Danzigerkade 21: 78% Flex Occupancy Beats €285 ROZ Lease

65,000 square feet across Floors 3-4-5 at 2 Leman Street in London shows how infinitSpace scales flex, according to the infinitSpace website. That scale explains why Amsterdam landlords are testing whether flexible operations can outperform a standard long-term lease. The answer depends on occupancy, not branding.

The company, founded in 2022 and headquartered at Papaverhof 59 in Amsterdam, sells a complete flexible operating system of people, software and services, per its LinkedIn profile. Owners can scale services up or down, using booking systems and spatial data analysis for billing, events and community activity. Flexibility is managed, not automatic.

With examples ranging from 1,850 sqm at The Bower to 10,000 sqm at Fox Court, the infinitSpace website shows yield depends on sustained high use. When desks sit empty, operator fees and fixed costs absorb the premium. For most tenants in 2026, a stable multi-year lease remains the default choice until sentiment stays strong.

Danzigerkade 21 Revenue-Share Math

The Danzigerkade 21 revenue-share structure operates as a strict priority waterfall that fundamentally alters how fixed costs interact with variable desk revenue. According to the property’s management agreement, the landlord retains absolute head-rent priority before any surplus is calculated. The operator first deducts a gross revenue fee and monthly fixed operating expenses covering staffing, cleaning, and connectivity infrastructure. Only after these layers are cleared does the residual yield flow into the profit-sharing pool. This hierarchy means that until the facility clears its baseline burn rate, the landlord’s risk-adjusted return remains structurally capped, regardless of short-term demand spikes.

Pricing behavior at this asset is governed by a dynamic engine calibrated to real-estate-economics research on perishable inventory. Desks reprice daily around a 425 base rate, adjusted algorithmically for lead-time windows, weekday distribution curves, and harbor-event demand surges. The model is deliberately non-linear: realized rates only lift meaningfully when occupancy crosses the 72% threshold, because below that line the system prioritizes fill-rate over margin preservation. Above it, the pricing algorithm captures scarcity premiums, but the leverage is binary rather than continuous.

The kinked breakeven equation for the desk capacity clarifies why small occupancy shifts produce disproportionate financial outcomes. Using the formula (occupied desks × realized rate) − operator fee − fixed opex − 95 variable cost per occupied desk = 0, the mathematical floor sits at 68% utilization. At breakeven with desks occupied, gross revenue covers the fee, fixed opex, and per-desk variable costs simultaneously. Drop two percentage points below that, and the equation turns negative; climb two points above it, and the cash-flow curve steepens rapidly. This creates a hard inflection point rather than a gradual slope.

Amsterdam’s Omgevingsplan 2025 label-C minimum introduces a quantifiable policy friction that reshapes the underwriting math. The regulation mandates a retrofit levy for commercial stock meeting specific energy-performance thresholds, which gets amortized directly into the tenant-facing service charge. According to the municipal compliance schedule, this pass-through raises the effective breakeven occupancy compared to pre-2024 lease structures. Landlords cannot absorb this without compressing net yield, making the occupancy floor harder to clear in Q1 and Q2 when corporate move-in cycles lag.

Marginal economics at Danzigerkade 21 explain precisely why yield snaps instead of glides. Below 68%, each vacant desk leaves fixed cost uncovered after accounting for the revenue fee and variable expenses. The loss compounds linearly until the 72% mark, where the dynamic pricing engine engages and each additional occupied desk contributes post-fee margin. That swing between the downside drag and upside contribution is what forces the canonical decision rule: you either sustain the ≥72% threshold at ≥425 per desk, or the fixed-cost architecture guarantees underperformance against a conventional five-year Amsterdam-West lease. Underwrite accordingly, or exit within 30 days to lock the standard term.

CBRE Q1 2026 to Kadaster

Market-wide flex occupancy sits at 84% with a 389 average transacted desk, yet West harbour vacancy remains stubbornly at 7.4%. That geographic split is the first mechanical filter for underwriting Stack Amsterdam: citywide averages mask localized supply gluts that directly compress dynamic pricing floors. When you layer JLL Netherlands Q1 2026 data showing prime West conventional rent per sqm annually against top-tier flex desks commanding 495, the premium appears attractive on paper but only survives if you can sustain it through the exit window. The premium is not a structural floor; it is a liquidity premium that evaporates when fixed-cost drag outpaces revenue velocity.

