(published in the VantagePoint newsletter)

In my last article I described applying a “PNI+3EMA” method to trading the S&P500 Emini.  In this article, as promised, I investigate adding other VantagePoint indicators to the analysis to improve results achieved thru backtesting.    

Review     

Regular readers of this Newsletter will recall the “PNI+3EMA” method I described and demonstrated previously for trading the USDJPY forex pair.  In my last article, I adapted the method to trading the S&P500 Emini.  In a nutshell, the trading rules for this method are:

Enter a Long trade when

(1) the Predicted Neural Indicator (PNI) changes from 0 to 1; and

(2) price closes above all 3 moving averages in VantagePoint’s Triple Predicted EMA set (P3EMA+1, P8EMA+2, P18EMA+3) on the day of or the day following the PNI change. 

For backtesting this Emini market, a stop loss was set 30 points below the analysis day’s closing price, and a profit target was set 15 points above the analysis day’s closing price.  Each trade was held open until the stop loss was hit, the profit target was hit, or a profitable daily closing price was achieved.  These rules are ‘reversed’ appropriately for short trades.

The chart below was included in the previous article.  It covers the 1 month period of January 2010, and the blue arrows show the indicated trade direction (long or short) while the colored ellipses indicate the trade results – green for ‘win’, red for ‘loss’.  In this time period, the PNI+3EMA method signaled 5 trades, 4 of which were profitable.  Backtesting this approach for all of 2010 (thru July 31) yielded 22 trade entries, with 18 wins and 4 losses, for a win rate of 82%.  The profit factor (ratio of gross wins to gross losses) was 1.46, and a total of 55 points was earned. 

Readers who have followed my articles for awhile know that I ‘often’, ‘usually’, turn to one or more of VantagePoint’s Predicted Difference indicators to help filter out less than ideal trades and improve the bottom line.  This situation is no different – but the final conclusion may surprise you!

Adding a Difference

Because we have had some success using VantagePoint’s Predicted Long Term Difference indicator (PTLDiff) to filter and confirm trades in the USDJPY forex market, I added it to the indicator suite being investigated for trading the S&P500 Emini, and repeated the backtesting.  The entry rules were expanded to incorporate this indicator by requiring that PTLDiff be moving in the direction of any trade entry signaled by a PNI change and closing price relative to the Triple EMA set.  In other words, if a long trade was signaled per the PNI+3EMA method rules described above, PTLDiff had to also be increasing in order to confirm the entry signal, or decreasing to confirm a short entry signal.  Also, if there was divergence between PTLDiff and price action, the trade was not entered.

The chart below shows the same 1 month period of January 2010 with the PTLDiff indicator added.  As shown by the green ellipses on the chart, 3 winning trades were signaled by the basic method and confirmed by PTLDiff during this period.  PTLDiff kept us out of a couple trades during the month:  at the point labeled ‘1’ on the chart, a long trade entry was not confirmed due to PTLDiff decreasing; and at the point labeled ‘2’ on the chart, another long entry was not confirmed due to divergence between PTLDiff and price action, which is also highlighted by the red dashed lines.  

 Applying this approach to the first 7 months of 2010 (thru July 31) produced disappointing results.  A total of 19 trade entries were signaled, resulting in 15 winning trades and 4 losing trades, for a win rate of 79%.  A total of only 28 points was earned, barely half that of the basic method, and the profit factor was 1.24.  These aren’t terrible results, profit is still being earned, but certainly very slowly. 

Continuing the hunt for a better bottom line, I investigated the impact of using VantagePoint’s Predicted Medium Term Difference (PTMDiff) in place of PTLDiff, using the same exact logic as described above.  These results were more encouraging, with a total of 18 trade entries signaled, 16 of which were winners, for a win rate of 89%.  A total of 89.75 points was earned, and the profit factor was 2.50. 

Similarly, I redid the backtesting yet again using VantagePoint’s Predicted Short Term Difference (PTSDiff) as the ‘filter’ or ‘confirmation’ on the basic PNI+3EMA method.  I found that if I ignored divergences between price action and movement/levels of the PTSDiff, results similar to those obtained using PTMDiff were achieved – 21 trades, 18 wins, and a win rate of 86%.  A total of 90.25 points was earned, and the profit factor was 2.0. 

Wrapping It Up

A trading method that produces a profit factor equal to or greater than 2 is desirable, since this means, in bottom line terms, at least $2 is earned in profit for every $1 that is lost.  So, with a profit factor of 2.0, profits replace every dollar lost, plus add an additional dollar in profit to the bottom line.  So the profit factor of 2.0 achieved by using PTSDiff is ‘good’, but not quite as good as that attained using PTMDiff (2.50). 

The Table below summarizes the backtesting results obtained and described in this article:

It is pretty obvious from the backtesting results that using VantagePoint’s PTMDiff or PTSDiff indicators in conjunction with the ‘basic’ PNI+3EMA method can significantly improve bottom line results in trading this market, since the win rates, profit factors, and netted points are definite improvements from those of the basic method.  This isn’t to say either of these are definitely ‘the only’ or ‘the best’ way to trade this market – they are just examples of some ‘quick strike’ approaches that can be used.  Certainly many other combinations and approaches are possible.

Another observation to be made is that obviously, ‘win rate’ doesn’t tell the whole story when it comes to evaluating and comparing trading methods, since a ‘good’ win rate may or may not indicate healthy bottom line profits, such as in the case of the basic method or the basic method plus PTLDiff.  Metrics such as profit factor and bottom line net earnings can and should also be examined to assess ‘goodness’.

In my last article, I pointed out that a basic VantagePoint trading method ‘should’ transfer from one market to another – ‘Basically, Yes’ – but that the devil is also in the details, because there are differences in the way markets work and indicators react -- and so I also said ‘Specifically No, Be Careful’.  The results described in this article demonstrate those ideas – since for a certain forex market we found that the PNI+3EMA method using PTLDiff as a ‘filter’ was preferable – and here, the basic method works ‘ok’, but using PTMDiff as a filter looks to be a more desirable approach than using PTLDiff.

Although I haven’t formally backtested it yet, logic says this trading approach should also perform well when applied to other similar index markets (NASDAQ, Russell, etc), since they tend to behave similarly. If it doesn’t - a really cool thing about VantagePoint is that there are many more indicators in VantagePoint’s world that can be combined to form a set that performs just as well or better in these other markets of interest….. and we live to profit another day!  

Cheers all, Good Trading!