Fading Gap Strategy?

I was wondering if anyone has experimented with fading the opening gap strategy on ASX stocks?

Would like to have a discussion about results, I have been testing this myself a bit now too.

Cheerio!
Gert

Yes. Its very effective but you'd be best using stock index futures rather than straight shares.

Some pointers;

1. The larger the gap, the better the fade but the lower the trade frequency
2. The stop should be dynamic and not exceed 30% of the average daily range
3. A substantial decrease in volatility will destroy the pattern. It pays to stand aside when vol. drops too much.
4. Usually a volatility expansion is a better method for individual stocks

Good observation. Keep at it.

Heres some random stuff on gaps from various places.

http://www.cbot.com/cbot/pub/cont_detail/0,3206,1070+14719,00.html

http://www.mypivots.com/articles/articles.aspx?artnum=43

http://stockcharts.com/school/doku.php?id=chart_school:trading_strategies:gap_trading_strategies

http://stockcharts.com/school/doku.php?id=chart_school:chart_analysis:gaps_and_gap_analysi

The pros know before an open that a gap play will happen. You can see the volume load up in the first few minutes after the start. There are programs that give you "percentage of gap closure" based on various factors, but Ive only seen them for US indices, as they also get more internal data than asian indices.
I read somewhere that countertrend gaps are more likely to close than with trend gaps, but cant say if it's true or not.
I guess if you see a gap, keep in mind the possibility it might close, then wait for price action or volume to confirm your suspicions, then enter when you can.
Use regular money management etc.

Nick Radge said:

Yes. Its very effective but you'd be best using stock index futures rather than straight shares.

Some pointers;

1. The larger the gap, the better the fade but the lower the trade frequency
2. The stop should be dynamic and not exceed 30% of the average daily range
3. A substantial decrease in volatility will destroy the pattern. It pays to stand aside when vol. drops too much.
4. Usually a volatility expansion is a better method for individual stocks

Good observation. Keep at it.

Click to expand...

Nick,

Freddo pointed out to me on another thread that opening gaps have a 'tipping point' where overnight holders will cover their positions on open...

We've seen this a number of times over the past few weeks where the market has gapped on open the pushed up further before reversing and closing the gap.

What are your thoughts ?

"Yes. Its very effective but you'd be best using stock index futures rather than straight shares. "

Click to expand...

Yes I've actually been using the S&P/ASX200 index values as a filter with the gaps, though with mixed results.

At the moment only using EOD data for test as well. Thanks for your advice on the stop, I was thinking of using a tight-ish trailing stop too on 5-min data although this seems to take quite a while to load (& run).

Test period from Feb2006 - Mar2008. Win ratio around 40%.

Can I ask what program you use to backtest systems?

-G

Thanks Broadway - some good articles!

Has anyone backtested a gap-fading strategy for the ASX market?

I have obtained some results as follows: ave_daily is ave. % daily return, so for 0.9% gap size, all trades generated are grouped by day, their % returns averaged (for each day) and the ave_daily then calc'ed as an average of all trading days.