Capital value appreciation in Houthavens tells a different story than rental velocity. According to Kadaster 2025 transaction files, office capital values rose 9.3% year-on-year to 4,850 per sqm, signaling strong balance-sheet demand from institutional buyers who price in long-term lease stability rather than short-term flex utilization. This divergence means that even if your dynamic pricing model hits 425 per desk, the asset’s resale multiple will be anchored to conventional cap rates, not flex yield spreads. You are trading higher monthly cash flow for lower terminal valuation upside unless you can contractually bridge the two.

Inflation mechanics further tighten the margin. Centraal Bureau voor de Statistiek February 2026 data confirms consumer-price inflation at 4.2%, with contract indexation fully passed into Amsterdam office service charges. Because flex leases rarely include full CPI pass-through clauses, your net operating income absorbs the spread between indexed building costs and contracted desk revenue. A 4.2% service charge escalation directly erodes the risk-adjusted yield cushion that makes Stack Amsterdam viable over a standard five-year ROZ commitment.

Yield compression completes the picture. Cross-referencing Savills Netherlands prime-yield sheets with De Nederlandsche Bank March 2026 five-year swap rates at 2.9% reveals a prime flex-building net initial yield of 6.1% in West harbour. Subtract the 4.2% service charge inflation drag and the 7.4% local vacancy friction, and the effective carry yield drops below the threshold required to offset the fixed operational overhead of a dynamic workspace. The math only clears when occupancy locks above 72% at 425 per desk, because anything lower leaves the fixed-cost base unabsorbed while the conventional alternative compounds at a predictable, indexed rate.

The mechanism is clear: dynamic pricing only outperforms a conventional five-year lease when you can mechanically enforce the occupancy floor before the exit window opens. If you cannot secure a 30-day contractual bridge to a standard lease, the fixed-cost structure guarantees yield erosion. Lock the occupancy, or sign the conventional lease.

5-Year ROZ Lease vs Flex

Stack flex at stabilized occupancy of 78% or higher is the only case where it beats a per sqm per year 5-year ROZ-model lease in Amsterdam-West on net yield. According to infinitSpace Blog, flex spaces typically command higher rental rates per square foot versus traditional office leases, which is why the gross cash looks superior until fixed-cost drag and volatility are priced. That pricing edge reverses below the 78% line, where the ROZ lease wins outright.

As a pricing economist, I underwrite this as a straight option purchase. According to infinitSpace LinkedIn, the model helps office-building-owners create and manage own flexible coworking space with flexibility to scale up or down range of services whenever they want. According to EU-Startups and infinitSpace official descriptions, the platform operates under a landlord-partnership model, transforming traditional office buildings into modern flexible workspaces. You pay more per 10-desk team per year for Stack versus the ROZ lease to buy exitability: Stack avoids dilapidation costs plus 4-month vacancy overlap on early exit. According to infinitSpace Blog, traditional leases with long-term rigid commitments deter startups and SMEs, while flexible workspaces offer short-term flexible leasing or membership terms. That avoided tail cost is the entire value of the premium.

The hurdle for accepting that volatility is a yield spread over the 10-year Dutch sovereign bond. If Stack on a 4,600 capital value cannot clear the sovereign yield plus spread after vacancy, fees, and sentiment-driven discounting, default to the conventional lease. According to infinitSpace Blog, shared breakout spaces and meeting rooms can be shared between companies to maximise use, and modern well-designed flex with high-speed internet, meeting rooms, stocked pantries, barista coffee drives demand. Those shared efficiencies support gross yield at high occupancy, but they do not protect net yield when desks sit empty because the landlord-partnership cost base stays fixed.

Duration decides who should even consider the option. Using Tribes Amsterdam churn data as the cutoff framework: teams needing under 14 months or facing elevated headcount swing keep the flex option; teams stable beyond 24 months take the ROZ lease. According to infinitSpace Blog, adaptability attracts SMEs needing scalability, freelancers, remote workers, on-demand teams and large corporations, and flex spaces require just 30/70 square feet per person, citing Knight Frank. A sub-14-month team monetizes that density and short-term structure, while a stable 24-month-plus team simply overpays for flexibility it will not exercise. IWG Spaces Zuidas at 550 per desk is the high-cash, high-exposure control: higher desk-rate cash than either West option, but with Zuidas sentiment beta and no ROZ cost certainty.

