Forex Automaton

Our primary goal is to create a public information service providing financial markets forecasts, based on our proprietary forecasting tools: an automated trading system -- a Forex Automaton™. Our secondary goal is to quantify and monitor the very existence of sustainable opportunities for arbitrage profit-making. Or simply put, to monitor the degree to which these markets are more predictable than a "fair game" -- to a trader without access to insider information.

Forex Automaton was launched in April 2008 with the ambitious mission of leveraging the specific algorithmic know-how to create a trading signal service geared toward retail forex traders. From the very beginning a two-prong strategy was adopted: first, development of the trading system product whose usefulness relies on secrecy of the relevant know-how. Second, white-paper research focusing on statistical properties of the market time series, especially those aspects which are potentially interesting from the point of view of algorithmic trading, however counter-intuitive, technical and remote from the mainstream picture of forex trading they may be. As of now, it is mostly the second prong that's visible to the website visitor. This document summarizes the main findings to emerge so far from a year of studies, including some glimpses into the progress made on the black-box algorithmic trading system front.

For the first time I am able to discuss performance of a ForexAutomaton system without the "benefit of hindsight" caveat: the results for January have been obtained in real time, hence no hindsight.

Executive summary

In the absense of a capital allocation and trade-idea disrimination system, the main figure of merit is the Pearson correlation coefficient between real and forecast logarithmic returns in day high, low and close. In January, these figures of merit appear to remain in line with longer range historical performance. A type of a trade strategy specifically designed to take advantage of the superior forecasting quality for daily high and low, while minimizing exposure to the forecast for close, is discussed, with an attempt to evaluate performance using the recorded output of the system in January.

In the course of the trading system optimization, best returns have been seen to be obtained by ignoring forecast moves  below a certain threshold, and acting on those above that threshold (that threshold was also called entry parameter). A small fraction of large returns has been seen to account for much of the positiveness of the Pearson correlation coefficient between actual logarithmic returns and their forecasts. Each Danica output contains forecasts for 14 forex rates, and a natural question is: how do I pick the ones to place trades? One might expect that the odds of success can be improved by selecting those markets where the next move is forecast to be large. I take version 0.5 of Danica forex system and study dependence of the correlation between the forecast and real logarithmic returns in day close on the relative strength of the forecast move.

Some opportunities for analysis are offered by the fact that the forex trading system such as Danica gives not just a forecast for the next close, but a combination of next high, low and close. This report is the first attempt to develop a selective approach to the forecasts, a discrimination algorithm of sorts, such that a decision to give the forecast further consideration or ignore it would be based on the information contained in the forecast itself.

This report gives a more detailed discussion of the optimization trade-offs made for Danica, as compared to the initial announcement. At the same time, it establishes a reference point for future studies of past performance.

It's now official: from now on, a forex forecast of low, high and close for the next 24-hour period will be posted on this site daily at 9am Eastern time. The system is named Danica following the naming convention where first names starting with D are assigned to systems with daily decision-making scale.

While it is true that past performance does not indicate the future, the only reliable information we have is about the past. A few important things make a difference between unbiased trading-system testing and self-delusion. Here I summarize my current understanding.

This document explains the output of Danica -- an experimental free (payless but closed-source) ForexAutomaton day-scale forex forecasting system.

Just like logarithmic returns can be defined and analyzed for daily close, they can be defined for daily high and low. Japanese candlestick charting techniques, believed to have predictive power, study patterns formed by open, low, high and close as the time series progresses. In this report I extend application of my newly developed forecasting figure of merit, Pearson correlation coefficient between the real and predicted logarithmic returns, to the daily high and low, taking another look at the dependence of the prediction quality on the magnitude of a forecasting parameter nicknamed Fred. As the prediction quality is seen to be much better for the next high and low than it is for next close, I am contemplating ways of improving quality for close.

As we are gearing up towards the launch of a real-time day scale forecasting service, I am using the newly developed forecasting figure of merit, Pearson correlation coefficient between the real and predicted logarithmic returns, to make a choice of the operating mode for the test mode of the service. I am taking another look at the dependence of the prediction quality on a day scale on the magnitude of a forecasting parameter nicknamed Fred for two distinct pattern recognition modes, namely that of completely independent analysis of the different exchange rates (so-called Step One), and the one in which the time series for the indvidivual exchange rates are considered jointly in order to utilize potential inter-market patterns (Step Three).

This interview appeared in the Forex Hunter Blog during that blog's brief but stormy life in November 2009. With Forex Hunter's permission, we reproduce it here.

The intention of this post is to tie together several topics which appeared on my radar screen in the course of the trading system optimization. First, it has been understandably hard to fully rid oneself of  vestiges of the mainstream financial theory based on the postulate of market efficiency, while building a wealth-generating tool relying explicitly on demonstrable market inefficiencies. The realization that Sharpe ratio does not let one make an objective choice of a portfolio was  there from the beginning, and I recall perceiving this fact as a "necessary evil". Then came the understanding of the fact that an  artithmetic average of returns gives one a biased picture of long-term return, and consequently, Sharpe ratio is built around biased quantities.

 I approach pairs trading with the correlations tool-box and basic algebra. Let's consider two time series, a(t) and b(t). It will be understood that these are  taken on a fixed time scale (second, minute, hour, and the like). Most explanations of pair trading fail to communicate the importance of non-zero correlations at non-zero time lags --  let alone the importance of their constructive interference (to be explained). Meanwhile, it is these subtleties that make a difference between just another roulette-like source of random outcomes and a reliable, little-risk source of arbitrage profits. 

This is a brief follow-up to the previous post on how I know my forex forecasting works. In that post I disclosed a measurement of a figure of merit I use to monitor the forecasting quality and optimize the algorithm, the figure of merit being the covariance of predicted and actual logarithmic returns on a day scale. The measurements were carried out for 16 values of the control parameter nicknamed Fred, which is currently the only "make it or break it" parameter responsible for the forecasting, and the only one being currently optimized. (As an aside note, there are other quantities which control the process like for example how big a chunk of data you look at. Those are believed to be more mundane and are currently fixed as some "reasonable" values -- which is not to say that I won't decide to take a more quantitative look at how reasonable those values are sooner or later.)  

The covariance of predicted and actual logarithmic returns is not the best quantity to look at when aggregating data for the different forex exchange rates: because of the somewhat different volatilities of those markets (different even despite the fact that the logarithmic returns take the absolute value of the exchange rate out of the picture), the resulting numbers for the major forex were volatility-weighted averages. Moreover, a quantity like 10^-6, even if it's more than 2 standard deviations above zero, does not communicate the result to the non-expert in the intuitive way the result deserves.

These are the reasons why I went over from covariances to Pearson correlation coefficients, and today I am presenting the updated measurements.

In the recent forex/CFD data, USD/CAD is negatively correlated with light oil (WTI) CFD. This is the same as saying that CAD, one of the commodity currencies, is positively correlated with oil. This is old news. In this article I take a deeper look at the issue and analyze the shape of the correlation peak. Analyzed on the hour time scale, the correlation peak is broad and somewhat asymmetric, indicating that it is much safer to rely on the guidance of USD/CAD in predicting the oil price, rather than other way round. The necessary caveat is that this is a time-integrated picture, covering a period from late August 2008.