gap win_perc lose_perc tot_trades ave_daily sharpe
0.500 0.380 0.620 5597 -0.103 -0.074
0.600 0.380 0.620 5406 -0.095 -0.067
0.700 0.381 0.619 5232 -0.082 -0.056
0.800 0.378 0.622 5057 -0.071 -0.047
0.900 0.374 0.626 4917 -0.058 -0.038
1.000 0.373 0.627 4764 -0.039 -0.025
1.100 0.373 0.627 4542 0.001 0.001
1.200 0.371 0.629 4375 0.023 0.014
1.300 0.373 0.627 4217 0.043 0.026
1.400 0.373 0.627 4029 0.101 0.059
1.500 0.373 0.627 3870 0.131 0.073
1.600 0.371 0.629 3715 0.150 0.081
1.700 0.371 0.629 3576 0.180 0.095
1.800 0.370 0.630 3467 0.195 0.101
1.900 0.368 0.632 3337 0.245 0.123
2.000 0.370 0.630 3212 0.270 0.132
2.100 0.365 0.635 3081 0.271 0.126
2.200 0.362 0.638 2964 0.278 0.128
2.300 0.362 0.638 2849 0.292 0.128
2.400 0.366 0.634 2727 0.361 0.156
2.500 0.366 0.634 2630 0.377 0.159
2.600 0.363 0.637 2511 0.379 0.152
2.700 0.363 0.637 2424 0.400 0.156
2.800 0.361 0.639 2313 0.516 0.201
2.900 0.359 0.641 2248 0.528 0.207
3.000 0.359 0.641 2170 0.543 0.210
3.100 0.360 0.640 2096 0.562 0.213
3.200 0.359 0.641 2029 0.569 0.207
3.300 0.358 0.642 1961 0.608 0.215
3.400 0.358 0.642 1895 0.601 0.207
3.500 0.358 0.642 1834 0.600 0.204
3.600 0.355 0.645 1764 0.625 0.210
3.700 0.358 0.642 1713 0.664 0.222
3.800 0.361 0.639 1650 0.706 0.231
3.900 0.356 0.644 1588 0.676 0.221
4.000 0.359 0.641 1536 0.702 0.227
4.100 0.362 0.638 1462 0.720 0.232
4.200 0.365 0.635 1382 0.734 0.235
4.300 0.362 0.638 1336 0.717 0.227
4.400 0.359 0.641 1269 0.730 0.229
4.500 0.358 0.642 1233 0.750 0.234
4.600 0.353 0.647 1180 0.767 0.232
4.700 0.351 0.649 1150 0.778 0.235
4.800 0.350 0.650 1086 0.822 0.241
4.900 0.350 0.650 1061 0.827 0.241
5.000 0.347 0.653 1036 0.796 0.228
5.100 0.351 0.649 993 0.795 0.224
5.200 0.347 0.653 971 0.775 0.216
5.300 0.344 0.656 936 0.726 0.202
5.400 0.343 0.657 926 0.729 0.203
5.500 0.345 0.655 898 0.742 0.202
5.600 0.351 0.649 857 0.770 0.205
5.700 0.349 0.651 836 0.731 0.196
5.800 0.346 0.654 817 0.739 0.197
5.900 0.349 0.651 785 0.770 0.202
6.000 0.349 0.651 773 0.778 0.204
6.100 0.348 0.652 749 0.783 0.204
6.200 0.353 0.647 731 0.797 0.207
6.300 0.351 0.649 700 0.737 0.189
6.400 0.350 0.650 682 0.709 0.182
6.500 0.349 0.651 662 0.738 0.196
6.600 0.352 0.648 647 0.764 0.203
6.700 0.354 0.646 618 0.735 0.197
6.800 0.350 0.650 602 0.722 0.193
6.900 0.355 0.645 588 0.741 0.196
7.000 0.352 0.648 576 0.735 0.194
7.100 0.351 0.649 561 0.754 0.198
7.200 0.350 0.650 529 0.729 0.193
7.300 0.352 0.648 517 0.736 0.195
7.400 0.354 0.646 508 0.748 0.198
7.500 0.357 0.643 498 0.782 0.207
7.600 0.355 0.645 485 0.736 0.197
7.700 0.365 0.635 455 0.733 0.196
7.800 0.362 0.638 448 0.735 0.197
7.900 0.360 0.640 436 0.705 0.189
8.000 0.355 0.645 420 0.683 0.186
8.100 0.358 0.642 411 0.710 0.193
8.200 0.360 0.640 397 0.718 0.191
8.300 0.363 0.637 391 0.739 0.199
8.400 0.364 0.636 360 0.762 0.203
8.500 0.365 0.635 353 0.769 0.199
8.600 0.365 0.635 348 0.769 0.199
8.700 0.365 0.635 337 0.757 0.195
8.800 0.363 0.637 328 0.747 0.193
8.900 0.368 0.632 318 0.723 0.182
9.000 0.369 0.631 314 0.726 0.183
9.100 0.376 0.624 290 0.656 0.169
9.200 0.379 0.621 285 0.658 0.170
9.300 0.378 0.622 283 0.639 0.167
9.400 0.377 0.623 273 0.638 0.173
9.500 0.377 0.623 265 0.642 0.175
9.600 0.378 0.622 262 0.644 0.175
9.700 0.373 0.627 255 0.631 0.171
9.800 0.378 0.622 249 0.642 0.173
9.900 0.382 0.618 246 0.642 0.173
10.000 0.379 0.621 235 0.614 0.170
10.100 0.372 0.628 226 0.580 0.162
10.200 0.376 0.624 221 0.579 0.161
10.300 0.372 0.628 215 0.551 0.154
10.400 0.374 0.626 214 0.551 0.154
10.500 0.368 0.632 204 0.574 0.161
10.600 0.369 0.631 198 0.564 0.159
10.700 0.368 0.632 193 0.528 0.151
10.800 0.363 0.637 182 0.528 0.151
10.900 0.362 0.638 177 0.524 0.150
11.000 0.354 0.646 175 0.514 0.147
11.100 0.356 0.644 174 0.517 0.148
11.200 0.358 0.642 162 0.476 0.138
11.300 0.357 0.643 157 0.454 0.133
11.400 0.355 0.645 155 0.442 0.129
11.500 0.353 0.647 153 0.436 0.128
11.600 0.354 0.646 147 0.451 0.134
11.700 0.352 0.648 145 0.454 0.135
11.800 0.358 0.642 137 0.459 0.139
11.900 0.353 0.647 136 0.454 0.138
12.000 0.353 0.647 133 0.440 0.136
12.100 0.354 0.646 130 0.422 0.132
12.200 0.354 0.646 127 0.405 0.128
12.300 0.352 0.648 125 0.403 0.127
12.400 0.352 0.648 122 0.424 0.135
12.500 0.361 0.639 119 0.423 0.135
12.600 0.354 0.646 113 0.437 0.138
12.700 0.351 0.649 111 0.421 0.134

