<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ilona Vänni | Macro Paper Warehouse</title><link>https://macropaperwarehouse.com/authors/ilona-vanni/</link><description>Ilona Vänni</description><generator>Hugo -- gohugo.io</generator><language>en-us</language><atom:link href="https://macropaperwarehouse.com/authors/ilona-vanni/index.xml" rel="self" type="application/rss+xml"/><item><title>Reading between the Lines — Uncovering Asymmetry in the Central Bank Loss Function</title><link>https://macropaperwarehouse.com/papers/reading-between-the-lines-uncovering-asymmetry-in-the-central-bank-loss-function/</link><guid>https://macropaperwarehouse.com/papers/reading-between-the-lines-uncovering-asymmetry-in-the-central-bank-loss-function/</guid><description>&lt;p&gt;This paper estimates the European Central Bank&amp;rsquo;s loss function directly, rather than inferring its shape from an estimated reaction function, and asks whether the Bank has disliked inflation above its target more than inflation below it. The method proxies the Bank&amp;rsquo;s realised loss by the tone (net negativity) of the Governing Council&amp;rsquo;s press conference introductory statements &amp;ndash; measured both with a purpose-built ECB-specific adaptation of the Loughran-McDonald finance dictionary and with the FinBERT language model &amp;ndash; and regresses that tone on the real-time inflation estimate and other real-time indicators available to decision-makers at each meeting. On 238 Governing Council meetings from January 1999 to June 2021, the authors report strong and robust evidence of asymmetry: the slope of the estimated piecewise-linear loss function is roughly three times steeper when inflation exceeds the target than when it falls short, though some specifications put the ratio at four, five or even six times, and in a few the below-target branch is essentially flat. They present this as a significant departure from the quadratic, symmetric loss function typically assumed in macro models, and note that their direct approach avoids having to assume a structure for the private sector or that policy is unconstrained by the effective lower bound &amp;ndash; assumptions that reaction-function studies, including Surico (2007b), which found the ECB&amp;rsquo;s price-stability objective symmetric, must make. The evidence is an association between communication tone and real-time inflation over this sample and this institution; the authors also estimate a de facto inflation target below 2%, with the level depending on which tone measure is used.&lt;/p&gt;</description></item></channel></rss>