Previsão de Preços Máximos e Mínimos para Green REITs
Previsão de Preços Máximos e Mínimos para Green REITs
-
DOI: https://doi.org/10.22533/at.ed.42624050222
-
Palavras-chave: REDES NEURAIS; PREVISÃO DE PREÇOS; GREEN REITS.
-
Keywords: NEURAL NETWORKS; PRICE FORECASTING; GREEN REITS.
-
Abstract: THE ACADEMY POSITS TWO RELEVANT HYPOTHESES ABOUT THE FINANCIAL MARKET: THE IMPOSSIBILITY OF PREDICTING ASSET PRICES AND THE POSSIBILITY OF PREDICTION. THEREFORE, PREDICTIVE METHODS WERE DEVELOPED TO HELP PEOPLE IN THEIR DECISIONS, SEEKING TO MAKE THE MOST PROFIT POSSIBLE WITH MINIMAL LOSSES. ARTIFICIAL NEURAL NETWORKS (ANNS) HAVE STOOD OUT IN THIS AREA, DUE TO THEIR ABILITY TO HANDLE A WIDE RANGE OF KNOWLEDGE AND PREDICT PRICES IN THE FINANCIAL MARKET. IN ADDITION, SUSTAINABILITY IN INVESTMENTS HAS GAINED IMPORTANCE, TAKING INTO ACCOUNT THE ENVIRONMENTAL AND SOCIAL IMPACTS OF COMPANIES. IN THIS WAY, INVESTORS SEEK PROFITS IN THE FINANCIAL REAL ESTATE MARKET, CONSIDERING SUSTAINABILITY. IN THIS CONTEXT, THIS STUDY USES ANN TO PREDICT MAXIMUM AND MINIMUM PRICES OF REITS CONSIDERED SUSTAINABLE DUE TO THEIR ESG CERTIFICATIONS AND ESTIMATE WHICH HYPOTHESIS FITS. DATA FROM 18 GREEN REITS FROM 2018 TO 2022 WERE ANALYZED, USING THE LONG SHORT TERM MEMORY (LSTM) AND RANDOM WALK (RW) MODELS FOR PREDICTIONS. THE RESULTS SHOWED THAT THE RW MODEL PERFORMED BETTER IN PRICE FORECASTING, COMPARED TO LSTM. THESE RESULTS CORROBORATED THE EFFICIENT MARKET HYPOTHESIS, INDICATING THAT THE PRICES OF GREEN REITS ALREADY REFLECT THE INFORMATION AVAILABLE IN THE MARKET.
- Hugo Martins Teixeira
- Daniel Vitor Tartari Garruti
- Flávio Luiz de Moraes Barboza