DECEMBER 2015 8 Lars Peter Hansen In March 1979, Lars Peter Hansen, a young Ph.D. fresh out of the University of Minnesota, submitted a paper to the prestigious Economet- rica. It described a statistical methodology that, in its final form, would allow economists to draw strong conclusions from models that weren’t completely specified (that is, not all variables, relationships or assump- tions were included or precisely defined). This “generalized method of moments” would give econometricians the ability to appraise alternative theories and investigate important economic phenomena without fully developing each of their elements. Researchers could rely on the most powerful explanatory variables and dispense with unnecessary assumptions. “GMM allows you to ‘do some- thing without having to do everything simultaneously,’” Hansen explains. But the GMM—abstract and mathematically challenging—was not immediately embraced by the field. (Indeed, Hansen’s initial draft was rejected by Econometrica, spurring him to refine and generalize his argument.) Hansen and his colleagues persevered, demonstrating the methodology’s power and range by applying it to exchange rates, as- set pricing models and rational expectations theory. These and other examples gradually convinced economists of its utility and, with time, GMM became the gold standard. In 2013, Hansen received the Nobel Prize in economic sciences for his methodology, specifically in reference to its ability to evaluate asset pricing models. Hansen continues to study asset prices, focusing on linkages be- tween financial markets and the broader macroeconomy. Recent work looks at uncertainty and risk tolerance in asset pricing behavior; he’s also developed methods to analyze and account for the uncertainty of the households and businesses that populate economic models, and also for the uncertainty that econometricians have about the adequacy of their models. Related research examines policymaking under uncertainty. At the top of his field, there are few major awards that he hasn’t received, but he’s especially pleased with his role in developing future generations of gifted economists. He is a major force behind the Becker Friedman Institute at the University of Chicago, a multidisciplinary re- search center that supports young scholars and others in economics, law, public policy and business. And he devotes a full page of his CV to a list of the Ph.D. students he’s advised. “I’m proud of my research accomplishments,” he observes in this Region conversation, “but I am also proud that I am able to associate with so many very good graduate students.” PHOTOS BY PETER TENZER BY PHOTOS 9 DECEMBER 2015 The Region THE GENERALIZED METHOD OF There’s a long history in formal GMM AND ITS APPLICATION MOMENTS AND ITS APPLICATION econometrics behind this approach. Some of my research had an important Is there a way that you Region: Let’s start, if we could, with as- antecedent, namely, the work of Denis set prices and the generalized method Sargan back in the late 1950s. I was also can study the connections of moments, known as GMM. When very heavily influenced by lectures Chris between asset prices and you were honored with the Nobel award Sims was giving back when I was a grad- macroeconomic outcomes in 2013 for your work on the GMM, the uate student. While developing a formal committee said that the GMM is “par- justification for an econometric method, without necessarily getting ticularly well suited to testing rational what excited me and helped my research all the other details exactly theories of asset prices.” Could you give gain some traction were the empirical correct? My aim was to a brief description of the method and its applications I and others came up with. importance to later work? develop methods that allow Region: What do you think was achieved for initial investigation of Hansen: Sure. I like to think about this by applying these methods, in your own linkages without requiring as the following: Economists can build work and in work by others? a fully fleshed-out model. a full-scale model of a macroeconomy with many different equations, where Hansen: My initial applications included they map out a really rich structure of fi- analyses of forward exchange markets nancial markets and the macroeconomy. with Bob Hodrick and investigation of The question is: Is there a way that you pricing a cross section of asset returns Hansen: Researchers have indeed fo- can study the connections between as- with Ken Singleton. The Hodrick paper cused on the forward-looking nature set prices and macroeconomic outcomes documented empirical challenges that of asset prices. Prices today depend on without necessarily getting all the other have altered how researchers model ex- what people think is going to happen in details exactly correct? For instance, change rate determination. The Single- the future. For instance, I may have ob- an econometrician may wish to avoid ton research exposed gaps in the