Welcome Gappers!

My name is Scott Andrews and I trade the opening gap. This site is a repository for my gap trading ideas and research.  Feel free to browse and contribute to the discussions.

Twitter

Quotes
Quote and Chart Search

You can search for stocks, futures,
and forex by symbol or name.
Subscribe for Free

Enter your email address to receive blog posts by email:

Concerned about privacy?

Email delivered by FeedBurner

Blogroll
Search

Blog Archives

Entries in Futures (10)

Friday
Oct172008

How Long For Open Gaps to Fill?

Comments Off DateFriday, October 17, 2008 at 4:46PM AuthorGap Guy tagged , in

One of the common perceptions about the markets is that "all gaps eventually get filled."  This is probably true, though I've never tested it.

On a related note, one of the questions I get frequently, is "since most gaps fill within a few days, why not hold your opening gap fade position until the gap fills and thereby avert the loss?" 

While it is true that most gaps do fill within a few days, many do not.  Check out the unfilled gaps in the E-Mini S&P 500 index that remain open since last Fall: 

Surprising, huh.

So, how long does it take for the typical gap, THAT DOES NOT FILL THE SAME DAY, to close?  Answer: 19 days.  Clearly this is exaggerated by the gaps that took much longer, but it is probably longer than most traders believe.  Here's the list of multi-day fills since Fall, 2007:

 

Print ArticlePrint Article
Monday
Oct132008

Gaps & Volatility, Part 3

CommentPost a Comment DateMonday, October 13, 2008 at 11:44PM AuthorGap Guy tagged , in

In the prior post, I showed the results of fading gaps in the E-mini S&P 500 during various periods of volatility, 1/1/1998 - 10/1/2008, using only a fixed 6 pt stop.  Today I'll show the results of using a variable stop based upon the 5 day average true range (ATR). For comparison purposes, the following table shows the results of fading opening gaps using a stop equal to 30% of the 5 day ATR:

 

Observations:  As volatility becomes extreme the variable ATR helps capture a greater percentage of the winning gap fades. However, this is more than offset by the lower win rates and profitability during less volatile periods where the percent of ATR appears to shrink the stop size too much. 

The one big outlier is the 30-40 pt 5 day ATR range. Oddly, and somewhat surprisingly, the fixed 6 pt stop shows much better historical results than the volatility adjusted stop. I really can't explain - any thoughts?

Bottom line: volatility adjusted stops may be helpful during certain conditions, but not overly so - at least  not at first glance without digging further into the numbers and/or not based upon the past 10 years of E-mini S&P futures data while using a 30% of ATR factor. Perhaps a different percentage would yield better results. 

Print ArticlePrint Article
Sunday
Oct122008

Gaps & Volatility, Part 2

CommentPost a Comment DateSunday, October 12, 2008 at 11:43PM AuthorGap Guy tagged , in

Yesterday I showed the impact of gap fading the E-mini S&P 500 during various periods of volatility, 1/1/1998 - 10/1/2008, using only an end-of-day stop. Today I'll show the results of using a fixed 6 point stop for the same scenario:

As you'd expect, the win rate drops as the volatility increases since the 6 pt stop is proportionally smaller as the ranges expand.  Interestingly, all volatility bands were profitable using a 6 pt stop. The reduction in win rate is offset by an increase in the win/loss ratio.

Now, this of course begs the question, "what if I used a variable stop based upon the ATR?"  I'll consider that in tomorrow's blog post.

Print ArticlePrint Article
Friday
Oct102008

Gaps & Volatility, Part 1

CommentPost a Comment DateFriday, October 10, 2008 at 10:43PM AuthorGap Guy tagged , in

With all the record-breaking volatility, I thought it would be interesting to see if gap fading tends to do better or worse during volatile and in-volatile periods.

The following table shows the results of fading 10 contracts for all size gaps in the E-mini S&P 500 futures, 1/1/1998 -10/1/2008, at the open, targeting gap fill (i.e. prior close), and using an end-of-day stop.  Results are segmented by the average 5 day ATR range.

