A range that has stopped growing is more likely to be left behind by the next open than a range that has already expanded. That is the whole idea. Toby Crabel spent the 1980s turning it into definitions a trader could mark on a chart and test. The narrow day is the contraction. The break of the next opening range is the expansion. A wide day means the expansion has already happened, and taking the same breakout afterward is a way to lose the edge.
The innovation was not a new indicator. It was a control. Crabel took the identical opening-range trade after a contraction, after an expansion, and on any day at all. The any-day trade shows what a breakout is worth without a pattern. Everything else is judged against that benchmark. He also fixed the entry distance, so a win after a narrow day and a win after a wide day counted as the same kind of win. Without that control, the patterns are stories. With it, they become a method, published in his 1990 book, Day Trading With Short Term Price Patterns and Opening Range Breakout. In what follows, every pattern is defined in Crabel's own words. The trade is never taken on the bar that just qualified. That bar establishes the setup. The order is placed for the next session.
Toby Crabel
William Toby Crabel was born in 1955. Before entering the markets, he played tennis: three years on the professional tour, a place in the men’s doubles at the 1980 French Open, and, before that, an All-American college career. The habit tennis left him was repetition under rules. Trading came alongside it. He was already active in the markets by the mid-1970s, while studying finance.
In 1980, he joined RB&H Financial Services in Chicago and traded customer accounts at his own discretion. The floor of the Chicago Mercantile Exchange provided the correction. What the locals did in the first minutes after the open looked repeatable. He stopped treating it as feel and started writing it down. The notes became a commodity letter, The Active Trader, distributed to floor traders, followed by a series of research reports under the name Market Analytics. In 1987, he formed Toby Crabel & Co., a sole proprietorship that became Crabel Capital Management. In 1990, Traders Press of Greenville published the reports as Day Trading With Short Term Price Patterns and Opening Range Breakout. The book is the pattern library and the tests.
In the early 1990s, he worked in research and trading for Victor Niederhoffer while maintaining a small customer book of his own. He left to run the strategies himself. In 1992, he launched Crabel Diversified Futures and spent the following years removing discretion from the ideas: the same research, expressed as rules, with software generating the trades. In 1998, the firm launched Crabel Multi-Product, built on those legacy strategies. For a time, the operation was based in a house in River Hills, Wisconsin, northwest of Milwaukee, with the European open monitored from a basement. The flagship fund closed to new investment in 2001. Accounts of its performance report no losing calendar year from 1991 through 2002.
"On the counter-trend side, perhaps the most well-known is Toby Crabel."
James Altucher, Financial Times, January 24, 2005.
By the middle of the decade, the results were public enough to be debated. On January 24, 2005, James Altucher, writing in the Financial Times, contrasted trend-followers with countertrend traders and identified Toby Crabel as perhaps the best-known figure on the countertrend side. The article credited him with more than $1.6 billion under management. It reported that 2004 had produced two losing months, with March down 0.79 percent and June down 0.39 percent, followed by a full-year gain of 3.36 percent. The average annual return was reported as 9.48 percent, and the worst drawdown as 4.23 percent. The fee structure was the institutional standard of 3 percent and 20 percent of profits. Altucher noted that the 1990 book was already changing hands on eBay for upwards of $1,000.
The July 2005 Absolute Return survey of US firms with more than $1 billion under management ranked Crabel Capital 101st out of 196. Reports for that year put the firm’s assets at $3.2 billion and its gain at 16.7 percent. The January and July figures do not represent the same measurement, and they are not smoothed here into a single number. Both were on the record before the decade turned. A 2010 profile of the firm put assets above $3 billion and its staff at no more than about eighty, including roughly thirty researchers and thirty traders, with capital divided between an offshore fund and a US partnership.
The method had already changed scale, but its subject had not. The book tests one trade—the opening-range breakout—against a handful of range patterns. By 2011, the firm had evolved into a collection of short-term models spanning many futures markets, with increasingly automated execution and a research staff dedicated to continued testing. The reason is already present in the book, even though the book contains only one system. A single pattern belongs to a particular market regime. It works while the open retains the characteristics observed in the test, and its effectiveness fades when the open changes. The durable innovation was not NR7. It was the procedure: define the behavior, trade it the same way every time, and maintain a control so that the pattern can be distinguished from the habitual practice of buying strength.