exist- servations on the market values of some specifying the precise information used ing macroeconomic models in terms of type of underlying dividend or cash flow by investors or detailing all of the ingre- their implications for asset pricing. This process. Or I may make assessments of dients that govern the evolution of the work in turn encouraged Ravi Jaganna- the market value of education, depend- macroeconomy. My aim was to develop than, John Cochrane and me to char- ing in part on guesses as to what the sal- methods that allow for initial investiga- acterize asset-pricing puzzles in more ary will be for different levels of educa- tion of linkages without requiring a fully general terms. Scott Richard and I be- tion going forward. fleshed-out model. gan to explore more abstract economic Asset prices, broadly conceived, reflect formulations of so-called stochastic dis- people’s beliefs about the future, but they Region: Which, of course, is always a count factor models, where such factors also reflect how people respond to uncer- challenge at the initial stages of model simultaneously discount the future and tainty. What makes people cautious and development. adjust for risk. The empirical challenge what makes them bold? From this van- to model builders is to understand why tage point, the challenge of an empirical Hansen: Right. The point is not even to the implied risk prices are large and asset-pricing model is to disentangle the make a pretense of believing that every- fluctuate over time in interesting ways. contributions coming from beliefs versus thing is fully specified. The econome- It’s been rewarding to observe the sub- those coming from concerns about un- trician’s jargon of a “partially specified sequent model extensions and refine- certainty. And I’ve been keenly interested model” applies where one tries to study ments motivated by empirical chal- in methods that allow us to try to disen- linkages without having to spell out all lenges. tangle these two different impacts. the various different details. As you build This aim to extract information from models to use for policy purposes, you ASSET PRICES AND THE forward-looking prices is not just an aca- do have to specify many of those details. MACROECONOMY demic exercise; it’s on the radar screen of But as an initial step, it’s nice to be able to policymakers as well. For instance, those make assessments without all that detail. Region: Asset prices are known to be engaged in monetary policy look to fi- I like to say that the GMM allows you to foward-looking and, as such, they contain nancial markets to try to see where the “do something without having to do ev- information about private sector beliefs. private sector thinks the macroeconomy erything simultaneously.” How does your work bear on this topic? is headed. So it’s very interesting to ask DECEMBER 2015 10 The Region what types of restrictions we impose on ASSET PRICES AND As uncertainty compounds over time, models that allow us to disentangle the THE MACROECONOMY random outcomes that don’t originally contributions of private sector investor look to have big consequences can grow beliefs from adjustments made for the The challenge of an empirical in magnitude as they play out. This re- uncertainties that they perceive. And asset-pricing model is to search provides a way to quantify such this has been a long-standing interest of disentangle the contributions phenomena—to actually measure this mine. compounding uncertainty and see how coming from beliefs versus it might be reflected in the behavior of Region: What methodological approach- those coming from concerns asset prices. es have you recently developed to further about uncertainty. And I’ve Just as growth rate uncertainties com- the understanding of linkages between pound over time, so do the market com- asset pricing and the macroeconomy? been keenly interested in pensations for those uncertainties. Alter- methods that allow us to native economic models have interesting Hansen: The empirical finance literature try to disentangle these two things to say about how compensations has focused on risk-return trade-offs differ across alternative investment ho- over a quarter or a month, or sometimes different impacts. rizons. even shorter horizons. But pricing im- This aim to extract plications extend over much longer ho- information from forward- rizons, and that’s the part that has been UNCERTAINTY AND less well-characterized, although it’s im- looking prices is not just an MARKET RETURNS portant for understanding investment academic exercise; it’s on the radar decisions and has macroeconomic im- screen of policymakers as well. Region: To measure the impact of un- plications. certainty and understand how it com- I’ve been working with collaborators, pounds over time, how do you capture including José Scheinkman and Jarda the link between uncertainty and finan- Borovička, to expand the existing tool cial market returns? kit for analyzing this problem.