.
• No downside case — institutional capital won't subscribe to a single-scenario underwrite.