Retail Was Supposed to Die. Office Might Be the Riskier Asset Now.
Retail is not staging a broad comeback. Something more investable may be happening: values have reset, the surviving formats are easier to understand, and some office risks that once looked cyclical are becoming structural.
Acquirepad Research ·
- EUR 1.9bn retail investment in H1 2026
- -35% retail investment year on year
- 44% of H1 volume in grocery and retail-warehouse formats
- +2.7% real German retail sales in 2025
- +10.1% real online and mail-order sales in 2025
The headline does not look like a comeback. German retail-property investment volume fell 35% year on year to EUR 1.9 billion in the first half of 2026. [1] If the thesis were simply that investors love retail again, the data would contradict it.
The more interesting change is beneath the volume. Higher-quality assets and portfolios are returning to market. Grocery stores, retail warehouses and retail parks accounted for 44% of H1 volume. In the first quarter, grocery-anchored formats represented 56% of activity, while shopping centres accounted for 22% and were increasingly being bought with repositioning or mixed-use strategies in mind. [1] [2]
Retail is becoming investable again in selected formats without becoming fashionable again.
| Signal | Latest data | What it suggests |
|---|---|---|
| H1 retail investment | EUR 1.9bn, -35% YoY | No broad capital-market boom |
| Grocery / retail warehouse share | 44% of H1 volume | Defensive formats still lead |
| Shopping-centre share | 22% of Q1 volume | Selective return via repositioning |
| Real retail sales, 2025 | +2.7% | Consumer spending recovered |
| In-store retail sales, 2025 | +1.1% | Physical retail grew, but slowly |
| Online / mail-order sales, 2025 | +10.1% | Digital channel keeps gaining |
Sources: CBRE and Destatis. Destatis notes that 2025 online growth was partly affected by a statistical restructuring effect. [1] [2] [3]
Retail has already paid for its disruption
E-commerce forced retail property to confront technological change early. Weak department-store models failed. Shopping-centre values repriced. Owners learned that tenant mix, food, convenience, leisure, services and alternative uses require active asset management.
That history matters for relative value. Investors no longer underwrite every retail asset as a bond-like rent stream. The market already assumes operating risk. In some office segments, by contrast, the full cost of hybrid work, energy capex and obsolescence is still being discovered.
This is not an argument that retail is safer in general. A declining high street with weak demographics can still be a poor investment. It is an argument against assigning an automatic structural-risk premium to the word retail while treating office as a single lower-risk category.
The consumer did not move entirely online
German retail sales increased 2.7% in real terms in 2025. In-store retail sales rose 1.1%, while online and mail-order sales rose 10.1%. [3] The online figure was partly boosted by a statistical restructuring effect, which is important when interpreting the growth rate.
The direction remains clear: online keeps taking share, but physical retail is not disappearing. The more useful question is which functions remain valuable in a physical location. Grocery, convenience, service, brand experience, returns, click-and-collect and local fulfilment all have different economics.
That is why format matters more than the aggregate retail label.
The office comparison is where the valuation question becomes interesting
German office investment rose 66% in H1 2026, while Big 7 vacancy reached 8.5% and prime rents still increased. [4] [5] This is a market in which top assets can perform while weaker stock deteriorates.
A functioning retail park can offer necessity-based demand, limited competing supply and an entry yield that already reflects years of scepticism. A secondary office can face hybrid-work risk, re-letting incentives, fit-out contributions, energy capex, refinancing pressure and uncertainty about whether the building will remain competitive after the next lease event.
Which one deserves the larger yield? There is no universal answer. That is precisely the point. The old ranking of sector risk should be rebuilt from the asset-level cash flows rather than inherited from the last decade.
Shopping centres are becoming land-and-location plays
The return of selected shopping-centre transactions is especially interesting because many centres sit on large, accessible urban sites. Their value does not have to be maximised by preserving the existing retail GLA forever.