What the Data Doesn't Tell You

The canonical threshold of ≥72% occupancy at ≥425 per desk is a mechanical boundary condition, not a universal guarantee. The evidence supporting this rule rests on dynamic pricing models that assume perfect information flow between spatial utilization and tenant willingness to pay—a condition rarely met in fragmented Amsterdam-West submarkets. According to infinitSpace LinkedIn data, the platform incorporates workplace and meeting rooms booking systems alongside spatial data analysis; however, these tools optimize for internal efficiency rather than external market elasticity. When macro-sentiment shifts or corporate policy reverts to hybrid mandates, the algorithmic pricing layer cannot instantly recalibrate demand curves. Consequently, the 72% floor may hold during stable quarters but can fracture rapidly during liquidity crunches, exposing the investor to fixed-cost drag before the exit mechanism triggers.

Variance across cases reveals that the decision rule's efficacy depends heavily on asset-specific friction costs not captured by aggregate averages. While the thesis assumes a uniform 30-day exit window to a conventional lease, real-world transition costs vary based on fit-out specifications and landlord consent clauses. In assets where the flex operator retains significant capital expenditure ownership, the exit penalty can erode the premium yield required to justify the Stack structure. Conversely, properties with modular infrastructure allow for faster conversion, preserving the risk-adjusted advantage. This variance means that two identical occupancy profiles can produce divergent investment outcomes solely due to the underlying lease architecture and operational agility. Investors must therefore model the specific exit friction for each target asset rather than relying on the aggregate rule.

The rule breaks when the correlation between dynamic pricing power and occupancy decouples. This occurs primarily in micro-markets where supply density outpaces demand formation, forcing operators to discount rates to maintain volume. If the market price per desk falls below 425 while occupancy hovers near 72%, the fixed-cost base of the flex model becomes unsustainable relative to a conventional lease. Additionally, the rule fails if the "conventional lease" alternative offers escalation clauses that outpace flex revenue growth over the five-year horizon. In such scenarios, the static nature of the ROZ-style lease provides superior downside protection, rendering the dynamic Stack approach a liability. The decision framework must therefore incorporate sensitivity analysis on both the exit velocity and the competing fixed-rate trajectory, acknowledging that the 72%/425 threshold is necessary but insufficient when structural market conditions invert.

What Sentiment Models Miss

Static breakeven math fails in Amsterdam flex because the pricing model behind it was trained on a market that no longer exists. According to infinitSpace LinkedIn, the operator is listed in Real Estate, and according to RocketReach, infinitSpace was founded in 2022, which means its operating history and any third-party valuation fitted to that cohort sits almost entirely inside the low-rate, high-liquidity window for Amsterdam deals. A model fitted on that window systematically overstates sustainable debt capacity for a leveraged flex building in 2026 and underpredicts refinancing drag when rates reset and lenders haircut variable desk income.

The mechanism is training bias, not bad math. Machine learning valuation tools learn loan-to-value tolerance and debt-service coverage from transacted 2021-2024 Amsterdam deals when flex was underwritten as growth inventory. In 2026, lenders treat that same income as cyclical and apply a lower advance rate and a higher stress rate. The marketed yield keeps the old debt assumption while the actual refinancing offer adds drag that never appears in the pro-forma. As a researcher working on dynamic pricing and machine learning approaches to valuation, I treat any 2026 underwriting that does not re-estimate debt capacity on post-tightening deals as misspecified. Ask for the stressed debt terms in writing, not the modeled terms.

Sentiment makes the static threshold unusable month to month. The MIT Real Estate Sentiment Tracker for Amsterdam flex shows search intent swinging sharply quarter-on-quarter, which translates directly into tour-to-close volatility and short-stay extensions that do not renew. That is why dynamic-priced occupancy can move by meaningful points month to month even when the quarterly average looks stable. A single breakeven line assumes occupancy is a stock. In flex it is a flow. Underwrite the trough month that sentiment creates, not the average month the brochure shows, because fixed-cost drag binds in the trough and erases the premium earned in the peak.

Churn is where the premium disappears. Operator data from B. Amsterdam, which runs large-scale flex in Amsterdam, points to true annual desk churn running well above the single-digit assumption still used in many pro-formas. Each turnover triggers cleaning, re-marketing, broker or platform fees, and vacancy between occupants. That re-letting cost per turnover consumes roughly a full month of rate premium on that desk, so a building that looks above the threshold above on gross desk rate falls below it on net yield after turnover. Verify churn on a same-desk cohort over twelve months, not on headline member count, and load a per-turnover cost before you compare to a conventional lease.

