The Impact of the Gross Book-to-market Ratio in REITs, and Corporate Reactions to PBGC Variable Rate Premium Increases
Type of DegreePhD Dissertation
MetadataShow full item record
The first two essays explore the usefulness of the gross book-to-market ratio in equity REITs. The first essay shows that the traditional calculation of the market-to-book (and book-to-market) ratio is flawed. We show that accumulated depreciation accounting leads to artificially high marketto-book ratios in REITs with highly depreciated assets, and this problem is exacerbated in REITs that utilize preferred equity. The three-tiered capital structure is employed by roughly half of the REITs, while only 12% of firms in a comparative non-REIT sample use preferred equity. We examine this effect in the context of capital structure studies and show that the gross book-to-market (where accumulated depreciation is added back to total assets and common equity) is superior to traditional metrics as a proxy for growth opportunities and stock valuations. We revisit a disagreement in the literature about the relationship between REIT market-to-book and leverage, and show additional support for pecking-order theory, while finding little evidence that market-timing drives capital structure decisions. We show that spurious relationships can yield T-statistics as high as 5 on the market-to-book, while the superior gross book-to-market is statistically insignificant. Our findings stress that future REIT researchers should use undepreciated balance sheet metrics, while also considering the three-tiered capital structure utilized by half of these firms. The second essay shows the gross book-to-market to be among the best predictors of aggregate REIT returns. The gross book-to-market is unavailable prior to 1995 due to data limitations, but we show that the ratio of paid-in capital to market capitalization (paid-to-market) contains similar information, because it excludes the effects of accumulated depreciation. Paid-to-market is the single best predictor of REIT returns across our full sample time period, with out-of-sample R2 statistics of 1.64% and 14.12% at the monthly and annual horizons, respectively. We show that REIT returns are predictable out-of-sample within an efficient market, as predictability stems from information connected to time varying risk aversion. Stripping accumulated depreciation from the balance sheet substantially improves forecasting models, which can provide investors with significant utility gains and larger Sharpe ratios. Outperforming the total returns from a buy-andhold strategy proves difficult though, as upside predictability is easier than downside predictability, especially during crises. The third essay explores a separate topic in US corporate pension plans. The Pension Benefit Guarantee Corporation (PBGC) is often criticized for its mispriced pension insurance formula, and for creating risk-shifting incentives that could encourage bad behavior by pension plan sponsors. Beginning in 2014, the Variable Rate Premium (VRP) charged to underfunded pension plans began rising. The VRP rate, set by Congress, effectively penalizes firms that underfund their pension plans. We overcome challenges in the data sources to take a comprehensive look at firm reactions to the threat of VRP payments. When faced with larger VRP payments, firms make larger contributions to their pension plans and allocate more of their plan assets to equities. The VRP rate is so high that in recent years, firms may have realized cost savings by funding their pension plans using borrowed funds. While anecdotal evidence exists for borrow-to-fund, we do not find a statistically significant relationship between debt issuances and pension contributions. Nor do we find that VRP rate hikes were associated with additional plan freezes. Overall, VRP rate hikes likely had both positive and negative effects on pension risk, as corporations made higher contributions while also increasing the riskiness of plan assets.