This paper uses simple monetary economic theory in order to extract implied BTC interest rates from exchange rates, interest rates and monetary supply data. Uncovered interest rate parity permits to derive a theoretical risk free BTC interest rate that is supposed to apply in a no arbitrage...

A new method for detecting low dimensional chaos in small sample sets is presented. The method is applied to financial data on low frequency (annual and monthly) for which few observations are available.

A new method for detecting low dimensional chaos in small sample sets is presented. The method is applied to financial data on low frequency (annual and monthly) for which few observations are available.

This paper shows how the method of surrogate data can help to assess statistical significance to the forecasting performance of a model. In an application to U.S. long term interest rates, the forecasts given by a non-linear predictor are evaluated for level and sign prediction. Surrogate data...

In an application to the CAC40 daily's returns, a trading rule based on predictions given by the nearest neighbour method is shown to exhibit higher profitability than a buy-and-hold strategy only for low transactions costs and during high volatility episodes. Those results are interpreted as a...

Predictive performance of the nearest neighbor method is compared with those by the martingale model and the mean value predictor for the euro/dollar exchange rates returns. Results bode poorly for conjecture that the euro/dollar dynamics contains nonlinearities exploitable for enhanced point...