The first pattern

The first pattern tests the two-day move following inside days that close higher or lower than the previous day’s close. The trade rules for long trades are:

Enter long at today’s close if:

1. Today’s low is above yesterday’s low.

2. Today’s high is below yesterday’s high.

3. Today’s close is below yesterday’s close.

Exit position at the close two days after entry. As formulas, these rules are:

1. If High < High[1];

2. Low > Low[1];

3. If Close < Close[1]

The rules for short trades are:

Enter short at today’s close if:

1. Today’s low is above yesterday’s low.

2. Today’s high is below yesterday’s high.

3. Today’s close is above yesterday’s close.

Exit position at the close two days after entry.

All trades were exited on the close two days after entry because initial pattern testing showed the early risk adjusted return was highest on that day: Further gains were probable with longer holding periods, but risk increased at a slightly faster pace.

The setup was tested on daily data in the EUR/USD pair from Dec. 31, 1998 to Dec. 30, 2008. Anominal initial account value of $25,000 was used and $10 was assessed per trade for commission and slippage.

Also, to see whether a trend filter might improve performance, the pattern was tested taking long trades only when the inside day’s close was above the close 40 days ago and taking short trades only when the close was below the close 40 days ago. This rule was not optimized in any way — 40 days was simply a representative intermediate-term look-back period.

1. Execute long trade signals only if today’s close is above the close 40 days ago (Close > Close[40]).

2. Execute short trade signals only if today’s close is below the close 40 days ago (Close < Close[40]).

The signals were tested with the filter because preliminary analysis indicated performance could be enhanced by accounting for trend direction — an important factor, considering the uptrend that has dominated the EUR/USD pair for much of its existence.

It compares the results of the signals with and without the filter. The no-filter results were profitable overall, thanks to the long trades; short trades actually lost money. Adding the filter made a dramatic difference: Despite cutting the number of trades in half, profitability more than doubled, the maximum drawdown was reduced by nearly twothirds, short trades became profitable, and the overall winning percentage increased by nearly 10 percentage points.

The results were profitable, but only mildly so. But they support the earlier analysis that suggested there was potential in fading the direction of an inside day’s close in the EUR/USD pair — as long as the trade is not fading the prevailing intermediate- term market direction. Now let’s look at a parallel setup that defines up-closing and downclosing days a bit differently.