More than a billion euros into Italian shopping centres in twelve months. The 250 basis point spread over high street prices management intensity and format obsolescence risk.
Retail closed the first half of 2026 as Italy's most capitalised asset class, with roughly 2.3 billion euros invested according to Dils, and a second quarter of 1.6 billion that ranks as the sector's strongest on record. Within that result, more than 80 per cent of the capital traces to two corporate transactions: the Via Montenapoleone 8 building in Milan and a pan-European outlet portfolio with significant Italian exposure, including Serravalle and Castel Romano.
A record built on two deals describes capital availability for trophy product. It says little about the rest of the sector.
The segment that genuinely restarted
Something less visible and more meaningful sits under the headline. Over the last twelve months shopping centres attracted more than a billion euros, after years in which activity had shrunk to a handful of opportunistic trades. Second-quarter transactions include a stake in an outlet in Milan's hinterland, the full acquisition of Genoa's Waterfront retail district, and entry into 31 per cent of the European Outlet Mall Venture.
The return does not cover the sector as a whole. It covers dominant assets, those with a consolidated catchment, a defensible tenant mix and long indexed leases. Everything else remains value-add material, or finds no buyer.
What the difference costs
Prime net yields state it precisely. High street at 4 per cent, shopping centres at 6.5, retail parks at 7, standalone supermarkets at 6.25, per BNP Paribas Real Estate for the first half. A PwC reading for the first quarter puts shopping centres at 6.75 and Milan and Rome high street at 3.70: the two sources do not match, and the twenty-five basis point gap on shopping centres reflects differences in perimeter and definition more than any market divergence.
The number that matters is the spread, around 250 to 275 basis points between street shop and shopping centre. That distance prices two things together: management intensity, since a centre requires tenant rotation, restyling capital and a cost structure a high street unit does not carry, and the obsolescence risk attached to the format itself.
Supply discipline is the real protection
Italy's estate holds more than 1,270 centres, over 40,000 stores and roughly 1.9 billion visits a year. Vacancy has run between 3 and 8 per cent over the last twelve months, a range indicating equilibrium. Seven new centres are planned through 2028, a deliberately restrained pipeline made largely of refurbishment and extension.
Seven openings in three years across an estate that size means nobody is building competition to existing centres. That, more than consumer demand, is what protects rents at dominant assets. A market where stock is effectively closed and space demand holds steady produces durable positional rents, and institutional investors are paying for precisely those.
The consumption figure that rewrites the rent roll
Direct sales in Italian centres reached 70 billion euros in 2024, up 0.7 per cent. The total conceals a clear recomposition: through the first half of 2025 services grew 3.1 per cent and health and personal care 2.5, while electronics lost 1.7, homeware 1.2 and apparel 1.
That redistribution touches the economic core of the shopping centre model. Apparel historically occupies the best floorplates and pays the highest rent per square metre; services and personal care take smaller units and generate higher-frequency footfall at lower ticket. If the trend continues, the space generating the most visits is no longer the space generating the most rent, and the arithmetic will only work at centres able to reshuffle the mix without losing footfall.
Anyone underwriting an asset of this kind today is not buying retail floorspace. They are buying the probability that management can rebuild the mix over the next decade, in a setting where the catchment does not grow and physical competition does not increase. A 6.5 per cent yield pays for that probability, and centres unable to demonstrate it will keep sitting outside competitive processes.