lol those results dont make much sense do they?!

Sorry gvanto, not to me.
I do remember reading in a past spi thread that 60% of spi gaps close in 2-3 days.
The problem is stops.
Good luck.

charttv (an ASF member) did some research on this (or similar strategy, might have been on SPI gaps)....worth searching for his old posts via the search tool. I might be way off line since my memory isn't perfect and it's been awhile since he posted....he might have some ideas eitherway.

Nick's probably the best on here when it comes to actually testing these things out so you can do much worse than following his pointers.

I traded an open gap system for over 6 years on various indices using futures. I don't do it anymore and I never did it on shares.

If you're going to use EOD you need use the open price and the anchor the stop to that. You cannot use any other stop method with EOD, unless you're using a stop exceeding the ATR. There was a system touted by Roger Montgomery called AGI which imploded immediately after release because, in my opinion at least, of this issue. It is my opinion that he used an intra day stop with EOD data.

Also remember that an opening gap can be defined differently. In the US they define it as any gap above or below the prior close whereas I define it as any gap above yesterdays high or below yesterdays low.

Nick Radge said:

I traded an open gap system for over 6 years on various indices using futures. I don't do it anymore.

Click to expand...

Why not?

Nick, I assume that you're no longer trading it (the gap open on indices using futures) because it no longer works? (makes perfect sense eh

But thanks for your in-depth answer.

Yes whatever gap strategies articles all seem to warn about the stops.

Can I ask why using an intra-day stop (for backtesting) while finding entries based on EOD data, is not a good thing to do?

For my stop I have the following conditions:
- Close either when prev. close is hit (gap has filled)
- Close when today's close is hit
(thus, always only staying in for the one day at max)

I first read about this gapping system in a book which had tested it on the US market. Has anyone tried backtesting on the US market? I am wondering if its more successful on there because of the higher volatility of the US (compared to ASX)

gvanto said:

I am wondering if its more successful on there because of the higher volatility of the US (compared to ASX)

Click to expand...

What do you base that on?

Trembling Hand said:

What do you base that on?

Click to expand...

Not sure - was thinking perhaps that the more volatile market experiences more gaps?

Ok, I'm gonna test this strategy but with extended stop:

- Hold until at least next morning
- Close when prior day's close is hit

Fingers crossed

ps: you guys daytrade for a living?

gvanto said:

Not sure - was thinking perhaps that the more volatile market experiences more gaps?

Click to expand...

I'm not sure it is more volatile. We have gone up more and down more by about 50% extra over the last year. Which would suggest ASX is more volatile.

I guess if there is more listed stocks as in the US you are more likely to find suitable candidates each day thou.

Gaps decrease as global markets open around the world. Australia and Asia have good gaps because of their immediate reaction to US events. Europe has lower gaps because they see how Asia reacts and also because they were open during the early US trading session.

Very in depth testing of Gap fill trades on the ES(S&P500 Futs) here. Think Edwood posted this one in a thread somewhere. Not sure how old it is.

http://www.mypivots.com/articles/articles.aspx?artnum=43

Thanks guys your help much appreciated!

Working on this gap strategy big time now and will look at using some volatility filtering too.

Luke did you get my message (email) ?

cheers
Gerry

Classic Gap closing today on the SPI. 70 points

Wow nice fill !

Everyone seems to be doing this gap-strategy for indexes - anyone tried it on aussie stocks?

G