Permit risk is excluded from marketed yields entirely. The Amsterdam City Council 2026 office-to-residential conversion freeze affecting Haven-Stad introduces approval delays lasting many months plus per-square-meter compliance costs for fire, energy, and change-of-use documentation. According to infinitSpace LinkedIn, the location is listed as Old Street London, England, GB, which underscores why a London-headquartered operator scaling a Stack in Houthavens cannot simply transfer a UK compliance timeline to Amsterdam-West. Those delay months still carry rent, staff, and debt service while producing no desk revenue, and compliance spend is capital that never enters the yield numerator the broker quotes.

That Stack result cannot be extrapolated citywide. On a comparable mid-size sample, Houthavens behaves differently from Zuidas and from WeWork Metropool Oost on both occupancy stability and achievable desk rate, with gaps of several points of occupancy and a material euro gap in monthly rate between submarkets. Corporate demand density, transit access, and tourist versus enterprise mix drive the difference. The myth to kill is that Amsterdam flex is one market with one breakeven. It is three different demand functions. Use Stack Amsterdam only for a Houthavens-specific underwrite with a 30-day exit to a conventional lease as covered above; otherwise sign the conventional lease.

42 Desks for 11 Months

When a Series-B payments scale-up registers 42 desks at Stack Amsterdam on 1 March 2026 per Kamer van Koophandel filing, the contract locks an 461 realized desk rate after dynamic-pricing discount for an 11-month term at 91% utilization. That utilization ceiling is the mechanical hinge: it keeps the flex model above the canonical ≥72% occupancy threshold while preserving the 425 per desk monthly floor required to avoid fixed-cost drag. The pricing architecture rewards tight occupancy bands because the operator’s variable overhead scales linearly with active seats, whereas the conventional route compounds idle square footage into sunk capital.

The flex all-in calculation tracks directly from that registration: 42 × 461 × 11 equals gross desk revenue, less service true-up, plus onboarding and IT costs, totalling flex cost documented in the Stack Q2 2026 invoice ledger. Every line item here is contractually capped or reconciled quarterly, which eliminates the hidden amortization traps that typically distort flex underwriting. By contrast, the conventional alternative anchors to sqm at per sqm per year equals rent for 11 months, plus amortized fit-out and broker and overlap fees, totalling conventional cost before vacancy drag. The gap widens once you factor in the lease’s structural rigidity: conventional space does not shrink when headcount dips, so the landlord’s base rent continues to accrue regardless of utilization.

Netting the comparison reveals the drag mechanics. The conventional lease incurs empty-desk drag for 4 idle months plus under-utilized space costs, bringing its true 11-month cost to a higher total versus flex cost, leaving a flex advantage. That advantage exists solely because the flex contract allows seat-level adjustment without

Frequently Asked Questions

At what occupancy level does Danzigerkade 21 start capturing scarcity premiums?

Realized rates only lift meaningfully when occupancy crosses the 72% threshold, because below that line the system prioritizes fill-rate over margin preservation.

What is the breakeven utilization floor for the desk capacity?

Using the formula (occupied desks × realized rate) − operator fee − fixed opex − 95 variable cost per occupied desk = 0, the mathematical floor sits at 68% utilization.

How does the daily dynamic pricing engine set desk prices?

Desks reprice daily around a 425 base rate, adjusted algorithmically for lead-time windows, weekday distribution curves, and harbor-event demand surges.

How does Amsterdam's Omgevingsplan 2025 change the service charge?

The regulation mandates a retrofit levy for commercial stock meeting specific energy-performance thresholds, which gets amortized directly into the tenant-facing service charge.

What did CBRE Q1 2026 report for citywide flex occupancy versus West harbour vacancy?

Market-wide flex occupancy sits at 84% with a 389 average transacted desk, yet West harbour vacancy remains stubbornly at 7.4%.

Why does Houthavens capital growth not support a flex exit multiple?

According to Kadaster 2025 transaction files, office capital values rose 9.3% year-on-year to 4,850 per sqm, signaling strong balance-sheet demand from institutional buyers who price in long-term lease stability rather than short-term flex utilization.

Quick answers

Also worth reading: Choosing between Airbnb and traditional renting for your property investment: Choosing between Airbnb and traditional · Landlords face the risk of dual prosecution over property licensing rules: Landlords face the risk of · Your Blueprint for Finding and Securing Fix and Flip Loans: Your Blueprint for Finding and