All market gains since 1993 have occurred after hours

[krugman25]

Link to article ?

Posted somewhere in the first few pages. Take a look.

 

[murray t turtle]

Wow, this I wasn't aware of. Thanks! I don't know if you have any other info for past quarters, but if so please post it here.

That is basically what the entire data set looks like. The hot and heavy selling happens during hours, less of it after hours.

%%
Good points on admitting you were wrong; this sentence also disagrees with you headline, as far as SPY, QQQ.By the way buddy if your whole data set looks like 4th quarter 2018=small sample

 

[krugman25]

%%
Good points on admitting you were wrong; this sentence also disagrees with you headline, as far as SPY, QQQ.By the way buddy if your whole data set looks like 4th quarter 2018=small sample

I think you think you understand those posts, but don't actually understand those posts. He was saying for the quarter ending 2018, the phenomenon was alive and well. He said buy and hold knocked off about a quarter of the gains compared to trading overnight only.

I am not sure where you got that I was admitting I was wrong from that. In any case, this thread isn't about me or whether I am right or wrong. It's a very simple study using daily OHLC data. The only way the study is wrong is if the data was altered. Are you saying the data was altered?

Now what a lot of people here have done is change the parameters and gotten different answers. If you change the parameters of the study then of course you will get different results, but I think that goes without saying. I invite people to post studies with different parameters and results. It's always beneficial to look at the data from different angles. There have been a number of those posts already, and are extremely useful in analyzing this phenomenon.

 

[murray t turtle]

Here is another study to add to the archive. This one shows the phenomenon is alive and well for gold also.

http://jayonthemarkets.com/2015/11/...Wm2Smq9mmYIwlJTIjVsSs8fHPczkW7HNC0ucLJkgKZJtI

%%
That is why i chose my words much more carefully than you do Kman 21/25. I admitted your headline could be correct in some markets[good gold points............-]
Sure the data is altered some, I record my weekly charts about 3;00 CST, reg market close, not settlement price.[Off course i could be off a minute easy+ i seldom change my weekly close to settlement price-maybe i should, but i am not]

Feel free to double check my math, actually i failed 4th grade math,LOL, but did ok rest of time LOL. I seldom use DOW as benchmark; but using your 4th quarter example=NO WAY @ ALL did SDOW make all the gains after hours LOL.

I used IBD data[investors.com] they hardly EVER make an error/charts,. No disrespect to IBD, but i caught them making a data error, on a 3X etf, they had the SDOW make a gain of 0.4 one week ; it was really 4.0%/honest mistake. So yes that data was altered accidentely there,

 

[krugman25]

%%
That is why i chose my words much more carefully than you do Kman 21/25. I admitted your headline could be correct in some markets[good gold points............-]

I am still not sure which words you think I have not chosen carefully enough. If you are talking about the headline, well those are for brevity. I am sure you know you have to go beyond the headline and into the actual content of the post. Actually, I think the issue is that others need slow down and be more careful when they read. About half of the posters in this thread understand what is being said here, how is that? It must have been clear enough for them.

I agree with the spirit of what you are saying, if you change the parameters you will indeed get different results. Again, I think that goes without saying. Not that changing the parameters are bad, as long as everyone understands that as soon as you do that it becomes a totally different study. Once I can get NinjaTrader tech support to help me resolve my data file issues, I plan on doing exactly that, running these tests with a number of different parameters. And if I am feeling really crazy I may pay $500 for the 11 years worth of 1 minute SPY data and have myself a lot of fun.

 

[murray t turtle]

I am still not sure which words you think I have not chosen carefully enough. If you are talking about the headline, well those are for brevity. I am sure you know you have to go beyond the headline and into the actual content of the post. Actually, I think this issue is that others need slow down and be more careful when they read. About half of the posters in this thread understand what is being said here, how is that? It must have been clear enough for them.

I agree with the spirit if what you are saying, if you change the parameters you will indeed get different results. Again, I think that goes without saying. Not that changing the parameters are bad, as long as everyone understands that as soon as you do that it becomes a totally different study. Once I can get NinjaTrader tech support to help me resolve my data file issues, I plan on doing exactly that, running these tests with a number of different parameters. And if I am feeling really crazy I may pay $500 for the 11 years worth of 1 minute SPY data and have myself a lot of fun.

%%
Yes, the headline.
AS far as 1 minute charts;
LOL. I seldom use those except............exits. If i tried to get my entry signals off 1 minute candle charts,i would do too much daytrading/commissions/dividend loss/losses.

SPY is such a liquidity leader/benchmark not sure why anyone would be crazy enough to buy 11 years of 1 minute SPY. But since 1 minute is not my timeframe mainly ,could be helpful LOL.

 

[krugman25]

LOL. I seldom use those except............exits.

Exactly right. I want to be able to adjust the parameters of the entry/exit for this overnight-only test. The daily OHLC data only gives you the immediate open and close price. I would like to run the tests with +/- 5, 10, 15, 30 etc minutes from open/close and see how it affects the results. Not to mention, getting a fill 5 minutes after the open is a lot more practical and economical then getting one 1 millisecond after the open, same with the close.

 

[murray t turtle]

Exactly right. I want to be able to adjust the parameters of the entry/exit for this overnight-only test. The daily OHLC data only gives you the immediate open and close price. I would like to run the tests with +/- 5, 10, 15, 30 etc minutes from open/close and see how it affects the results. Not to mention, getting a fill 5 minutes after the open is a lot more practical and economical then getting one 1 millisecond after the open, same with the close.

%%
I like hourly + 5 hour sometimes; but on hourly, have to cut 1 or 2 in half LOL/reg market hours.
I may or may not daytrade more as i get older; a daytrade is good about keeping you awake LOL.

 

[volente_00]

All the articles proves is that gap ups always fill eventually in a mean reverting instrument


So either by and hold or a contrarian. It never chase price

 

[krugman25]

Alright folks, I finally worked through the issues with NT8 back-testing. It was a combination of needing to adjust the "market hours" template as well as massage the OHLC data in such a way that when I turned the open and close of the OHLC into ticks, the NT historical data engine properly processes it.

Below are the vanilla results of the daily OHLC data, converted to ticks, and processed through NinjaTrader 8. This data goes back to the inception of SPY. There are no commissions applied. There are 6600 trades taken, so if you want to run your own commissions + fees + slippage numbers, be my guest.

Now that I have the data working and have these vanilla tests as a baseline, I plan on digging a bit deeper to see if a trading strategy can be had.

Results show that cumulative % returns (roughly) for market hours only = -17% and cumulative returns for overnight hours only = over 700%.

Open Market Hours Only

Overnight Only