Catalogo Articoli (Spogli Riviste)

OPAC HELP

Titolo:
Financial volatility trading using recurrent neural networks
Autore:
Tino, P; Schittenkopf, C; Dorffner, G;
Indirizzi:
Aston Univ, Neural Comp Res Grp, Birmingham B4 7ET, W Midlands, England Aston Univ Birmingham W Midlands England B4 7ET 7ET, W Midlands, England Slovak Tech Univ, Bratislava, Slovakia Slovak Tech Univ Bratislava Slovakia ak Tech Univ, Bratislava, Slovakia Austrian Res Inst Artificial Intelligence, A-1010 Vienna, Austria AustrianRes Inst Artificial Intelligence Vienna Austria A-1010 Austria
Titolo Testata:
IEEE TRANSACTIONS ON NEURAL NETWORKS
fascicolo: 4, volume: 12, anno: 2001,
pagine: 865 - 874
SICI:
1045-9227(200107)12:4<865:FVTURN>2.0.ZU;2-L
Fonte:
ISI
Lingua:
ENG
Soggetto:
SYMBOLIC SEQUENCES; REPRESENTATIONS;
Keywords:
financial indexes; Markov models; options; prediction suffix trees; recurrent neural networks; straddle; volatility;
Tipo documento:
Article
Natura:
Periodico
Settore Disciplinare:
Engineering, Computing & Technology
Citazioni:
35
Recensione:
Indirizzi per estratti:
Indirizzo: Tino, P Aston Univ, Neural Comp Res Grp, Birmingham B4 7ET, W Midlands, England Aston Univ Birmingham W Midlands England B4 7ET Midlands, England
Citazione:
P. Tino et al., "Financial volatility trading using recurrent neural networks", IEEE NEURAL, 12(4), 2001, pp. 865-874

Abstract

We simulate daily trading of straddles on the financial indexes DAX and FTSE 100. The straddles are traded based on predictions of daily volatility differences in the underlying indexes. The main predictive models studied inthis paper are recurrent neural networks (RNNs), In the past, applicationsof RNNs in the financial domain were often studied in isolation. We argue against such a practice by showing that, due to the special character of daily financial time-series, it is difficult to make full use of RNN representational power. Recurrent networks either tend to overestimate the noisy data, or behave like Finite-memory sources with a relatively shallow memory, in fact, they can hardly beat (rather simple) classical fixed-order Markov models, To overcome the inherent nonstationarity in the data, we use a special technique that combines "sophisticated" models fitted on a larger data set, with a fixed set of simple-minded symbolic predictors using only recent inputs, thereby avoiding older (and potentially misleading) data. Finally, we compare our predictors with the GARCH family of econometric models designed to capture time-dependent volatility structure in financial returns. GARCH models have been used in the past to trade volatility. Experimental results show that while GARCH models are not able to generate any significantly positive profit, by careful use of recurrent networks or Markov models,the market makers can generate a statistically significant excess profit. However, on this type of problems, there is no reason to prefer RNNs over much more simple and straightforward Markov models, We argue that any reportcontaining RNN results on financial tasks should be accompanied by resultsachieved by simple finite-memory sources combined with simple techniques to fight nonstationarity in the data.

ASDD Area Sistemi Dipartimentali e Documentali, Università di Bologna, Catalogo delle riviste ed altri periodici
Documento generato il 15/07/20 alle ore 07:59:00