Following the recent developments in the asset pricing literature of geographic diversification, we abandon classic theories based on information asymmetry and explain the short-run performance of REIT IPOs over history though an investor base mechanism. We analyse the US market and find evidence that issuers are less likely to underprice (or more likely to overprice) when the REIT is more diversified and therefore the investor base is bigger. Even if both types of diversification are significant, geographic diversification shows a stronger impact on initial returns than property-type diversification. Our argument and the deadweight cost theory complement each other as lower deadweight cost associated with the IPO weakens the influence of the geographic diversification on the initial returns of IPOs. Our results are robust to different measures of private market returns and time fixed effects. In the end we show that a Herfindahl index should be preferred as a measure of geographic diversification, which is not the case for property-type diversification.