Excluding the ATR range of 50 - 60 points (which only has a sample size of 10 trades), the "% profitable" was fairly consistent  for all ATR ranges.  The 20 - 30 point 5 day ATR range shows the highest historical win rate and profit factor while the 30 - 40 point range shows the lowest for both categories.

What the above table doesn't factor is the result of using a fixed size stop during these various periods of volatility.  I'll evaluate the results of using a fixed 6 point stop to the same scenario in tomorrow's blog.

Print ArticlePrint Article
Wednesday
Oct082008

Gap Tip #6: "The worse they look..."

Comments Off DateWednesday, October 8, 2008 at 12:49PM AuthorGap Guy tagged , in

Today's price action reminded me of Gap Trading Tip #6 from my book, Understanding Gaps:

6.  The worse it looks, the better it works. And vice versa (most of the time). Don’t let pre-market action overly influence your decision to fade a gap or not.

This morning "Otis" (my gap system) passed on the scary looking 26 point gap down following the coordinated, joint Fed rate cut.  The reason was simple: fading "down" gaps below the low of a prior "down" day (I categorize these as D-1 gaps) is highly problematic since these gaps have lower than average fill rates and often end the day bearishly (below the open.) 

However, based upon the initial reaction which was wildly bullish pre-market (60 pt rally), I was suspicious of the sudden sell-off before the open.  So, I tested the results of fading various size D-1 gaps during weak markets (price < 50 DMA), targeting gap fill, and exiting at end of the day.  The optimization table below shows the results (assuming 10 contracts of the E-mini S&P 500 were traded).  The first column shows the minimum size gap as as percent of the prior closing price. The results were quite interesting:

Note how all 7 gaps > 2% (of the index price) in size filled or at least finished the day profitable. Of those >1.5%, eight of ten were profitable and generated the optimal profit expectancy.  As the smaller gaps are included, the profitability goes down. Meaning: that the smaller gaps in the D-1 zone have a negative profit expectancy. However, the big down gaps following down days, during weak markets, appear to be quite profitable and consistently so (though the sample size is admittedly very small). 

With this in mind, I went long about 2 minutes after the open at 1274 as prices sold off a little and market internals showed some signs of buying pressure.  My plan was to scale out half at 6 pts and lock in a winning trade (since my stop was also 6 pts).  Members in my trading room were impressed as the first target was literally hit in seconds. Then I realized that I had not yet adjusted my 2nd target which was defaulted to 10 pts.

Just as I was trying  to move it, the rally hit my target and proceeded to fill the monster gap (20 pts higher than my 2nd target). In total, the S&P moved 30 points in 7 minutes. Amazing. Though nicely profitable, that one seemingly minor mistake cost me 20 pts ($1,000) per contract. Grrrrr.  Note to self...

Print ArticlePrint Article
Tuesday
Oct072008

Follow-up to "Tomorow's Post Crash Gap"

Comments Off DateTuesday, October 7, 2008 at 5:04PM AuthorGap Guy tagged , in ,

As shown in the prior post, 9 of 10 "post crash gaps" have filled historically.

Though Otis my system did not generate a gap fade signal for today's 20 point gap up, per the "post crash gap" data in yesterday's blog, I knew that a gap fill was probable.  So, I waited for an opportunity to short after the opening.  Here's what I called out for members of the Trading Room and posted in my daily "Gap Wrap" at  www.MasterTheGap.com:

"After some weak action following the bell, I set up a short with a 4 pt stop just under the opening price. I was run over like a turtle on a country road as prices ran up and tested the overnight high. Prices then failed quickly and it became clear that sellers were "in the house" so I shorted at 1170 with another 4 pt stop. This was an aggressive (read that: sloppy) entry and I survived a hard bounce by the skin of my teeth. But prices then tanked and I scaled out for +3 and +16 pts."