Contraction and Expansion Price Range Patterns
A short-term price pattern, as Crabel explains in his introduction, is recent price action interpreted through previous closes, openings, range size, and the move off the open. The principle underlying the tests is contraction and expansion. A range that has stopped growing is more likely to be followed by one that breaks away from it. The narrow day represents contraction. The breakout of the next opening range represents expansion. A wide day means that the expansion has already occurred."One should be looking to go with a forceful move off the open after a contraction and not willing to do so after an expansion."
Crabel tested the same opening-range breakout after two types of contraction and two types of expansion, comparing all four against the identical trade taken on any day."An average of all percentages under the column labelled NR7 was 64%, NR4 64%, Control 57%, WS4 54% and WS7 52%. The more pronounced the expansion, the smaller the percentage chance of following through in the direction of an ORB. Profits were seven times larger for ORB trades after contractions than after expansions."
His average, not a fresh count. Contractions beat the control 30 times out of 32. Expansions were worse 29 times out
of 32. Gross profit was about $710,000 on 7,313 contraction trades and about $102,000 on 7,524 expansion trades.
No slippage. No commission.
Table A of the contraction study. NR7 and NR4 are the columns to trade. WS4 and WS7 are the columns that show the move already spent.
Stretch
Stretch is the entry distance. It is not a pattern, nor is it a ten-day average of the daily range.
“The Stretch is determined by looking at the previous ten days and averaging the sum of the differences between the open for each day and the closest extreme to the open on each day.”
For each of those ten sessions, take the smaller of the high minus the open and the open minus the low. Average the ten values. A session with an open at 1,250, a high at 1,258, and a low at 1,240 contributes 8, not 10 and not the 18-point daily range. The smaller value is the distance to the closer extreme, not the daily range and not a pivot. Pivots play no part in the calculation and are not a condition of the trade.
Opening Range Breakout (ORB)
Two entry distances appear in the book, and they do not define the same trade. In the first chapter, the stop sits beyond the opening range. In the later tests, Crabel drops Stretch and uses a constant distance from the open so that every pattern can be compared with every other pattern.
"An Opening Range Breakout is a trade taken at a predetermined amount above or below the opening range. When the predetermined amount (the stretch) is computed, a buy stop is placed that amount above the high of the opening range and a sell stop is placed the same amount below the low of the opening range. The first stop that is traded is the position and the other stop is used as a protective stop."
The glossary defines the opening range as the first thirty seconds of trading. In an ORB, both stops are active. The first to be filled establishes the position; the other becomes the protective stop.
"In later testing you will notice that I use a constant value off the open rather than the stretch point. Experience has shown this to be a better method. As long as I used the same value throughout, I could generalize about market activity."
Using a fixed distance from the open and closing the position at that day's close allowed every pattern to be evaluated on the same basis. The practical entry method, however, remains Stretch beyond the thirty-second opening range. The tests specify no price target; the measured exit is the close.
Opening Range Breakout Preference (ORBP)
An Opening Range Breakout Preference is an ORB with the side already chosen. The choice comes from the market’s own bias, or from a clear supply or demand indication. It does not come from a second indicator laid on top of the pattern."There are days when a trade in only one direction is taken, this is called an Opening Range Breakout Preference (ORBP). Usually this is done in a market with a strong bias in one direction or just after a clear supply or demand indication. The procedure is similar to the ORB but the only order entered is the stop in the direction of the entry. The protective stop is entered only after the trade has been entered. One qualification to this is if the market trades to the stretch in the opposite direction first; the ORBP is nullified and the resting order is cancelled."
After the Fill
"The ORB can be utilized as a general indicator of bias every day. Whichever side of the Stretch is traded first will indicate bias in that direction for the next two to three hours of the session."
"Early Entry is defined as a large price movement in one direction within the first five minutes after the open of the daily session. This is ideal price action when using an ORB for entry. The open should act as one extreme."
In general, stops should be moved to breakeven within one hour of entry. Markets with a greater tendency to trend should be given less time. The S&P 500, Crabel notes, usually takes no more than five to ten minutes to establish a clear move away from the open.
Trading more than one contract allows some positions to be closed as the move continues, while the remaining contracts are trailed with the stop moved to breakeven. If the initial breakout is missed, a retracement of three-eighths to one-half of the range already established can provide another entry, with the stop placed beyond the five-eighths level. He permits two such retracements; the third, he warns, is treacherous. A fill that comes late in the session is a weaker entry. Position size should decrease as the day progresses.
How a Range Is Compared
Range is the high minus the low. Each earlier bar is compared individually with today's bar. Today’s range must be strictly narrower or strictly wider; a tie with any bar in the comparison window disqualifies the pattern. The patterns are nested: an NR7 is also an NR4 and an NR. Crabel never calls the one-day case NR2, nor does he define an NR5 or a WS5. When he writes that the ORB is effective after an inside day with a range smaller than those of the previous four or five days, he is describing the trade, not introducing an additional pattern.
