All market gains since 1993 have occurred after hours

[murray t turtle]

One thing I find fascinating is how during the 20% selloff last December, it was barely a blip with the overnight only test.

%%
Maybe true;
but not true of major liquid US markets OCT-DEC, 2018. And QQQ,for example looked bearish/bear market ralley downtrend last week DEC. But now we know that saint nick rally was a trend change; back to bull market uptrend.

{Personal note; i like green charts but not 100% green, i like some red , even if one uses blue for green UP,LOL}Thanks/100% green; @ least i could read them

 

[cruisecontrol]

Yes, this is my thread and I started this thread for a specific reason and I will be as inviting or uninviting as I need to be.

That's not how this works. Get a blog or your own forum if you want to control everyone.

I have to admit I enjoy watching them flail around like a fish out of water.

That's pretty immature sounding, hopefully you are young and will age out of this attitude.

 

[krugman25]

That's not how this works. Get a blog or your own forum if you want to control everyone.

Thanks for the tip.

That's pretty immature sounding, hopefully you are young and will age out of this attitude.

31, and never too old to enjoy it.

 

[Cuddles]

Backtests mean nothing. NOTHING! Only the real world experience of opening a real money trade at the close, and closing it at the open will make a true difference.

And nobody here has the stones to do it with real money, and would rather armchair quarterback it with "backtests".

I wonder why that is?

This sounds like someone who refuses to see the advantages of coding and/or algos

 

[krugman25]

This sounds like someone who refuses to see the advantages of coding and/or algos

Honestly, I get where he is coming from. There are people on each side of the spectrum. One side is that data and testing is the only thing that matters. The other is that data and testing doesn't matter at all. I try to take a middle of the road approach. The data and testing is to reveal something interesting and exploitable that a trading strategy can be built on. But real life trading is where the rubber meets the road and it is proven to have an edge or not. Still, I am not going to risk hard earned cash on a strategy until I have enough data/testing to show it's worth while trying out. That is what I consider middle of the road.

 

[HelloDollar]

Good grief.

1. Other than FOMC operations, most "market moving" news occurs outside of RTH.
2. Since 1993, $SPX has rallied about the same amount as the buy the close, sell the open strategy. The only advantage is being flat for five hours each day.
3. There's a cost to getting flat. Someone who merely bought "the market" in 1993 and stayed long, incurred no commission costs and will be taxed on cap-gains rather than earned income.
4. This data provides us little context as to if this is an index only phenomena. Because of the mega-length of the modern U.S. stock rally (essentially, 1982 on), saying that "something works" for 25 years has great sway. But, take a look at one year rallies in a plethora of instruments, be it Bonds or Soybeans, and you'll see that markets open more often in the direction of the trend than not. Indeed, "they opened limit up" is a lexicon in Big Bull Markets.
5. Hence, I suspect that if in the year 2050, $SPX is printing 330, we'll be talking about the wonders of that "go home short every night and cover on the open" strategy.

 

[krugman25]

1. Other than FOMC operations, most "market moving" news occurs outside of RTH.
2. Since 1993, $SPX has rallied about the same amount as the buy the close, sell the open strategy. The only advantage is being flat for five hours each day.
3. There's a cost to getting flat. Someone who merely bought "the market" in 1993 and stayed long, incurred no commission costs and will be taxed on cap-gains rather than earned income.
4. This data provides us little context as to if this is an index only phenomena. Because of the mega-length of the modern U.S. stock rally (essentially, 1982 on), saying that "something works" for 25 years has great sway. But, take a look at one year rallies in a plethora of instruments, be it Bonds or Soybeans, and you'll see that markets open more often in the direction of the trend than not. Indeed, "they opened limit up" is a lexicon in Big Bull Markets.
5. Hence, I suspect that if in the year 2050, $SPX is printing 330, we'll be talking about the wonders of that "go home short every night and cover on the open" strategy.

Thanks for the thoughtful comments. Here are my thoughts on these.

1. Research has been done on possible reasons for this phenomenon. The theory you mentioned is just 1 of many presented and at the end of the day are just theories.
2. Not really, some of the markets such as the NASDAQ are down significantly when looking at trading hours only. That means it is not just a case of being flat five hours a day, but in same cases it means avoiding non-trivial losses. From what I can see in the data, at best, market hours have added significant risk for no gain, and in some cases(such as NASDAQ), losses.
3. Indeed. This is simply data and research on the phenomenon itself, not a trading strategy. Those things would have to be taken into consideration to see if the improved risk adjust returns justify the added overhead costs.
4. A number of posts have been made that look at both smaller 5 year time frames as well as the larger 25 year time frame. The phenomenon has been present in every case. Of course you can isolate 1 week, 1 month or 1 year, where it has under performed but most traders will be trading for more than 1 week, 1 month or 1 year. I am sure you know that any strategy will not outperform every week or every month, so that is just a straw man argument.
5. If the SPX prints 330, likely every long strategy under the sun would be toast, including the buy and hold strategy which is the baseline that these tests are compared against. This is another straw man.

 

[dozu888]

Backtests mean nothing. NOTHING! Only the real world experience of opening a real money trade at the close, and closing it at the open will make a true difference.

And nobody here has the stones to do it with real money, and would rather armchair quarterback it with "backtests".

I wonder why that is?

it does mean something... tells the pattern when the upward drift actually happens.

real money, of course, I just hold them all the time lol, that's enough stones maybe

 

[Maverick2608]

I assume if you buy SPY on the close every day, you will always be entitled to the dividend, correct?

 

[krugman25]

I assume if you buy SPY on the close every day, you will always be entitled to the dividend, correct?

My understanding is yes.