System for foreign exchange trading using genetic algorithms and reinforcement learning

Foreign exchange trading has emerged recently as a significant activity in many countries. As with most forms of trading, the activity is influenced by many random parameters so that the creation of a system that effectively emulates the trading process will be very helpful. A major issue for traders in the deregulated Foreign Exchange Market is when to sell and when to buy a particular currency in order to maximize profit. This paper presents novel trading strategies based on the machine learning methods of genetic algorithms and reinforcement learning.

A machine learning approach to intraday trading on foreign exchange markets

Foreign Exchange trading has emerged in recent times as a significant activity in many countries. As with most forms of trading, the activity is influenced by many random parameters so that the creation of a system that effectively emulates the trading process will be very helpful. In this paper we try to create such a system using Machine learning approach to emulate trader behaviour on the Foreign Exchange market and to find the most profitable trading strategy.

The price elasticity of the Euro to movements in foreign reserves through European Central Bank intevention

The Euro has been used as the largest weighting element in a basket of currencies for forex arrangements adopted by several Central European countries outside the European Union (EU). The paper uses a new time-series approach to examine the relationship between the Euro exchange rate and the level of foreign reserves. It employs Zero-no-zero (ZNZ) patterned vector error-correction (VECM) modelling to investigate Granger causal relations among foreign reserves, the European Monetary Union money supply and the Euro exchange rate. The findings confirm that foreign reserves may influence movements in the Euro's exchange rate. Further, ZNZ patterned VECM modelling with exogenous variables is used to estimate the amount of foreign reserves currently required in order to again achieve a targetted Euro exchange rate

Time-resolved spectroscopy of the metal-to-metal charge transfer excited state in dinuclear cyano-bridged mixed-valence complexes

Visible pump-probe spectroscopy has been used to identify and characterize short-lived metal-to-metal charge transfer (MMCT) excited states in a group of cyano-bridged mixed-valence complexes of the formula [(LCoNCMII)-N-III(CN)(5)](-), where L is a pentadentate macrocyclic pentaamine (L-14) or triamine-dithiaether (L-14S) and M is Fe or Ru. Nanosecond pump-probe spectroscopy on frozen solutions of [(LCoNCFeII)-Co-14-N-III(CN)(5)](-) and [(LCoNCFeII)-Co-14S-N-III(CN)(5)](-) at 11 K enabled the construction of difference transient absorption spectra that featured a rise in absorbance in the region of 350-400 nm consistent with the generation of the ferricyanide chromophore of the photoexcited complex. The MMCT excited state of the Ru analogue [(LCoNCRuII)-Co-14-N-III(CN)(5)](-) was too short-lived to allow its detection. Femtosecond pump-probe spectroscopy on aqueous solutions of [(LCoNCFeII)-Co-14-N-III(CN)(5)](-) and [(LCoNCFeII)-Co-14S-N-III(CN)(5)](-) at room temperature enabled the lifetimes of their Co-II-Fe-III MMCT excited states to be determined as 0.8 and 1.3 ps, respectively.

Constructing internationally comparable real income aggregates by combining sparse benchmark data with annual national accounts data: A state-space approach

The importance of availability of comparable real income aggregates and their components to applied economic research is highlighted by the popularity of the Penn World Tables. Any methodology designed to achieve such a task requires the combination of data from several sources. The first is purchasing power parities (PPP) data available from the International Comparisons Project roughly every five years since the 1970s. The second is national level data on a range of variables that explain the behaviour of the ratio of PPP to market exchange rates. The final source of data is the national accounts publications of different countries which include estimates of gross domestic product and various price deflators. In this paper we present a method to construct a consistent panel of comparable real incomes by specifying the problem in state-space form. We present our completed work as well as briefly indicate our work in progress.

