Publicação
A deep learning test of the martingale difference hypothesis
| dc.contributor.author | Bastos, João A. | |
| dc.date.accessioned | 2025-03-27T08:35:42Z | |
| dc.date.available | 2025-03-27T08:35:42Z | |
| dc.date.issued | 2025-03 | |
| dc.description.abstract | A deep learning binary classifier is proposed to test if asset returns follow martingale difference sequences. The Neyman-Pearson classification paradigm is applied to control the type I error of the test. In Monte Carlo simulations, I find that this approach has better power properties than variance ratio and portmanteau tests against several alternative processes. I apply this procedure to a large set of exchange rate returns and find that it detects several potential deviations from the martingale difference hypothesis that the conventional statistical tests fail to capture. | pt_PT |
| dc.description.version | info:eu-repo/semantics/publishedVersion | pt_PT |
| dc.identifier.citation | Bastos João A. .(2025). “A deep learning test of the martingale difference hypothesis”. REM Working paper series, nº 0374/2025 | pt_PT |
| dc.identifier.issn | 2184-108X | |
| dc.identifier.uri | http://hdl.handle.net/10400.5/99770 | |
| dc.language.iso | eng | pt_PT |
| dc.publisher | ISEG - REM (Research in Economics and Mathematics) | pt_PT |
| dc.relation.ispartofseries | REM Working paper series;nº 0374/2025 | |
| dc.subject | Martingale Difference Hypothesis | pt_PT |
| dc.subject | Convolutional Network | pt_PT |
| dc.subject | Variance Ratio Test | pt_PT |
| dc.subject | Portmanteau Test; | pt_PT |
| dc.subject | Exchange Rates | pt_PT |
| dc.title | A deep learning test of the martingale difference hypothesis | pt_PT |
| dc.type | working paper | |
| dspace.entity.type | Publication | |
| rcaap.rights | openAccess | pt_PT |
| rcaap.type | workingPaper | pt_PT |