In Part 1 and 2 of this series, I reviewed three approaches to trading the USD/JPY forex pair using VantagePoint’s  “predicted neural index” (PIndex) as the primary entry signal, and demonstrated how money/risk management can be applied to back test results. I showed how the money/risk management strategy a trader couples with a trade entry-exit system can have surprising effects on bottom line profitability.

In this article, I will present back test results of employing the PNI+3EMA method during 2009, and I will also describe a “tweak” to a method that can further improve the outcome.  

(Note:  the results presented in here were generated following an upgrade in my version of VantagePoint, and there are some resulting quantitative differences in method performance. The qualitative features noted and discussed in each Part of the series are not affected.)  

Review

The “PNI+3EMA” method I presented uses changes in the predicted neural index in conjunction with the analysis day’s closing price relative to the Triple predicted EMA set provided in VantagePoint (3EMA+1, 8EMA+2, 18EMA+3). If PIndex changes value from “0” to “1” and the analysis day’s closing price is above all three predicted EMAs, a long trade entry is signaled and “confirmed;” if PIndex changes value from “1” to “0” and the analysis day’s closing price is below all three predicted EMAs, a short trade entry is signaled and “confirmed.”   

Stop losses were set 20 pips above the analysis day’s high for short trades and 20 pips below the analysis day’s low for long trades. Entries were at the analysis day’s closing price, the profit target was 100 pips, and the trade was exited when either (a) the profit target was reached; (b) the stop loss was reached; or (c) the trade was in profit at a daily closing – whichever came first.

Although this is a very simple method to implement, it leverages and benefits directly from all of the computational and predictive power that is inherent in VantagePoint’s predicted neural index as well as the predicted moving averages.  

2009 Backtest Results

Backtesting this method from January 1 through December 15, 2009, a total of 37 trade entries were signaled and 30 of these (81%) were winners. The method achieved a profit factor of 2.22. As shown in the chart below, despite a discipline-testing drawdown period in the middle of the year, the method netted nearly 1,000 pips.  Although this is a modest total, keep in mind it was produced by placing on average three trades per month, each were of typical duration (1-2 days) using a daily method that takes minimal time to check and implement. 

For this small number of days of exposure in the market, we might have gained $1,000 trading one mini-lot per trade, or $10,000 trading one full-size lot per trade.  As I demonstrated in Part 2 of this series, sizing trades to risk a chosen percentage of current account balance, rather than entering with a fixed trade size, would change these results, and should be examined as part of the backtesting process.

An Observation

In the process of backtesting this method, I noted that at times the PIndex changed value indicating a possible trade entry, but the closing price did not confirm the entry by closing on the corresponding side of the three moving averages until the following day. Since PIndex changes value based on differences in a three-day moving average predicted two days in advance (a short term outlook), while price closing above or below the three EMA set is more about trend and/or momentum, it stands to reason that in some instances, such as changes in momentum direction, it may take a little extra time for the price action confirmation to occur. 

Reviewing the year’s data I found that allowing an extra day for the price action to confirm a trade entry often (but not always) resulted in a successful trade.  An example of this behavior is shown in the VantagePoint chart below.   All three trade entries shown on this chart were winners, with the third trade confirmed and entered one day after the original PIndex signal. 

Revisiting the Backtesting

Redoing the backtesting, allowing this type of “delayed entry”, in addition to the original PNI+3EMA method rules, yielded positive results. A total of 46 trade entries were signaled, and 38 of these were winners, for a win rate of 82.6%. The profit factor was 2.74. 

As shown in the chart below, this approach netted nearly 1,600 pips, and the drawdown period in the middle of the year did not get to the point of challenging the “back to square one” area!

In keeping with results presented in Part 2 of this series, the chart below shows the effect of adding a money management strategy to the backtesting, risking 5% of current account balance on each trade, with an initial balance of $1,000.  

The “standard” PNI+3EMA method netted $1,403 (a gain of 40%) for the year, and the “delayed entries allowed” method netted $1,639 (64% gain) for the year.  While these may seem like “small” dollar values, earning these levels of returns while risking only 5% of account equity and devoting only minimal time to executing the method are results many will be very happy to achieve.  As the saying goes, “this ain’t your grandfather’s mutual fund!”  

It should also be noted that increasing the initial balance and/or increasing the risk percentage can have a dramatic effect on results.  For example, increasing the risk level to 7% of current account balance would result in gains of 59% and 97% respectively for the two methods, and things can “go exponential” from there. Not too shabby for a simple VantagePoint method trading one forex pair.

Cheers all, good trading!