Here's today's 5 minute chart of the E-mini S&P 500:

Print ArticlePrint Article
Monday
Oct062008

Tomorrow's "Post Crash" Gap

CommentPost a Comment DateMonday, October 6, 2008 at 11:17PM AuthorGap Guy tagged , in ,

Wow - what a day. At one point today prices in the E-mini S&P 500 were down 100 points from Friday's close.

In preparing for tomorrow's gap play, I checked the results of fading gaps following the current price pattern:  4 successive lower daily closes, close of 2 days ago greater than high of prior day by more than 20% of the 5 day ATR (translation: a HUGE, butt-ugly, unfilled gap), and prices are trading below the 50 day MA. 

Here are the results:

Clearly the sample size is too small to read too much into and only an end-of-day stop was used, but it is interesting nonetheless.


Print ArticlePrint Article
Monday
Sep292008

Gaps Inside the Prior 4 Days 

Comment4 Comments DateMonday, September 29, 2008 at 3:56PM AuthorGap Guy tagged , in

This morning's gap generated a discretionary "go with" signal from Otis, my gap trading system.  Meaning, my research showed that not only should one not attempt to fade the gap, but historical probabilities favored trading in the direction of the gap.  I waited until after the open to enter and nabbed 5 pts to the south side.  Unfortunately, Otis did not tell me that the ES would sell off almost 100 pts. Wow.

This morning I also shared the following interesting research nugget with members at www.MasterTheGap.com:  fading gaps that open between the low and high of EACH of the past 4 days' trading range has been problematic historically (past 10 years).

Specifically, fading all gaps that opened inside the highs and lows of each of the prior 4 days, targeting gap fill, and using a reasonable stop (25% of the 5 day ATR), shows a 50% win rate and a negative profit factor. Here's the Tradestation strategy summary: 

Print ArticlePrint Article
Sunday
Sep282008

Gap Fades By Day of Week

Comments Off DateSunday, September 28, 2008 at 11:05PM AuthorGap Guy tagged , in ,

The following are the results BY DAY OF WEEK of fading all opening gaps (excluding holidays) in the E-mini S&P 500 futures index, from 1/1/1998 - 6/30/2008, using 10 contracts, minimum size gap = 1 pt, target = gap fill, no stop, exit at end of day if gap did not fill.  This is NOT a recommended strategy.

Note: the first column entitled "factor" shows the day of week:  1 = Monday, 2 = Tuesday, .... 5 = Friday.

Two observations:

1)  "Turn-around Tuesday" is more than just a saying on the street. It is a proven historical fact as witnessed by the NEGATIVE profits from fading gaps on this day. Meaning: though the historical win rate (~72%) for fading is similar to other days, Tuesday is the day of the week that is most likely to "trend" and finish beyond, and in the same, direction as the opening gap.

2) Wednesdays and Thursdays have the best long term gap fade win rates. Further, as net profit shows, these two days are the most likely to finish inside, and possibly in the opposite direction, of their opening prices.

Print ArticlePrint Article
Saturday
Sep272008

Gap Fading the S&P 500 Index

Comments Off DateSaturday, September 27, 2008 at 9:49PM AuthorGap Guy tagged , in

The following are the results of fading all opening gaps (excluding holidays) in the E-mini S&P 500 futures index, from 1/1/1998 - 6/30/2008, using 10 contracts, minimum size gap = 1 pt, target = gap fill, no stop, exit at end of day if gap did not fill.  This is NOT a recommended strategy, but it does show the historical bias of all equity/index gaps to fill the same day.  Note:

- the very high win rate of 73%

- the max consecutive winning vs losing trades (27 to 7)

- average annual return of 12.3%

- and the time in market of about 7.3%.  This number is MUCH higher than I and most gappers average since we use a stop for each trade. Not many trading strategies offer such a compelling return versus exposure to market risk ratio.


Print ArticlePrint Article