Income Real Estate Cap Rates: Risk Premium by Asset Type

In income-producing real estate filings, Cap Rate is not always the capitalization rate used in the valuation. Sometimes it is a derived return: representative or normalized NOI divided by the fair value, or net attributed value, of income-producing assets after company-defined adjustments. A horizontal study of cap rates therefore has to start with metric classification, not with collecting every number shown under the same label. After that filter, the Israeli capitalization-assumption map sits mostly from the mid-5% area to the high-7% area, with quality logistics and industrial assets sometimes below offices, core retail in the middle of the range, and weaker retail or special-use assets around 8% or higher. Q1 2026 adds clean valuation data points, especially at Gav-Yam and Isras, but it does not prove a broad move in cap rates. The professional conclusion is a map, not a trend: the filings show where the risk premium sits, but a consistent same-method comparison is still needed before saying the market changed assumptions.

Metric Definition: Valuation Input vs Derived Yield

Three different numbers have to stay separate. The first is a rate used in the valuation: a direct capitalization rate, a DCF discount rate, a terminal capitalization rate or WACC explicitly presented as part of the valuation model. In this article, "cap rate" is the umbrella term for that valuation-input layer.

The second is a derived yield. In those rows, the company divides representative or normalized NOI by the fair value of income-producing real estate, or by a net attributed value after excluding land, rights, vacant space, development assets or other adjustments. This is useful for testing rent against asset value, but it is not the assumption used in the valuation. It is also not the public company's stock-market value.

The third is a public-market implied yield, such as NOI against market capitalization or enterprise value. That is a different analysis: it tests how the stock market prices the company, not which rate was used to value the asset. It stays outside this article.

Cap-Rate Ranges by Asset Type

After derived-yield rows are removed, the table becomes smaller but more precise. It does not answer which company has the lowest rate. It answers where a higher risk premium is required by asset type.

These ranges are not an automatic ranking of conservatism between companies. A company that owns a mall, shopping center, office building and logistics park can disclose several different valuation assumptions in the same filing. The correct comparison unit is therefore asset type and valuation method, and only then the company name.

What Q1 Adds

Gav-Yam provides a clean example of the required separation. In the same disclosure layer, it presents a 6.7% weighted return derived from income-producing property, and separate average rates according to valuations of 6.7% for high-tech and offices, 6.3% for logistics and industry, and 8.0% for retail. Only the second group belongs in the valuation-assumption table.

Isras adds a more asset-level layer. Its Q1 material-valuation appendix shows capitalization rates of 7.0% for the Jerusalem Technology Garden, 5.75%-7.0% for Har Hotzvim, 5.75%-6.7% for Ramat Hahayal, 6.5% for Holon, 6.75%-7.0% for Rehovot and 5.75%-7.0% for Park Gisin. Most of those valuations are dated December 31, 2025, and the March 2026 value update is small for many assets. Park Gisin is more notable, rising from NIS 858.0 million at year-end 2025 to NIS 876.5 million at the end of March.

Valuation Sensitivity to Capitalization Assumptions

The distinction between a valuation assumption and a derived yield changes the balance-sheet reading. Amot discloses a sensitivity analysis showing that, based on normalized NOI of NIS 1.078 billion, a 25 bps change in the weighted capitalization rate changes the fair-value adjustment by NIS 680 million, or about NIS 524 million after deferred taxes. At Giron, a 50 bps increase in the capitalization rate reduces fair value by about NIS 117 million, while a 50 bps decrease adds about NIS 137 million.

Those figures explain why a table of property yields is not enough. A company can have high occupancy, stable NOI and index-linked leases while remaining highly sensitive to a small change in the capitalization assumption used in the valuation. For accountants, analysts and valuers, this is a question of fair value, equity and leverage, not only operating yield.

What Stays Outside the Core Sample

Weighted-return rows should remain in the appendix, not in the core table. They can help test the reasonableness of asset value against NOI, but they do not replace the assumptions used in the valuation. If the filing does not explicitly disclose the rate used in the valuation, it should not be inferred from an output Cap Rate alone.

Special models should also stay outside the average. Residential rental, senior housing, data centers, gas-station complexes and development assets may be investment property or income-producing assets, but their economics are not the same as stabilized offices, malls or logistics assets. Non-Israeli assets are useful for geographic risk-premium comparison, but not for the local core index.

What Can Be Concluded Now

The current research is enough to build a capitalization-assumption map by asset type, but not enough to establish a broad cap-rate trend. To claim that the market changed assumptions, the comparison has to follow the same valuation data across several periods: the same asset or asset group, the same method, and the same type of rate.

If value rises while cap rates stay stable, the explanation will be NOI, indexation, new leases, occupancy improvement or development completion. If the rates themselves begin moving, that is a change in the valuation layer, with a faster effect on fair value and equity for leveraged companies. At this stage, the map shows where the risk premium sits, not that it has already changed across the sector.