Green rows are the contractions. A wide day is in the book, and it is the control that lost. WS4 is "a day with a daily
range that is larger than any of the previous three day's daily ranges." WS7 is the same against the previous six.
"The ORB is effective after inside days that have a smaller daily range than the previous four or five days and for that matter after any day that has a daily range less than the previous six days (NR7) whether an inside day or not."
Inside Day (ID)
"Inside Day (ID) is defined as a narrow range day that has its daily range completely within the previous day's range. More specifically, an Inside Day's high is less than the previous day's high and the Inside Day's low is greater than the previous day's low."
Inside Day Narrow Range 4 (IDnr4).
Both extremes must be strictly inside. A touch of the prior high or the prior low is not an inside day. A plain inside day names no side. Calling it trend continuation is the special case, not the definition.
"The hypothesis is that Inside Days precede trend day activity and consequently successful opening range breakouts."
The exception is a market that is already running. Then the inside day is a springboard in the direction of that run, and the trade is an ORBP with the run. Drop the stop against it. A gap in the direction of the run is continuation. A break the other way is the one to skip. He says direction is not that predictable on an inside day that is not inside a run.
Opening-range breakout after an inside day, bonds, the S&P, beans, and cattle. The inside day is the setup.
The numbers are the next session. In bonds and the S&P; the pattern beats the same trade taken on any day.
Outside Day (OD)
An outside day is the reverse bar: a high above the prior high and a low below the prior low. Crabel uses the term. He does not give it a glossary entry, and it is not one of the contraction patterns. It is an expansion that took both sides of the prior bar. Do not treat the next morning as a fresh narrow-day breakout.
Hooks
"A Hook day is any day that opens above or below the previous day's high or low then proceeds to reverse the previous day's close but does so with a narrowing daily range relative to the previous day."
"Bear Hook is defined as a day in which the open is below the previous day's low and the close is above the previous day's close with a narrow range relative to the previous day. There is a tendency for the price action following a Bear Hook to move to the downside."
The sale is the next session, not the hook itself. He does not buy a bear hook. Thirteen of sixteen tests were profitable as sales. The exception he allows is a gap above the hook that holds through mid-session.
Bear Hook, then the ORB. Thirteen of the sixteen tests pay as sales. The S&P’s
buy-side profits are the bull-market exception, and he says so. It is not a buy.
"Bull Hook is defined as a day in which the open is above the previous day’s high and the close is below the previous day’s close with a narrowing range."
The name says bullish. The tests only lean that way, and they lean less than the bear hook leans down. Eight of sixteen trades were profitable as buys. Gross profits were $30,051 on the buys and $7,854 on the sales. It is a better trade if the next open is below the hook’s close. A fast early move against the hook is a reason to stand aside. Where his diagrams mark Day 2 and Day 1, Day 2 is the hook and Day 1 is the session that is traded.
Bull Hook Day. The buys made about four times the sales.
That is the whole case, and it is the weaker of the two hooks.
The Longer Contractions
The longer narrow ranges close the library. The object is one span, the highest high minus the lowest low, compared with earlier spans of the same length. It is not a streak of narrow days. A two-bar narrow range can contain one wide day, provided the combined span is the narrowest.
"2 Bar NR is the narrowest range from high to low of any two day period relative to any two day period within the previous twenty days."
"3 Bar NR is the narrowest range from high to low of any three day period relative to any three day period within the previous twenty days."
"4 Bar NR is the narrowest range from high to low of any four day period relative to any four day period within the previous thirty days."
"8 Bar NR is the narrowest range from high to low of any eight day period relative to any eight day period within the previous forty days."
NR7 ORB.
They are used like an NR7. They do not pick a side. The next session is a two-sided ORB, or an ORBP with the run if the market is already running.
Glossary definitions. The window is not the same for every pattern. Four bars look back thirty sessions. Eight bars look back forty.
What to Do with the Row
A narrow day, an inside day, an IDnr4, or a 2BNR through an 8BNR does not pick a side. Take a two-sided ORB the next session. Stretch is fixed before the open. The buy stop is one Stretch above the high of the first thirty seconds. The sell stop is one Stretch below that low. The first stop filled is the trade. The other is the protective stop.