Simulated rainwater effects on anion exchange capacity and nitrate retention in Ferrosols

Nitrate (NO3) accumulations (up to 1880 kg NO3-N/ha for a 12-m profile) in the soils of the Johnstone River catchment (JRC) may pose a serious environmental threat to the Great Barrier Reef lagoon if the NO3 were released. The: leaching of artificial rainwater through repacked soil columns was investigated to determine the effect of low NO3/low ionic strength inputs on the NO3 Chemistry of the JRC profiles. Repacked soil columns were used to simulate the 11.5-m profiles, and the soil solution anion and cation concentrations were monitored at 10 points throughout the soil column. As the rainwater was applied, NO3 leached down the profile, with substantial quantities exiting the columns. Anion exchange was discounted as the major mechanism of NO3 release due to the substantial net loss of anions from the system (up to 2740 kg NO3-N/ha over the experimental period). As the soils were dominated by variable charge minerals, the effect of changing pH and ionic strength on the surface charge density was investigated in relation to the release of NO3 from the exchange. It was concluded that the equilibration of the soil solution with the low ionic strength rainwater solution resulted in a lessening of both the positive and negative surface charge. Nitrate was released into the soil solution and subsequently leached due to the lessening of the positive surface charge. Loss of NO3 from the soil profile was slow, with equivalent field release times estimated to be tens of years. Although annual release rates were high in absolute terms (up to 175 kg NO3-N/ha.year), they are only slightly greater than the current loss rates from fertilised sugarcane production (up to 50 kg NO3-N/ha.year). In addition to this, the large-scale release of NO3 from the accumulations will only occur until a new equilibrium is established with the input rainwater solution.

Spatial patterns of aerobic and anaerobic mineralization rates and oxygen penetration dynamics in coral reef sediments

Oxygen consumption rates (OCR), aerobic mineralization and sulfate reduction rates (SRR) were studied in the permeable carbonate reef sediments of Heron Reef, Australia. We selected 4 stations with different hydrodynamic regimes for this study. In situ oxygen penetration into the sediments was measured with an autonomous microsensor profiler. Areal OCR were quantified from the measured oxygen penetration depth and volumetric OCR. Oxygen penetration and dynamics (median penetration depths at the 4 stations ranged between 0.3 and 2.2 cm), OCR (median 57 to 196 mmol C m(-2) d(-1)), aerobic mineralization (median 24 to 176 mmol C m(-2) d(-1)) and SRR (median 9 to 42 mmol C m(-2) d(-1)) were highly variable between sites. The supply of oxygen by pore water advection was a major cause for high mineralization rates by stimulating aerobic mineralization at all sites. However, estimated bottom water filtration rates could not explain the differences in volumetric OCR and SRR between the 4 stations. This suggests that local mineralization rates are additionally controlled by factors other than current driven pore water advection, e.g. by the distribution of the benthic fauna or by local differences in labile organic carbon supply from sources such as benthic photosynthesis. Carbon mineralization rates were among the highest reported for coral reef sediments, stressing the role of these sediments in the functioning of the reef ecosystem.

The genetic structure of Australasian green turtles (Chelonia mydas): exploring the geographical scale of genetic exchange

Ecological and genetic studies of marine turtles generally support the hypothesis of natal homing, but leave open the question of the geographical scale of genetic exchange and the capacity of turtles to shift breeding sites. Here we combine analyses of mitochondrial DNA (mtDNA) variation and recapture data to assess the geographical scale of individual breeding populations and the distribution of such populations through Australasia. We conducted multiscale assessments of mtDNA variation among 714 samples from 27 green turtle rookeries and of adult female dispersal among nesting sites in eastern Australia. Many of these rookeries are on shelves that were flooded by rising sea levels less than 10 000 years (c. 450 generations) ago. Analyses of sequence variation among the mtDNA control region revealed 25 haplotypes, and their frequency distributions indicated 17 genetically distinct breeding stocks (Management Units) consisting either of individual rookeries or groups of rookeries in general that are separated by more than 500 km. The population structure inferred from mtDNA was consistent with the scale of movements observed in long-term mark-recapture studies of east Australian rookeries. Phylogenetic analysis of the haplotypes revealed five clades with significant partitioning of sequence diversity (Phi = 68.4) between Pacific Ocean and Southeast Asian/Indian Ocean rookeries. Isolation by distance was indicated for rookeries separated by up to 2000 km but explained only 12% of the genetic structure. The emerging general picture is one of dynamic population structure influenced by the capacity of females to relocate among proximal breeding sites, although this may be conditional on large population sizes as existed historically across this region.