This research was previous published in Quantifiable Edges, and investment newsletter available at http://www.quantifiableedges.com/ , and published by Rob Hanna, CMT. It is reprinted here with permission.
Editor’s Note: The Opening Range Breakout is a classic trading technique employed by many very successful traders. However, very little has been written about it in the literature, perhaps because it works so well. This article provides more than enough detail to allow motivated market students to study the technique and builds on the existing body of knowledge by using pre-market data in the trading strategy.
Opening Range Breakouts in tight markets
Opening Range Breakouts (ORB’s) were popularized years ago by Tony Crabel. In his work he looked for narrow-range bars to occur as part of the setup. He would then expect an expansion in volatility and attempt to profit from that expansion by trading a breakout of an early morning range. Using slightly different tools I looked at some basic methods of trading ORB’s and did a statistical study to see if I could identify an edge. Below I will first describe how I measured a volatility contraction, then I will apply this to ORB’s. Lastly I will examine the concept of using pre-market data in calculating opening ranges.
Measuring the volatility contraction
As background, in July I conducted a study that looked at how the market has acted following sharp short-term contractions in volatility. In that study I used the 3-day historical volatility and compared it to the 10-day historical volatility of 3 days ago. The reason I offset the 10-day historical volatility was so that there was no overlap in the calculation. When dividing the 3-day by the 10-day offset a result below 1 would indicate the last 3 days have been less volatile than the previous 10. A result above 1 indicates a recent uptick in volatility.
Looking back to 1960, I then examined what a sharp contraction in historical volatility over the last 3 days might indicate about the next 3 days. What I found is that when the 3-day over the offset 10-day dropped to 0.25 or lower the next 3 days were 5.5 times as volatile as the recent 3 days. This was based on 1,111 occurrences since 1960 – or about 9% of all trading days. Going back to just 1999 gave similar results, as the 3-day historical volatility increased by 5.4 times.
Testing ORB’s with the volatility contraction
So the 3/10 offset historical volatility study suggests a short-term volatility expansion is expected. Let’s now look at some simple tests utilizing the SPY and 5-minute bars to determine whether the ORB concept might help in establishing an edge. Tradestation was used to conduct all of the below tests.
This first test looks to enter in the direction of the breakout. It then places a stop on the opposite end of the opening range. The trade is exited at the close if it is not stopped out before the close. Should the breakout prove false and the entire range is reversed in the next 5 minutes, an entry in that direction will be taken. Otherwise, no 2nd entry is taken on the day.
You’ll note the longer you wait the higher the winning percentage. While the winning % is relatively low across the board the size of the winners is much larger than the losers and therefore there is a positive expectancy. I also tested this without using a stop and just selling all ORB’s at the day’s close. That also was profitable although slightly less so than using a stop. The edge here isn’t huge, but with some trade management it should be able to be improved. By doing things like trailing stops, taking partial profits, or exiting on a breakdown of a topping (or bottoming) formation traders should be able to improve upon a simple end-of-day exit.
Using the 30-minute ORB (first 6 bars) I broke out the results by longs and shorts to make sure they were both profitable. Below is the longs:
Here we see a mild edge. Now the shorts.
Again a fairly mild edge, but a little bit better than the longs.
I also checked to see that the low HV actually had a positive influence and it wasn’t just the ORB that provided all the edge. To do this I check all instances over the last 10 years that didn’t come after a very low 3/10 offset ratio.
What we see here is a mild negative expectation if the market isn’t doesn’t have a low 3/10 offset HV ratio. I actually diced it up a little more than this and found generally positive ORB trade results below 0.75 and generally negative results above 0.75.
Examining use of the pre-market range
In examining the results of this study I also considered whether using pre-market action could also provide an edge. Below is a 5-minute chart of SPY on 8/27/09. It includes the pre-market trading from 8am -9:30am eastern.
The problem in this scenario of try to play a break of the 1st 30 minutes of trading is that the SPY spent almost the entire 1st 30 minutes selling off. It wasn’t a sideways consolidation pattern you’d be looking to take a break of. Instead the market was already very extended. Shorting the 10:05 bar simply isn’t a good idea with the market having already free-fallen for the 1st half hour.
An alternate entry here and one that would have done quite well as long as you trailed a stop or took profits at some point would be shorting the breakdown of the premarket trade. But how would using the premarket range as your entry triggers have worked over time?
Here we see that from 2002 – the beginning of my data – through mid-2007, using the pre-market range to play a breakout would have led to small but consistent losses. So what changed in 2007?
I looked at the data a few different ways and devised this chart, which I thought was quite interesting. It measures the size of the range (in percentage terms) of the 8 – 9:30 premarket action in SPY.
What we see here is that in 2007 there was an explosion in volatility in pre-market trading. The explosion peaked in early 2009. And while is has been steeply downtrending lately, it is still about 3 times larger than it was for most of the period prior to 2007. I envision at least 2 implications of the expanded range that may have aided in the success of breakout trading them.
First, with ranges of only 0.25%, pre-2007 trades had stops that were too tight. It was too easy to get stopped out of one of those trades and it happened a lot.
Second, the larger range would also suggest more volume. With more volume and a bigger range, support and resistance levels established pre-market would have more substance. Breaks of these levels would be more impactful since real buying and selling occurred. Low volatility, low volume, premarket drift pre-2007 did not establish meaningful ranges. It appears the more active premarket trading of the last 2 years has helped to establish more meaningful ranges that could be used to a traders advantage. Traders may want to keep an eye on premarket activity moving forward. Should the above chart continue to quickly erode, then the edge of using premarket levels for support and resistance would also likely erode. If we see a leveling off or move back up in the above chart, then that might suggest using premarket levels for trade triggers could continue to provide an edge.
Final Thoughts
These tests are in no way meant to represent a complete trading system. The takeaway here is that when the market is trading unusually tight on the daily charts, you will often get a strong move and expansion in volatility. Under these circumstances, if a range is established either pre-market or during early trading with clear levels of support and resistance, traders could look to play a breakout of that range and be confident that risk/reward levels were skewed in their favor.
I also believe some reasonable trade management, stop trailing, and perhaps enhanced entry filters could greatly improve this edge. Traders comfortable trading in an intraday timeframe could look to exploit these edges when they occur.
This report has been prepared by Hanna Capital Management, LLC and is provided for information purposes only. Under no circumstances is it to be used or considered as an offer to sell, or a solicitation of any offer to buy securities. While information contained herein is believed to be accurate at the time of publication, we make no representation as to the accuracy or completeness of any data, studies, or opinions expressed and it should not be relied upon as such. Robert Hanna, Hanna Capital Management, LLC or clients of Hanna Capital Management, LLC may have positions or other interests in securities (including derivatives) directly or indirectly which are the subject of this report. This report is provided solely for the information of Hanna Capital Management, LLC clients and prospects who are expected to make their own investment decisions without reliance upon this report. Neither Hanna Capital Management, LLC nor any officer or employee of Hanna Capital Management, LLC accepts any liability whatsoever for any direct or consequential loss arising from any use of this report or its contents. This report may not be reproduced, distributed or published by any recipient for any purpose without the prior express consent of Hanna Capital Management, LLC. Copyright © 2009 Hanna Capital Management, LLC.
Letter from the Editor