If the market is already running, an inside day is not two-sided. Drop the stop against the run and take the ORBP with it. A bear hook is a sale. A bull hook prefers the buy and is the weaker signal. A wide day (WS), a WS4, a WS7, or an outside day is not a new contraction. The expansion has already left. Stand aside or wait for the next narrow day.
Whichever side of Stretch trades first is the bias for the next two or three hours, even with no position. Do not fade that first move unless it is against the intermediate trend and the pattern agrees with the trend. Move the stop to flat within an hour, sooner in a market that trends easily. There is no fixed target in the tests. The measured exit is the close of the entry day. Hold longer only if that session is already substantially profitable.
One practical limit, learned from running the patterns rather than from the book. The bar has to be one session. A holiday, a feed that revises the high, or a futures roll that splices two deliveries into one lookback will print a pattern the market did not make. Confirm the high, the low, and the close against the exchange before trading the row.
What these Definitions Are Not
The stop in the tests is a fixed distance off the open. The stop in the trading rule is one Stretch beyond the high and the low of the first thirty seconds. Using the test as the order throws away the opening range he spent the first chapter defining.
NR5 and WS5 are not in the book. Neither is a pattern called NR2. NR is simply a range smaller than yesterday's. A narrow day is not, by itself, a trend-continuation signal. Continuation is what he does when the direction is already known: an inside day (ID) inside a run, or a hook. The narrowness only says the range is ready to leave, and the leaving is a two-sided breakout.
A WS4 or a WS7 is not a second setup. In his own table the opening-range trade after them was worse than the same trade on any day. The wide bar is the result of the contraction, not the start of another one.
Nothing in the Stretch calculation refers to a pivot, a moving average of the close, or a second study. The ten numbers are distances from the open to the nearer extreme. That is the entire formula.
Summary of Patterns
NR (Narrow Range): A price bar whose range is narrower than the previous bar. Its opposite is WS. Next to NR4, NR5, and NR7 it is technically NR2.
NR4 (Narrow Range 4): A price bar whose range is narrower than each of the previous three bars, marking the narrowest range in four bars. Its opposite is WS4. NR, NR5, and NR7 are also closely watched patterns.
NR5 (Narrow Range 5): A price bar whose range is narrower than each of the previous four bars, marking the narrowest range in five bars. Its opposite is WS5. NR, NR4, and NR7 are also closely watched patterns.
NR7 (Narrow Range 7): A price bar whose range is narrower than each of the previous six bars, marking the narrowest range in seven bars. Its opposite is WS7. NR, NR4, and NR5 are also closely watched patterns.
WS (Wide Spread): A price bar whose range is wider than the previous bar. Its opposite is NR. Next to WS4, WS5, and WS7 it is technically WS2.
WS4 (Wide Spread 4): A price bar whose range is wider than each of the previous three bars, marking the widest range in four bars. Its opposite is NR4. WS, WS5, and WS7 are also closely watched patterns.
WS5 (Wide Spread 5): A price bar whose range is wider than each of the previous four bars, marking the widest range in five bars. Its opposite is NR5. WS, WS4, and WS7 are also closely watched patterns.
WS7 (Wide Spread 7): A price bar whose range is wider than each of the previous six bars, marking the widest range in seven bars. Its opposite is NR7. WS, WS4, and WS5 are also closely watched patterns.
ID (Inside Day): A bar whose high is below the previous high and whose low is above the previous low, so the session sits inside the prior bar.
OD (Outside Day): A bar whose high is above the previous high and whose low is below the previous low, so the session takes out both sides of the prior bar.
IDnr4 (Inside Day Narrow Range 4): An inside day that is also an NR4. The bar sits inside the prior bar, and its range is narrower than each of the previous three bars.
2BNR (2 Bar Narrow Range): The two-bar range, higher high minus lower low, is the narrowest two-bar range of the last 20 sessions.
3BNR (3 Bar Narrow Range): The three-bar range, higher high minus lower low, is the narrowest three-bar range of the last 20 sessions.
4BNR (4 Bar Narrow Range): The four-bar range, higher high minus lower low, is the narrowest four-bar range of the last 30 sessions.
8BNR (8 Bar Narrow Range): The eight-bar range, higher high minus lower low, is the narrowest eight-bar range of the last 40 sessions.
Bear Hook: An NR with the open below the previous bar's low and the close above the previous bar's close.
Bull Hook: An NR with the open above the previous bar's high and the close below the previous bar's close.
Stretch: The 10-session average of the smaller open-to-extreme distance, open to high or open to low. It sets the distance for an ORB or ORBP entry and stop.
Reference:
.jpg)
