Essays in financial economics
Abstract/Contents
- Abstract
- This dissertation offers contributions to frontier topics in the financial economics literature. Its three chapters are based on three papers that I developed while a student at the Stanford Graduate School of Business. The first two chapters are focused on the characteristics of risk and return in residential real estate markets. Housing has emerged as an important topic of study in financial economics. This is because most households are homeowners, and their house constitutes a large fraction of their wealth. Moreover, housing is a highly levered investment frequently used as collateral by homeowners, and fluctuations in house prices affect households' financial constraints. A key aspect of housing is that investments are lumpy. Households buy a specific, indivisible asset. Moreover, households own undiversified portfolios of houses, which in most case consists of an individual property. The first chapter of the dissertation studies asset-level, or idiosyncratic, risk in housing. The empirical evidence in the chapter is based on a unique dataset that I built with Victor Westrupp, from the merge of information on house prices and remodeling expenses at the level of individual family homes in the main metropolitan areas of California. Relatively little research has been devoted to the topic of idiosyncratic risk in housing markets. The prevailing view in the literature is that idiosyncratic risk scales with the holding period of a house, and constitutes more than 50% of the standard deviation of capital gains earned by individual homes. The results in chapter one challenge this view, showing that the standard deviation of idiosyncratic shocks does not depend on the holding period. Since the variance of market-wide fluctuations is close to scaling with time, the idiosyncratic component of house capital gains determines a different fraction of total house price standard deviation at different holding periods. This fraction is more than 50% after one year, but less than 20% after five years. Thus, I find that idiosyncratic risk behaves as a one-time shock, taking place at the time a house sale occurs. Housing is an illiquid asset class, traded other-the-counter. An interpretation of my finding is that idiosyncratic risk in this market is indeed driven by illiquidity, and determined by price uncertainty at the time a house sale occurs. In the second part of the chapter I develop and calibrate to the data a quantitative portfolio model, which I use to compute the value of insurance against the idiosyncratic component of house-level risk. I show that insurance premiums crucially depend on household mobility. The second chapter of the dissertation is the product of joint work with Victor Westrupp. I have mentioned that housing is an illiquid asset class, traded over-the-counter. Most over-the-counter markets are characterized by the presence of asset dealers, acting as middlemen. These intermediaries improve asset allocations and provide liquidity by buying and re-selling assets to counter-parties with different valuations. In doing so, they extract economic rents. Everyday experience and empirical studies suggest that most intermediation in housing markets is carried out by specialized brokers (real estate agents). Surprisingly, middlemen do not appear to play a major role in housing markets. In the chapter, we develop an empirical strategy to identify professional middlemen. We then provide empirical evidence on the volume of transactions undertaken by middlemen in the main housing markets of California. Their activity is not negligible. They intermediate between 2% and 7% of all house sales in each year from 1998 to 2012, for a total investment in property acquisitions above thirty billion dollars over the entire period. However, this is still a relatively small fraction of the total volume of house transaction. A rationale for the limited involvement of middlemen in housing markets could be that they are not able to extract substantial economic rents from their activity. We, therefore, use our data to measure the abnormal capital gains earned by middlemen with respect to local market fluctuations and appreciation of comparable houses. A key strength of our dataset is that it captures not only the prices at which individual properties were bought and sold, but also intermediate investments in remodeling. We find that middlemen are on average able to extract abnormal capital gains with respect to the local markets where they are trading. However, it is not clear that these average abnormal capital gains can be interpreted as economic rents. A large intermediary, carrying out all middlemen transactions in the market, would earn the average abnormal rent and diversify away transaction level risk. In the data, the industrial organization of middlemen is atomistic, and individual middlemen portfolios are poorly diversified. Most middlemen are processing not more than three house transactions in the same year. Risk at the level of the individual transactions is substantial, shadowing the average abnormal returns. Thus, while on average profitable, middlemen trading activity in housing markets is extremely risky. We believe that our contribution opens an interesting debate on the frictions constraining the scale of middlemen operations, and on the potential rents that could be extracted by a large asset dealer in housing markets. The third and final chapter is based on joint work with Kristoffer Laursen and Kenneth Singleton, and focuses on learning and its effects on asset prices. Most work in empirical asset pricing measures risk compensation assuming a rational expectations framework with full information. This implies that the representative agent or the agents in the economy know the dynamics of the key state variables, and adjust their behavior to price financial assets accordingly. However, this is a very demanding set of assumptions. In the data, economic agents are unlikely to be endowed with a full understanding of economic dynamics, and are instead likely to make inference based on the available historical information. In other words, agents are likely to be learning the dynamics of the economy while pricing assets. The chapter focuses on the implications of learning for US Treasury bonds. We extend the Gaussian dynamic term structure framework, allowing for real-time Bayesian inference of the pricing factor dynamics. We bring our framework to the data, and find that real-time inference delivers different physical dynamics for yields when compared to traditional full-sample estimates based on full information rational expectations. However, the risk neutral dynamics, or the cross-sectional relationships between yields, are extremely stable over time. When analyzing out of sample performance, we show that allowing for time-varying parameters in the physical dynamics delivers more accurate forecasts of yields. Our set of state variables initially consists only of yields portfolios. We then extend it to include measures of disagreement among professional forecasters and macro-economic fundamentals. We show that including disagreement among forecasters improves predictive performance, and we analyze the channels through which disagreement affects the forecasts generated by the model.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2017 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Giacoletti, Marco |
|---|---|
| Associated with | Stanford University, Graduate School of Business. |
| Primary advisor | Binsbergen, Jules H. van |
| Primary advisor | Singleton, Kenneth J |
| Thesis advisor | Binsbergen, Jules H. van |
| Thesis advisor | Singleton, Kenneth J |
| Thesis advisor | Grenadier, Steven R |
| Advisor | Grenadier, Steven R |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Marco Giacoletti. |
|---|---|
| Note | Submitted to the Graduate School of Business. |
| Thesis | Thesis (Ph.D.)--Stanford University, 2017. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2017 by Marco Giacoletti
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...