Via G1:
Carole Ghosn, mulher de Carlos Ghosn, está de volta ao Japão, onde deve ser ouvida pela Justiça na investigação sobre as suspeitas de fraudes que teriam sido cometidas pelo ex-presidente da aliança Nissan-Renault-Mitsubishi - informou seu advogado, François Zimeray, nesta quarta-feira (10).
Ela havia deixado o Japão no último fim de semana, depois da recente detenção do marido, afirmando que "se sentiu em perigo". Segundo diferentes jornais japoneses, porém, a Justiça queria interrogá-la.
[...]
De acordo com a agência de notícias Kyodo, o escritório do promotor de Tóquio suspeita que somas de dinheiro desviado teriam passado por uma empresa administrada por ela.
Carlos Ghosn foi preso em 4 de abril em Tóquio por novas suspeitas de crimes financeiros, um mês depois de ser libertado sob fiança.
Sua mulher, que estava presente quando foi preso, disse que voltou para a França, apesar de seu passaporte libanês ter sido confiscado pela polícia japonesa. Carole disse que usou seu outro passaporte, americano.
11 abril 2019
TPM
Segundo o Google Analytics, 56% do nosso público é composto por mulheres. Esta é uma postagem "off topic", mas achei válida por demonstrar de uma forma bem sintetizada algo que acontece com a maioria de nós a cada ciclo menstrual. Haja fôlego!
09 abril 2019
Austeridade Fiscal: quando funciona?
[...]
At the moment the consensus among economists seems to be that governments need to approach any fiscal consolidation with caution since they risk disrupting economic growth or making a recession worse, and should take care to balance spending cuts with tax increases. This book argues this approach is a mistake.The authors think that the evidence of various austerity programmes over the last four decades suggests that the negative impact of economic austerity on growth has been exaggerated. While tax hikes can indeed be bad for the economy, cuts in public spending can actually boost economic growth, by increasing confidence among investors that spending is under control. What’s more, austerity is far from the “political kiss of death” that many people believe it to be – voters are willing to back a government with the “courage” to make difficult sacrifices to get a country’s house in order.
[...]
Fonte: aqui
08 abril 2019
Crise da Replicação pode ser bom para a ciência
Há alguns anos um erro em uma planilha Excel foi descoberto graças a replicação de um trabalho.
Um texto do The Conversation afirma que a “crise da replicação” é boa para ciência. E apresenta diversos argumentos onde o esforço de replicação pode ajudar no progresso científico.
Alguns periódicos agora exigem que os cientistas disponibilizem seus dados e códigos de computador, o que torna mais provável que outros detectem erros em seu trabalho. Além do mais, alguns cientistas podem agora "pré-registrar" suas hipóteses antes de iniciar seu estudo - o equivalente a chamar sua chance antes de começar.
Combinado com o compartilhamento aberto de materiais e dados, o pré-registro melhora a transparência e a reprodutibilidade da ciência, assegurando que uma fração menor de estudos futuros não seja replicada.
Além disto, a discussão pode fazer com que os pesquisadores tenham uma visão melhor sobre os métodos usados. Recentemente, um grupo de cientistas defendeu o abandono da significância estatística.
A incrível possibilidade de que uma grande fração das descobertas de pesquisas publicadas possa ser apenas casual é exatamente o motivo pelo qual as pessoas falam da crise de replicação. Mas não é realmente uma crise científica, porque a consciência está trazendo melhorias na prática de pesquisa, novos entendimentos sobre inferência estatística e uma apreciação de que descobertas isoladas devem ser interpretadas como parte de um padrão maior.
Em vez de minar a ciência, sinto que isso está reafirmando as melhores práticas do método científico.
Estimativas Contábeis via Machine Learning
Resumo:
The progressive migration of financial reporting from the cost basis to fair values has created the need for a wide range of estimates to be included into the process. These values are in general clearly an improvement on obsolete cost based measurements but also create the possibility of self-serving manipulation of critical financial information. This study shows that critical estimates in one particular industry are quite weak and finds that the usage of a set of different algorithms including the ones used in machine learning to predict estimates produces far superior results.
