Professional stuff

This is the second chapter of my Ph.D. dissertation, co-authored with my advisor Nobel Laureate Rob Engle. It's a purely theoretical econometrics paper that extended Rob's univariate work (predicting volatility of a single asset) to a multivariate context (predicting volatility and interrelationships of several assets).

Interesting anecdote: It took about seven years to get this paper published because it took that long for me to figure out how to fix an error that the referee found. (The referee was very good.) While awaiting publication, the paper became notorious for being one of the most heavily cited unpublished papers in the field. I was almost sad to see it published.

This paper, coauthored with classmates Tim Bollerslev and Ray Chou, surveyed the volatility modeling literature.

Interesting anecdote: As of the late 1990s and early 2000s, this was one of the 100 most influential papers in the history of the social sciences (as measured by citations). It's probably since been surpassed by many other papers.

Interesting anecdote: The French translation had an additional coauthor. We included him on the translation as a reward for helping get this paper off the ground.

Interesting anecdote: This paper was reprinted twice. Once in Volatility: New Techniques for Pricing Derivatives and Managing Financial Portfolios (edited by Robert Jarrow), Risk Publications 1998. And the second in Recent Developments in Time Series (edited by Paul Newbold and Stephen Leybourne): Edward Elgar Publishing, 2003

This paper, coauthored with my boss Richard Grinold, gives a framework for forecasting long-run stock market returns (the "equity risk premium"). The paper then applies the framework to show that equities should outperform bonds by about 3% to 3½% over the long run.

Interesting anecdote: This paper is easily my most-read and most-studied paper. Every CFA candidate from about 2002 to 2016 (and maybe still today...I don't know) had to study this paper, because it was on the CFA level 3 exam virtually (if not literally) every year. This means that about 100,000 finance professionals since 2002 had to learn its key concepts in order to get their professional CFA certification, and explains why there are a number of web pages that help students understand this model.

Interesting anecdote: Someone created a short Wikipedia article on the model. The link to this article is included in the "Press releases, articles and the like" panel to the left.

Being able to forecast the future value of a currency (or commodity or stock or any other security) would be the Holy Grail in investing. Therefore, hundreds of papers were written to test whether information from the futures markets provides accurate forecasts of currency prices (or commodity prices or stock prices). This was a popular research topic both because it's pretty important and because everyone was getting different results, depending on how they conducted their tests. Robin Brenner and I applied the then-new concept of Cointegration to show theoretically why everyone got the results they were getting.

Interesting anecdote: This paper won a Financial Management Association "best paper" award when it was written.

Interesting anecdote: One of my Ph.D. advisors, Clive Granger, eventually won the Nobel Prize for developing the concept of Cointegration. This paper demonstrated just how relevant Clive's work was.

Stock market risk changes over time, both because the risk of each underlying stock changes over time, and because the stocks sometimes move together and sometimes move apart. For example, Google and Facebook risk are each time-varying, and sometimes one is zigging while the other is zagging but sometimes they both zig and zag together. Modeling these dynamic behaviors is so important that many different ways to do it were developed (one in my Ph.D. dissertation). In geek-speak, there were eventually several ways to parameterize multivariate volatility models, each with slightly different properties. The papers below, written with classmate Victor Ng, developed a novel way to compare and contrast the models, showing that each is appropriate for different purposes.

Interesting anecdote: The JASA paper was reprinted in the book Volatility: New Estimation Techniques for Pricing Derivatives (ed. by Robert Jarrow), Risk Publications 1998

I wrote several papers that were all kind of the same. They all answered questions like, "If I own a particular asset (say, the dollar), how much of another asset (say, dollar futures) should I own in order to minimize the riskiness of my portfolio?" In geek-speak, these papers all used multivariate volatility models to compute optimal financial hedge ratios. Here are some of those papers. For the most part, they differ in that they are each applied to different markets. But the questions addressed and the tools used are pretty similar.

Interesting anecdote: The fixed income paper...I didn't write a word of it. I didn't perform a single piece of empirical analysis. I never even read the paper. And I don't think I ever met my coauthors. I hope it's a good paper.

The euro was introduced in 1999. In this 1998 paper, BGI colleague Kevin Coldiron and I analyzed how the euro's introduction should impact the way that investors decide which countries and asset classes they invest in.

Interesting anecdote: This paper won the Institutional Investor Award for best international paper in any Institutional Investor journal in 1999.

The options markets give you one forecast of volatility, and time series models (like GARCH) give you another. My World Bank coauthor (Stijn), one of my Ph.D. students (Kevin) and I showed that you get better forecasts by blending the two techniques than by using either of them stand-alone.

Interesting anecdote: This paper was reprinted in Economic Forecasting, edited by Terence C. Mills, Edward Elgar Publishing, 1999

In the 1990s, research on interest rate movements assumed that risk (or volatility) was a function of only the level of the interest rate. High interest rates meant high risk. By allowing for richer volatility dynamics, University of Arizona colleagues Robin Brenner, Rich Harjes and I were able to model interest rate movements suprisingly more effectively than the existing literature. We were also able to show that the level of the interest rate didn't have as big an impact on risk as the existing literature assumed.

Big insurance companies should be better than small ones at screening out bad drivers, and therefore should insure better risks and have fewer claims as a result. My undergrad professor, Doug West, and I (more him than me) developed a theoretical model for this observation, and confirmed it empirically.

Interesting anecdote: This is one of two papers I wrote with my undergraduate professor, Doug West. Truth is that for this paper, I was just his research assistant while I was home for summers from grad school. He felt so guilty about how hard I worked on the paper that he rewarded me by naming me a coauthor.

You ever notice that auto dealers choose to locate next to other auto dealers? Ditto for furniture stores, etc. Why would they voluntarily choose to park right beside a competitor? In the 1930s an economist presented a theory for why this happens, which came to be known as the "Christaller theory of central places." This paper, coauthored with two of my undergrad professors, was the first to evaluate his hypothesis empirically.

Interesting anecdote: This paper was published while I was an undergraduate, in one of the top journals in economics. It probably helped me get into grad school at UC San Diego. (I could never get in today.)

If a country needs to go into debt, which country should it borrow from? World Bank economist Stijn Claessens and I demonstrate an innovative way to make that decision, under the (admittedly unrealistic) assumption that a country cares deeply about the riskiness of that debt portfolio.

Interesting anecdote: This was chapter three of my Ph.D. dissertation.

Interesting anecdote: This paper was written over the summer between finishing grad school and starting at Univ of Arizona, while I was consulting for the World Bank.

A simple paper in a trade journal that summarizes different ways to forecast risk, and highlights how each of them are the same and how they differ. It was relatively insightful, given its simplicity.

Interesting anecdote: The editor asked me to write this paper after he saw me give a presentation on this topic at practitioner conference.

Most economists argue that flexible exchange rates are good, in that they give a nation more control over its own economic policy. But exchange rate risk could partially offset those benefits by adversely impacting foreign trade. In this paper, Bill Lastrapes and I showed that trade was indeed impacted by currency risk.