Essays on inflation, expectations and central bank communication
Date
2024-01-25
Authors
Gabrielyan, Diana
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Praeguse ja oodatava inflatsioonidünaamika hea mõistmine on keskpankade jaoks oluline teadlike otsuste langetamiseks. Ootused muutuvad eriti oluliseks äärmuslike majandussündmuste ajal või siis, kui majandusmaastik on ebakindel. Käesolevas doktoritöös esitatud uurimustes on inflatsioon ja keskpankade teabevahetus kas otseselt või kaudselt ühine nimetaja ning neis rõhutatakse ootuste rolli inflatsiooni mõistmisel ja prognoosimisel ning asjakohase teabevahetuse tagamisel. Seega on käesoleva töö põhieesmärk inflatsioonieesmärgistamise raamistikus analüüsida inflatsiooniootusi, keskpanga sõnumeid, leibkondade majanduslikke valikuid ja nende kolme nähtuse omavahelisi seoseid.
Esimene mõõde, mille kaudu käesolev väitekiri panustab majandusteaduslikku kirjandusse, on erinevate masinõppe ja traditsiooniliste ökonomeetriliste meetodite võrdlus. Uudiste andmete kasutamine on käesoleva töö teine panus, mis võimaldab luua asjakohaseid prognoose ja testida majanduslikke seoseid. Viimane käesoleva töö oluline panus on heita valgust meedia rollile majandusagentide ootuste kujunemise protsessis.
Käesoleva väitekirja I ja II uuringus prognoositakse inflatsiooni vastavalt Phillipsi kõvera ja masinõppe mudeli abil, kusjuures viimane tugineb inflatsiooniootuste andmetele. Uuring III esitab väljakutse küsitlusel põhinevate inflatsiooniootuste mõõdikute kasutamisele ja pakub välja uue, mitmedimensioonilise ja potentsiaalselt reaalajas toimiva inflatsiooniootuste mõõdiku, mis kajastab kodumajapidamiste inforuumi majanduse olukorra kohta. Uuring IV annab oma panuse keskpankade teabevahetuse mõju analüüsi käsitlevase kirjandusse, uurides empiiriliselt Euroopa Keskpanga kommunikatsiooni reaktsioonifunktsiooni.
Having a good understanding of current and future inflation dynamics is important for central banks to make informed decisions. Expectations become especially relevant during extreme economic events or when the economic landscape is full of uncertainty. The studies in this thesis have inflation and central bank communication as the common denominator, either directly or indirectly and emphasise the role of expectations in understanding and forecasting inflation and ensuring relevant communication. Therefore, this work's aim is to provide evidence that the key to the inflation targeting area of economic policy is in understanding and forecasting inflation and inflation expectations and ensuring relevant communication. The first dimension through which this thesis contributes to economic literature is through experiments comparing various machine learning and traditional econometric methods. Employing news data is another novelty and the second contribution of this thesis, allowing for state-of-the-art forecasts and test economic relationships. The last major contribution of this thesis is casting light on the role of media in the expectation formation process of economic agents. In studies I and II of this thesis, inflation is forecasted using the Phillips curve and a machine learning model, respectively, with the latter building on the data of inflation expectations. Study III challenges the usage of survey-based measures and instead proposes new, high-dimensional and potentially real-time measure of inflation expectations that is able to capture household's true information space about the economy. Study IV contributes to the literature on analysing the effects of central bank communication through an empirical study on the communication reaction function of the European Central Bank.
Having a good understanding of current and future inflation dynamics is important for central banks to make informed decisions. Expectations become especially relevant during extreme economic events or when the economic landscape is full of uncertainty. The studies in this thesis have inflation and central bank communication as the common denominator, either directly or indirectly and emphasise the role of expectations in understanding and forecasting inflation and ensuring relevant communication. Therefore, this work's aim is to provide evidence that the key to the inflation targeting area of economic policy is in understanding and forecasting inflation and inflation expectations and ensuring relevant communication. The first dimension through which this thesis contributes to economic literature is through experiments comparing various machine learning and traditional econometric methods. Employing news data is another novelty and the second contribution of this thesis, allowing for state-of-the-art forecasts and test economic relationships. The last major contribution of this thesis is casting light on the role of media in the expectation formation process of economic agents. In studies I and II of this thesis, inflation is forecasted using the Phillips curve and a machine learning model, respectively, with the latter building on the data of inflation expectations. Study III challenges the usage of survey-based measures and instead proposes new, high-dimensional and potentially real-time measure of inflation expectations that is able to capture household's true information space about the economy. Study IV contributes to the literature on analysing the effects of central bank communication through an empirical study on the communication reaction function of the European Central Bank.
Description
Väitekirja elektrooniline versioon ei sisalda publikatsioone
Keywords
keskpangandus, kommunikatsioon, rahapoliitika, inflatsioon, prognoosimine (majandus), majandusmudelid, central banking, communication, monetary policy, inflation, economic forecasting, economic models