Fonte: Ding, Kexing and Lev, Baruch Itamar and Peng, Xuan and Sun, Ting and Vasarhelyi, Miklos A., On Creating Accounting Estimates using Machine Learning (September 21, 2018). Available at SSRN: https://ssrn.com/abstract=3253220 or http://dx.doi.org/10.2139/ssrn.3253220
[...]
To prove this I teamed up with a great group of ML experts at Rutgers University (Miklos Vasarhelyi, Kexing Ding, Xuan Peng, and Ting Sun). In the first research on the use of ML in accounting, we focus on property & casualty insurance companies. The reason: a major cost item of these companies is the “future loss estimate.” Explanation: For insurance policies written (earned) in a given year, the major cost item are the payments made on insurance claims (car accidents, home damage, etc.). But that’s not the end of the story. People can, and often do claim damages years later, like long-term effects of concussions. So, in addition to claims paid during the insured year, there will be claims to be paid in the future years (“insurance long tails”). Thus, a proper measurement of insurance companies’ earnings requires accounting for both claims paid during the year, and an estimate of future claims― “future loss estimates.” These are generally large estimates having a material effect on insurance companies’ reported earnings. We focus on these estimates in our ML research.
Summarizing, this is the first demonstration of the use of Machine Learning in improving the quality and reliability of financial information. In addition to insurance companies’ loss estimates, ML can be applied to a wide range of managerial accounting estimates, like the bad debt reserve, future payments on warranties, pension estimates, etc., etc. Very promising indeed, leading me to the final figure:
The progressive migration of financial reporting from the cost basis to fair values has created the need for a wide range of estimates to be included into the process. These values are in general clearly an improvement on obsolete cost based measurements but also create the possibility of self-serving manipulation of critical financial information. This study shows that critical estimates in one particular industry are quite weak and finds that the usage of a set of different algorithms including the ones used in machine learning to predict estimates produces far superior results.
Fonte: Ding, Kexing and Lev, Baruch Itamar and Peng, Xuan and Sun, Ting and Vasarhelyi, Miklos A., On Creating Accounting Estimates using Machine Learning (September 21, 2018). Available at SSRN: https://ssrn.com/abstract=3253220 or http://dx.doi.org/10.2139/ssrn.3253220
[...]
To prove this I teamed up with a great group of ML experts at Rutgers University (Miklos Vasarhelyi, Kexing Ding, Xuan Peng, and Ting Sun). In the first research on the use of ML in accounting, we focus on property & casualty insurance companies. The reason: a major cost item of these companies is the “future loss estimate.” Explanation: For insurance policies written (earned) in a given year, the major cost item are the payments made on insurance claims (car accidents, home damage, etc.). But that’s not the end of the story. People can, and often do claim damages years later, like long-term effects of concussions. So, in addition to claims paid during the insured year, there will be claims to be paid in the future years (“insurance long tails”). Thus, a proper measurement of insurance companies’ earnings requires accounting for both claims paid during the year, and an estimate of future claims― “future loss estimates.” These are generally large estimates having a material effect on insurance companies’ reported earnings. We focus on these estimates in our ML research.
In essence, we conducted a horse-race between managers’ annual loss estimates and the ML predictions of future claims, both compared to the actual claims paid over the subsequent 10 years. Humans, with all their frailties (manipulation) vs. machines, like former world chess champion Gary Kasparov playing against a computer.
[...]
By using ML, you can cut managers’ errors by more than a third! And this is just our initial run of ML predictions. Improvements are coming. (The third row in the table show the impact of one of our predictors―top management compensation. Without this key variable, the average error improvement decreases by 5-10%. Second vs. third rows).
Summarizing, this is the first demonstration of the use of Machine Learning in improving the quality and reliability of financial information. In addition to insurance companies’ loss estimates, ML can be applied to a wide range of managerial accounting estimates, like the bad debt reserve, future payments on warranties, pension estimates, etc., etc. Very promising indeed, leading me to the final figure:
Assinar:
Postagens (Atom)
Postagens
