Essays on Higher Moments in Macroeconomics and Finance

by Harri Turunen

Institution: University of Helsinki
Department: Department of Economic and Political Studies; University of Oxford
Year: 2015
Keywords: taloustiede
Record ID: 1143585
Full text PDF: http://hdl.handle.net/10138/154151


I study the importance of variation in the higher moments of macroeconomic and financial quantities. The first essay considers the effects of uncertainty on the fiscal multiplier when the economy has hit the zero lower bound (ZLB) on the nominal rate in a relatively standard New Keynesian Dynamic Stochastic General Equilibrium (DSGE) model. As the ZLB is a very strong form of nonlinearity, the model is solved using a numerical method. Uncertainty in government spending and productivity are modeled as stochastic volatility. Confirming previous research (e.g. Christiano, Eichenbaum and Rebelo, 2011), the multiplier is found to be higher when at the bound, and the effects of volatility shocks are found to be noticeable (e.g. Basu and Bundick, 2015). Uncertainty is found to have an impact on the multiplier: when future spending is uncertain, the multiplier is high, but when future productivity is uncertain, the multiplier is low. The second essay studies whether or not DSGEs are able to generate simulated realizations with realistic third and fourth moments. Many time series in macroeconomics and finance exhibit either excess kurtosis or skewness or both. However, as was shown by Ascari et al (2013), standard DSGEs such as the neoclassical growth model or the model of Smets and Wouters (2007) are unable to produce realizations with reasonable moments, regardless of shock distribution or the order of the Taylor approximation applied. My results however indicate that this is mostly due to lack of nonlinearity in the models, since especially a model with a very strong form of nonlinearity, such as the ZLB, is able to generate non-Gaussian realizations. The third essay considers the pricing of macroeconomic risk. The theory of Merton (1973) implies that there can be other sources of priced risk than the risk associated with the return on the market portfolio and that an appropriate measure for sensitivity of a stock to this risk is the covariance of the return of that stock with the source of that risk. I apply the multivariate volatility model of Engle (2002) to estimate time-varying covariances of US stock portfolios with a variety of US macro time series. The finding is that inflation and unemployment are priced in the market and earn a negative premium, while the growth rate of industrial production and the Case-Shiller house price index are not priced. Useimmat ihmiset ovat riskinkaihtajia. He pyrkivät välttämään tilanteita, joissa esimerkiksi heidän tuleva palkkansa on epävarma. Samoin ihmisten hallinnoimat instituutiot, kuten yritykset, toimivat usein riskiä kaihtavasti: investointeja vältetään tilanteissa, joissa niiden tuottoon liittyy epävarmuutta. Taloustieteessä eräs tapa mallintaa riskiä tai epävarmuutta on olettaa, että mallien satunnaiskomponenttien korkeammat momentit, kuten varianssi ja kovarianssi, ovat vaihtelevia - esimerkiksi tutkija saattaisi olettaa, että ne ovat ajasta riippuvia ja itsekin satunnaisia. Näin voidaan tehdä, sillä vaihtelu varianssissa kykenee huomioimaan käsitteellisesti…