Residential, healthcare, leisure, hospitality, services, education and last-mile functions can all form part of a repositioning strategy where planning and economics allow. This creates an option that a conventional retail valuation can miss: the asset may be worth more as a mixed-use location than as a shopping centre with the current tenant mix.
That option is not free. Conversion requires capex, planning risk and often complex phasing. But it means the downside is not determined only by current retail rent.
A comeback can happen without a consumption boom
Property returns depend on the gap between expectations and reality. Retail does not need consumers to rediscover 2007 shopping behaviour. It needs operating performance to be better than the pessimism already embedded in price.
The reverse can be true for assets whose historical pricing assumed stability. If secondary office income proves less durable than expected, a lower starting yield can deliver a worse return even when the office sector as a whole is more liquid.
This is why the most interesting retail opportunity may be relative rather than absolute: sectors that have already repriced can outperform sectors where the obsolescence bill is still being calculated.
The comparison should be made through forward cash flow, not reputation
A proper relative-value decision compares the durability and capital intensity of the cash flows. The matrix below is intentionally qualitative because public German datasets do not provide a consistent asset-level panel of yields, incentives, capex and financing terms across these formats. It shows the questions the headline needs to survive.
| Underwriting dimension | Repriced retail park / grocery-led | Secondary office |
|---|---|---|
| Demand risk | Local catchment, tenant category and operator strength | Hybrid work, submarket depth and tenant consolidation |
| Capex risk | Reconfiguration, ESG, tenant fit-out, site works | Energy retrofit, fit-out, incentives, technical obsolescence |
| Supply risk | Planning and land availability by catchment | Competing quality stock plus obsolete vacancy |
| Financing risk | Depends on format, covenant and lease profile | Can be highly sensitive to quality, capex and lease expiry |
| Residual-value option | Alternative retail mix / selective mixed-use potential | Refurbish, convert, redevelop or accept structural vacancy |
Acquirepad Research framework. It is not a claim that either column is universally lower risk.
The comparative thesis becomes investable only when the entry price is added. At a 6% yield, one unit of NOI is capitalised at 16.7x; at a 5% yield it is 20.0x. A 100bp yield difference therefore represents roughly 20% more price for the same current NOI. Starting yield can absorb or amplify operating mistakes, which is why a sector that has already repriced can outperform without having better absolute fundamentals.
What could invalidate the thesis?
A renewed drop in real consumption, a wave of retailer insolvencies, accelerating online substitution in categories that still support physical stores, or aggressive new retail supply would weaken the case. So would a faster-than-expected stabilisation of secondary offices combined with lower retrofit costs and stronger tenant demand.
The changed belief is deliberately narrow: retail is no longer automatically the structurally riskier property sector. For some assets, the market may already have priced more disruption into retail than it has priced into secondary office.
Methodology and limitations
The note combines CBRE/JLL transaction and office-market evidence with Destatis retail-sales data. The central retail-versus-office claim is a relative-value hypothesis, not a finding that all retail is safer than all office. Public data do not provide a clean matched sample of German retail parks and secondary offices with consistent entry yields, capex, incentives, financing and realised returns. The underwriting matrix therefore states the variables that should be compared; the 5%/6% yield example is a mathematical sensitivity, not a current market quote.
Notes and sources
Numbers in the article are drawn from the sources below. Interpretive statements and scenario analysis are Acquirepad Research unless otherwise stated.
- [1] CBRE Germany, Retail investment market subdued in H1 2026 but market supply growing, 6 July 2026.Source
- [2] CBRE Germany, Germany investment market Q1 2026, retail trends.Source
- [3] Statistisches Bundesamt, Retail sales in 2025 up 2.7% in real terms, 2 February 2026.Source
- [4] CBRE Germany, Real estate investment market on the up in H1 2026, 8 July 2026.Source
- [5] JLL Germany, Big 7 office market dynamics, Q2 2026, 14 July 2026.Source
- [6] JLL Germany, Germany retail market dynamics, H2 2025, 27 February 2026.Source