In a previous article, backtesting was conducted utilizing moving averages of varying time periods to determine their most profitable application as entry points on the AUD/USD daily chart. This article continues that backtesting to fine-tune those results.

In the previous testing, the 10-day over 30-day combination was the most profitable, with a return of 1,435 pips; however, it generated an early entry point in one trade. The next most profitable combination was 21 days over 89 days with 795 pips, but due to its lack of sensitivity, the entry points were not particularly close to the trend beginnings. Based upon these results, it seems logical that a combination of numbers somewhere between these two should provide a better balance between profitability and sensitivity.

To keep the testing on an equal basis, this week’s round makes use of the same static chart used previously. Again, the green line will always represent the moving average with the shorter time period, the yellow will represent the longer one, and therefore the entry point will be signalled by the green line crossing above the yellow one.

The dataset selected for this round of testing includes a 19-day moving average paired with a 70-day moving average; 16 and 55 days; and 13 and 40 days.

As illustrated immediately above, utilizing the 19/70 dataset gives an entry signal, marked with a vertical red line, at approximately 0.8467 on 27 September 2007 (exit near the top of the trend at 0.9200), and another on 4 February 2008 at around 0.8859 (top near 0.9350), for a profit of 1,224 pips. There is certainly nothing wrong with such a profit; however, a close examination of the chart above shows that, again, neither entry point is particularly close to the beginning of the trend and more sensitivity could be obtained.

The 16/55 moving averages generates buy signals on 26 September 2007 at 0.8450, and on 24 January 2008 at 0.8797. By making use of the same exit points as above, this generates a profit of 1,303 pips, slightly better than the 19/70 dataset. Examining the chart shows that, while the second entry point is nicely at the beginning of the trend, the first one remains stubbornly in the middle and is only slightly ahead of that generated by the 19/70 dataset.

Finally, the 13/40 dataset generates entry signals on 19 September 2007 at 0.8343, and on 10 January 2008 at 0.8766, for a profit of 1,441 pips. This is by far the best showing of the three selected datasets; however, the chart examination shows that the second entry point comes, not at the beginning of the trend, but in the middle of the previous consolidation, giving the trader an uncomfortable month of January while not advancing the previous entry point to such an appreciable degree.

As stated in the previous article, all traders must make their own decisions as to what constitutes an acceptable level of risk and reward, and where the balance must lie between profitability and sensitivity when using moving averages or any other technical indicator. This form of backtesting on static charts assists traders in determining where their own particular balances lie.