Gap Probabilities for Tuesday 7/6/2010

IMPORTANT NOTICE: Futures trading has large potential rewards, but also large potential risk . You must be aware of the risks and be willing to accept them in order to invest in the futures and options markets. This is neither a solicitation nor an offer to Buy/Sell futures or options. No representation is being made that any account will or is likely to achieve profits or losses similar to those discussed on this web site. The past performance of any trading system or methodology is not necessarily indicative of future results. There is no claim or guarantee that the following information are free of errors,including typos, program bugs, error in modeling the data, internet errors, etc.
Trade at your own risk!


Not a single loosing trade so far!

Yesterday's Forecast and Recap
Friday's Forecast:
Before Gap - According to the  modeling I use for market, there is not much data to be able to calculate probabilities based on the current market structure.  However, in a case the market opens above the high of yesterday (i.e.1029.75), the timing data suggests a win rate of 67% with a profit factor of 2.32.  This is not
one of the top setups, but I will trade it for smaller targets.
Gap trading usually works best when market is consolidating in a range.  The current market is not the ideal one for these types of trade.
After Gap - No trade for today.
Friday's Recap:
Before Gap - ES opened at 1026, under the 1029.75 which was the criteria to go short.  So there was no trade at the open.  However, the relative high probability of winning a short trade if the market opened above 1029.75 and the fact that market dropped immediately after the news spike before the open pushed the price above that high, suggested that this market may not want to stay above the high of prior day.  So I was looking for an opportunity to go short.  At 9:40 NYSE ticks were on the high side for the day, price hit the upper B band which was close to the high of yesterday, and the market was expected to consolidate for the day immediately after 3 consecutive breakout days.  So I went short at 1028.25.  I was happy to make 5.75 points before the long weekend.  This is an example of how I use the probability numbers indirectly to set up other trades which are not directly related to the opening gap.
After Gap - There was no trade.

Today's Forecast Summary:
Before Gap
Both market structure and timing data shown in Table-1 and Table 2 below agree that if the market opens in C-L zone, below the close but above the low of Friday(1014.25 to 1010.75), a long trade has a good chance to succeed.  I am going to take this trade.  A word of cautious is that Tuesday is the first day after the long holiday and it may have some inherent behavior that is not taken into account in these data. 
After Gap
 C-L zone is also a good long trade for After Gap.  I usually take my profit for Before Gap play, then wait for a pull back and go long again for an additional 4 points above close(1018.25).  Another high probability zone is the H-O zone.  So if the market opens below high but above open of Friday and it trades down to close(1014.25), then I wait for a pull back and I go short targeting 4 points below the close(1010.25).  Since the low is only two ticks above the target and may generate some resistance, I would target 1011 instead.
Additional Comments about important weekend research
Over this long weekend I found some time to do some research which I share here.  My objective was to study the level of confidence of my probability numbers from a completely different point of view.
I said if there is any advantage in these numbers at the open, this advantage must fade out gradually by time.  After the market opens other factors such as news and the chaos of trading kicks in more and more and it should wear off the initial advantage of these probabilities.  Therefore, if I can identify some advantage that wears off by time, this is an indication of the reliability of the advantage based on opening gaps.  This is specially the case, if the decline is gradual and without any bumps.  This is very important data and it is beneficial to understand it even if you are not a gap trader.  Here is the result of one study:  This study is based on 12 years of data.

Minutes PCT Resolved PF
10 24% 10.58
15 40% 3.48
20 47% 2.76
30 57% 2.16
40 67% 1.68
50 72% 1.65
60 76% 1.62
70 79% 1.55
80 82% 1.51
100 84% 1.49
120 87% 1.43
150 89% 1.40
Minutes = Minutes after the open.
PCT Resolved = PCT of the trades that either hit the target for profit or hit the stop and resulted in loss.
PF = Profit  factor = Total Win $ divided by Total Loss $.
This is an amazing result.  24% of all trades are completed in the first 10 minutes with more than 10 times winning than losing. As the time passes, more and more trades come to end but less and less are profitable.  By noon 89% of trades are completed with profits only 1.4 times the losses.  Here is a chart of PF VS Time:

