Trading Articles

Range Expansion By Toby Crabel

Rate this post

This is a study of range expansion and its impact on the market. Other analysts have assumed that Range Expansion has an impact on price action on the day or days following range expansion. With the computer it becomes very easy to validate or invalidate this assumption. To begin with, it is necessary to set up a control or standard of comparison, so I borrowed from a previous study (Price Pattern Studies 1 – Bonds Open/Close Studies). In that study I tested all two, three, four, and five-day patterns in the T-Bond market, 1978 -1986. In order to make a proper comparison, I did the exact same test with only one added variable, an expanding daily range relative to the previous day’s range on the second to the last day of each price pattern. From there it was possible to make direct comparisons between Price Pattern Studies (August/September issue of Gann & Elliott Wave) and the Price Pattern with the added variable of range expansion.

To briefly review what Table A is showing, refer to Pattern #5. Simply, this pattern is two lower closings and a lower open. This is read from left to right. An entry is assumed on the open or the minus to the far right of the pattern. Exit was on the close of the same day, thus the same open to close patterns. All patterns are set up in the same manner, but as you can see, become more complex as more days are included. Regardless of the number of days, the (+) or (-) sign in the pattern is always relative to the day immediately preceding it, and are closings unless it is the last (+) or (-) sign in the pattern which is an open.

My conclusion from that comparison is that there is a definite bias in the direction of the range expansion the day immediately following. Over 75 percent of the time, the price pattern showed either a percentage profitability favoring the bias of the range expansion, and/or an average win/loss ratio that made it clear that the range expansion was influencing the price pattern. There was one problem with the comparison–there was an obvious upward bias throughout the period. This was reflected in the inability of the range expansion to have an appreciable impact on patterns ending with (–) or (++).

Read Book: Square The Range Trading System 

In the control study, almost all patterns ending with (–) showed an upward bias from open to close. The same was true for the study on range expansion and in some cases it actually produced a better trade. This was contrary to expectations. Logically, the market should exhibit a bias in the direction of the range expansion, but in fact the range expansion seemed to be acting as a selling climax rather than a momentum increase.

This was baffling and the thought occurred to me that this was not a vigorous enough definition of range expansion. Consequently I decided to make a comparison between very simple two-day patterns and five variables, one for each test (refer to Table B). The variables were: the narrowest range relative to the previous three days (NR3), Narrow Range relative to the previous day (NR), control which included all possible patterns either narrowing or expanding, Range Expansion relative to the previous day (RE), and a Range Expansion greater than the previous three days (RE3). There was thought to be enough of a contrast between Range Expansion 3 and Narrow Range 3 to give decisive evidence one way or the other on the Range expansion impact on subsequent price action.

See also  Do Chart Patterns Still Work By Thomas Bulkowski

The four patterns studied were: a higher close followed by a higher open (++), a lower close followed by a higher open (-+), a higher close followed by a lower open (+-), and a lower close followed by a lower open (–). I assumed that a simple comparison between the NR3 and the RE3 in each pattern group would give me the information that I was looking for. In the case of the (±+) pattern both the NR3 and the RE3 were profitable sales with similar win/loss ratios. The only difference was the amount of total profit which favored the NR3. This is logical and in the case of an expanding range to the upside and subsequent higher open (RE3) should not be as good a sale as on a day without as much definition preceding it. It was not totally conclusive.

A comparison of RE3 and NR3 in the (-±) shows some clear differences. Both are profitable buys, but percentages favor the NR3 pattern and win/loss ratio with total profits obviously favoring the NR3 over the RE3. Again, this is logical for a day that exhibits an expanding range to the down-side theoretically is increasing momentum, and the subsequent higher open should not be as favorable a buy relative to a day lacking such definition. The clearest difference comes in the (+-) pattern where the profitable trades are actually in the opposite direction. The narrowest range in three days followed by a lower open proved to be a profitable sale 61 percent of the time, with a win/loss ratio of 1.17 :1, and total profits of $15,18l.

The same pattern with the largest range in three days to the upside prior to a lower open proved a profitable buy 48 percent of the time with 1.10 : win/loss ratio and profits of $417. Total profits obviously are not dramatic in the case of the buy on the RE3 pattern, but there is a distinct contrast between the RE3 and the NR3 supporting the conclusion that range expansion does have an impact on subsequent price action. Where it is not so clear is in the pattern of a lower close followed by a lower open (–) where a momentum increase actually produced a better buying opportunity than on a day where there was no such range expansion. This can be explained by the dominant upward bias throughout the test period.

Nonetheless, it still seemed inconclusive that a range expansion was consistently impacting the subsequent price action. Therefore a third test sequence was done, this time taking into account only a higher close or a lower close. Again the variables of an expanding range relative to the previous three days, and a narrowing range relative to the previous three days were used in a direct comparison between the two. Entry was on close to open (C/O) and close to close (C/C) for either an overnight trade into the next day’s open or a trade from the given day’s close into the next day’s close. The thought here was that if range expansion was actually affecting the price action it would most likely show up from close to open, and if it was a particularly strong affect it would carry into close of the next day.

See also  Market Trend And Midas By Andrew Coles and David Hawkins

On a lower close and the close to open trade, the RE3 proved to be a better sale in all respects except for a slight difference in win/loss ratio. A lower close with a close to close trade showed buys for both RE3 and NR3, but as expected the NR3 proved to be the better buy with win/loss ratio and total profits reflective of this. In the case of a close to open trade, after a higher close there was a distinct difference favoring a buy in the direction of an RE3 over that of an NR3. In the case of a close to close trade after a higher close, the RE3 actually reflected a profitable buy, whereas the NR3 was a profitable sale, clearly demonstrating the RE3’s affect on the following day’s price action. From the tests exhibited in Table C and Table B, it is clear that the open acts as a climactic element in the price pattern. My observation is that most of the profits were made on the close to open trade, not on the open to close trade. In fact, the momentum increase had run its course by the time of the next day’s open. This would suggest that if one was interested in trading in the direction of a momentum increase, one should consider profit–taking on the next day’s open after the range expansion.

These tests conclusively support the concept of range expansion affecting subsequent price action, but there are some difficulties as demonstrated and it would appear that profit margins drop considerably the longer one is in the trade after the momentum increase. Also, and I think probably the most important piece of information taken from these tests, is that the trend is a stronger influence than short-term momentum increase or decrease. In other words, the underlying intermediate long-term trend will take precedence over a momentum increase indicated by a range expansion against the trend. Range expansions with trend hold more significance than if against trend. The psychology involved could be something as follows: In a strong up-trend, new shorts are constantly entering the market trying to pick the market top; the isolated range expansion to the down-side brings in more shorts; the high percentage open lower the following day and has very little follow through, and somewhere around mid-session after a bullish day, structure shorts begin to cover.

Emphasis should be placed on the necessity of integrating other market concepts along with range expansions. Without a firm grasp of the concept of day structure and a clear picture of the underlying trend, a trader will run into difficulties using range expansion on its own.

Leave a Reply

Your email address will not be published. Required fields are marked *