All market gains since 1993 have occurred after hours

  • Thread starter Thread starter krugman25
  • Start date Start date
krugman25 said:
So if we want to start talking realism, then we need to reinvest gains. There is no way we are going to be trading the same lot size in 2019 as we were in 1993. I added a few lines of code to take the gains and buy that many whole numbers of shares. The results are a 61% increase in profits by reinvesting, from $68,000 to $110,000.

Not Reinvested
View attachment 201061

Reinvested
View attachment 201062
More...

I added 6 dollars round trip for commissions and fees, and 1 cent spread per share. It's coming out to be about 10-20$ cost per round-trip trade. Does that sound about right?

After adding that the return is $90,000 when using test #3 and reinvesting profits.
 
I see that you ran this on the SPY. My questions are, where is your entry and exit? Since I ran my study on the ES, I looked at entering in the last second before the market closed and exiting one second before the market opened. SPY does have extended hours trading, so its possible to do this, but how does your study take into account actually trading this?

I would love to see what your numbers would be if you ran the test with the same time frame that I used. Either time frame of end of 2013 to present, or in October just before the drop would be sufficient. I'm just curious if our numbers make sense. I already showed that trading this in the ES would not have been profitable once you factor in commissions. Even without commissions, there was barely 1 point per day profit. So I'm curious what your data would say if you ran the test in the time frame that I did.
 
Seaweed said:
I see that you ran this on the SPY. My questions are, where is your entry and exit? Since I ran my study on the ES, I looked at entering in the last second before the market closed and exiting one second before the market opened. SPY does have extended hours trading, so its possible to do this, but how does your study take into account actually trading this?
More...

I used SPY because I had access to that data. I am using daily OHLC data and then converting the O and C into ticks. I then feed that tick data into NinjaTrader8. This is essentially entering the market right at the close and entering right at the open. I don't know how it would affect the results if I moved the entry/exit by a minute or two. I don't have access to that resolution of data, but I would love it.

Seaweed said:
I would love to see what your numbers would be if you ran the test with the same time frame that I used. Either time frame of end of 2013 to present, or in October just before the drop would be sufficient. I'm just curious if our numbers make sense.
More...

I will run a 2013 to present test and post it here. I will do a no filters test, and then run one with the filters from my test #3, since that seemed to significantly improve the performance and reduce trading occurrences.

Seaweed said:
I already showed that trading this in the ES would not have been profitable once you factor in commissions. Even without commissions, there was barely 1 point per day profit. So I'm curious what your data would say if you ran the test in the time frame that I did.
More...

My tests as showing it would be profitable with 10-20$ of commissions per trade. My test #3 with gains reinvested shows 11k in commissions to make 110k profit. I am not a scalper or day trader, but that seems like it would be in par?
 
krugman25 said:
..
My tests as showing it would be profitable with 10-20$ of commissions per trade. My test #3 with gains reinvested shows 11k in commissions to make 110k profit. I am not a scalper or day trader, but that seems like it would be in par?
More...

If it were that easy, everyone would have done it. It WILL NOT WORK. Your backtests are flawed. Not your fault, per se, it is just that it doesn't work in RL.
 
Overnight said:
If it were that easy, everyone would have done it. It WILL NOT WORK. Your backtests are flawed. Not your fault, per se, it is just that it doesn't work in RL.
More...
Aren't you just precious.
 
Seaweed said:
I would love to see what your numbers would be if you ran the test with the same time frame that I used. Either time frame of end of 2013 to present, or in October just before the drop would be sufficient.
More...

I ran it from 2013 to present. This has no gain reinvestment, commission+fees, spread (1c/share), and no additional trade filters.

upload_2019-4-18_21-21-59.png


upload_2019-4-18_21-32-55.png
 
Last edited by a moderator:
Overnight said:
If you believe in it, do it with your own real money, and stop with the backtests. Sheesh! Go on, make a fool out of me. Do the Nike thing. "Just do it."™
More...
"Patience my friend. More testing to do."

I am not into proving people wrong, that makes others opinions way more important than they really are. I don't care about others opinions, I just want to make money. Now shush, I am working here.
 
Seaweed said:
I would love to see what your numbers would be if you ran the test with the same time frame that I used. Either time frame of end of 2013 to present, or in October just before the drop would be sufficient.
More...
Now here it is with all of the filters/rules on. Only trade Monday/Tuesday, only trade if the market has dropped by at least -0.5% in the past 5 trading days, reinvest returns.

It's not completely obvious from the graph but when looking at the trade metrics, this test has significantly better performance.

upload_2019-4-18_21-31-38.png


upload_2019-4-18_21-31-58.png
 
You must log in or register to reply here.
Back
Top