(Please note the minute axis is not quite linear but that does not change the idea that I am trying to cross.)
This graph strongly supports the significance of the probability numbers I come up with everyday.  If the numbers were random or result of optimization and curve fitting, this graph would not have been completely downward.  It would have had ups and downs.  I am very amazed by this result and I hope you realize its significance.
I also took a different approach verifying the above result.  In this approach I calculated the correlation between the time passed and the profit made.  I found out that there is a statistically significant negative correlation.  That means as the time passes, the profits are less (confirming the first approach).  "statistically significant" is defined as the state that the amount of evidence exceeds the "significance level" at which point we can say it is unlikely that the event was random.  In this approach I found out that the [negative] correlation between time passed and profit starts at a level below the "significance level" at the open. It increases to a maximum of almost double the "significance level" after 40 minutes and then drops for the rest of the day.  The profit factor for the trades that complete after 40 minutes and before noon is only 1.03 which is almost break even.  The profit factor for the afternoon is 0.66, a huge money looser.  So I pick that 40 minutes area (granularity is 5 minutes) as the time that the Gap statistics is starting to loose its advantage. This is the point that I will start to look to see if there is any other reason to stay in the trade or to get out.  Otherwise its going to be a close to a 50/50 gambling trade soon.
Having that afternoon profit factor of 0.66 is tempting me to have another study to find out how much of this disadvantage comes from the morning actions vs afternoon actions.  If the afternoon actions can be brought to spotlight, it may be possible to come up with a strategy to trade the reverse of morning gap in the afternoon.

Today's Data:

ES pivot points:

R3 1043.50
Fib 2 1040.25
R2 1036.25
Fib 1 1034.00
Day 2 Hi 1029.75
High 1029.00
Open 1026.00
R1 1025.25
Pivot 1018.25
Close 1014.25
Low 1010.75
S1 1007.00
Day 2 Low 1006.00
Fib 1 1005.75
S2 999.75
Fib 2 999.50
S3 988.75
Table - 1        Gap Fade Probabilities (based on Market Structure)
Profit Target = Gap Close, Stop = 5 points.
Opening Zone > H H-O O-C C-L < L All All


Short Short Short Long Long Longs Shorts
PF
0.48 0.90 1.12 1.59 0.84 0.98 1.02
PCT Prof
33% 52% 70% 79% 57% 66% 66%
Avg Loss
$261 $241 $232 $222 $294 $270 $239
Net Loss
$2,613 $4,825 $19,225 $5,100 $15,275 $22,400 $29,613
Loose
10 20 83 23 52 83 124
Avg Win
$250 $197 $113 $91 $185 $134 $128
Net Win
$1,250 $4,325 $21,438 $8,113 $12,763 $21,975 $30,275
Win
5 22 190 89 69 164 236
Avg P/L
$91 $12 $8 $27 $21 $2 $2
Net
$1,363 $500 $2,213 $3,013 $2,513 $425 $663
Total 
15 42 273 112 121 247 360

Table - 2     Gap Fade Probabilities(based on timing)
Profit Target = Gap Close, Stop = 5 points.
 
Opening Zone > H H-O O-C C-L < L All All


Short Short Short Long Long Longs Shorts
PF
0.46 0.67 1.34 1.33 0.80 1.14 1.05
PCT Prof
29% 53% 76% 78% 47% 68% 67%
Avg Loss $361 $329 $382 $315 $298 $288 $330
Net Loss $4,338 $5,600 $10,688 $4,413 $8,938 $25,888 $34,675
Loose
12 17 28 14 30 90 105
Avg Win
$400 $198 $165 $118 $263 $154 $169
Net Win
$2,000 $3,763 $14,313 $5,875 $7,113 $29,400 $36,350
Win
5 19 87 50 27 191 215
Avg P/L
$138 $51 $32 $23 $32 $13 $5
Net
$2,338 $1,838 $3,625 $1,463 $1,825 $3,513 $1,675
Total 
17 36 115 64 57 281 320

Table - 3     After Gap probabilities(based on Market Structure)
Profit Target = 4 points, Stop = 5 points. Conditional probabilities after the gap is filled.

Opening Zone
> H H-O O-C C-L < L All All


Short Short Short Long Long Longs Shorts
PF
0.80 1.81 0.81 1.33 0.81 1.10 0.90
PCT Prof
50% 71% 52% 63% 54% 60% 54%
Avg Loss
$250 $273 $257 $261 $284 $272 $255
Net Loss
$750 $1,638 $25,425 $8,875 $9,075 $17,950 $29,088
Loose
3 6 99 34 32 66 114
Avg Win
$200 $198 $193 $200 $194 $198 $194
Net Win
$600 $2,963 $20,500 $11,800 $7,388 $19,788 $26,063
Win
3 15 106 59 38 100 134
Avg P/L
$25 $63 $24 $31 $24 $11 $12
Net
$150 $1,325 $4,925 $2,925 $1,688 $1,838 $3,025
Total 
6 21 205 93 70 166 248


If you need help with reading and using these data please visit Fading Opening Gaps 101.
If you need help with trading in general, the best place I have found so far is LBR Group headed by market wizard Linda Raschke. Beware that most other sites are purely scams.

Good luck and please direct any questions or suggestions!

Copyright © Ali Mohseni 2010 Charts are created using Tradestation software.





































No comments:

Post a Comment