Custom Search

In my search to determine whether price patterns signal the next day's market direction, I have tested all combinations of four closing prices followed by either a higher or lower open the day after the last close. I used the soybean market between 1970 and 1988 and studied 32 patterns based on open, high, low and close. The only criterion to determine the direction of the next day's trade was gross profits. The five-day patterns in Figure 1 are symbolized by a + for an up closing relative to the previous day and a - for a down closing, with the exception of the last symbol in the pattern, which is the open and the point of entry for the day of the trade.

For example, pattern 30 starts on day 1 with a lower close than the previous day (-). Day 2 has a lower close than day 1 (-). Day 3 has a lower close than day 2 (-). Day 4 has a higher close than day 3 (+). Day 5's open is lower than day 4's close (-). A sale is assumed and entry is taken on day 5's open with trade exit on day 5's close.

Because a positive gross profit was the only criterion determining trade direction (buy or sell), all patterns in Figure 1 show a profit. In some cases, the percentage of profitable trades is below 50%, but the high value of average wins pulls gross profits into the black. In other cases, average wins are less than average losses, but the high percentage of winning trades produces a positive gross profit. These are borderline patterns that may or may not have any predictive value. A trader should consider the percentage of profitable trades, the average win-to-average loss ratio and gross profits when assessing the validity of a pattern.