Yang Bai

Should households buy their homes? Contrary to popular personal-finance and academic expert advice, our block-bootstrap life-cycle simulation suggests they should. Using 150 years of data across 16 countries, we find that homeownership raises wealth and consumption-equivalent welfare relative to saving-rate-matched benchmark strategies investing solely in financial assets. The gains come from lower portfolio risk, rent-risk hedging, access to home equity late in life for enhanced retirement consumption, and higher bequests. The gains vary with income, house-price conditions at purchase, and mortgage rates. Mortgages enable early homeownership access at the cost of lower liquidity and consumption-equivalent welfare.

Using labor skill demand disclosed in job postings as a proxy for firms' growth strategies, we find that similar growth strategies increase the likelihood of two firms merging. In particular, a firm is more likely to become a target as its labor skill demand becomes more similar to that of its potential acquirer. Similar growth strategies ameliorate post-merger integration challenges in facilitating merger deals. Following the merger, the combined firm continues hiring the same skills, consistent with the growth strategy persisting. These types of mergers experience more synergies and superior operating performance.

Classification outperforms regression across matched machine learning models in portfolio construction. A stacking ensemble of gradient boosted trees, random forest, and neural network yields a value-weighted annualized Sharpe ratio of 2.08 for classification and 1.39 for regression. This outperformance strengthens with class granularity and persists across subsamples and after transaction costs. Spanning tests show that classification retains economically large alphas after we control for regression, whereas regression alphas shrink substantially once we control for classification. These results indicate that classification extracts more return information than matched regression. Our diagnostics trace classification's advantage to more precise separation of return deciles.