Tick vs M1

I just want to hear the forum’s take/opinion on using Tick data vs M1 data.

From the SQ e-Book by Zdenek Zanka, there is a paragraph which he said that he personally prefers/recommends M1 than Tick. Reason being that the back test results are identical in most cases between Tick and M1, and also the huge benefit of using M1 is the 50 times smaller data size which results in much faster strategy development process.

As for me, I’ve been using tick data, but I think using M1 data does have a significant advantage due to the much smaller file size to download and the much faster speed in data mining. It also takes up less space in your HDD especially when you’ve multiple instances of SQ. I’ve 4 instances of SQ running, and the tick data size is enormous. However I managed to overcome that by using junction folder which is a brilliant solution to save tons of space.

Anyway just interested to hear your thoughts, opinions, comments on your personal preference over Tick or M1  🙂

Safe trading…

Regards,

Dave

Ticks if scalping,

M1 only from Asirikuy if else.

I use NT, and they recommend using second instead of tick, as it’s significantly smaller data size.

One challenge in seconds in NT data are built on ticks, and you typically only get ~120 days back of tick data from RT data providers, and buying historical tick data can be both expensive to purchase, and expensive to store (if you have multiple symbols)..

Having said that, I use minutes anyway, as I do higher time frame anyway.

i recommend you to finaly also retest on tick data to see the real past performance aslo.

I am currently using Dukascopy tick data, and i am using Timeframe for developing strategies, tick data for Retest to see the actual performance with real spread, and then M1 for RT.
I am also using Tickstory and Tick Data Suit as a final test to see its real performance on MT4 backtester. All these tick data takes up enormous amount of disk space, so I am thinking of using M1 instead of tick if the backtest results don’t differ much. But of course the downside of using M1 is the lack of using real spread in back test.

i will recommend you to buy an SSD or even better a PCI-e hard disk,

all things related to SQ+MT4 will be only on that hard disk,

this way you will get maximum speed and the storage wont bother you that much,

the bigger the hard disk the better for the upcoming future

Yep, i am using a 128GB SSD as primary hdd, and a 1TB 7200rpm sata hdd as secondary. I have put all SQ related files including the tick data on SSD for best performance while using the secondary hdd as a data store for storing strategies.
But 128GB SSD is alittle too little, so i am trying to make the best out of it to get the best performance of SQ.

Yep, i am using a 128GB SSD as primary hdd, and a 1TB 7200rpm sata hdd as secondary. I have put all SQ related files including the tick data on SSD for best performance while using the secondary hdd as a data store for storing strategies.
But 128GB SSD is alittle too little, so i am trying to make the best out of it to get the best performance of SQ.

For sure, thats smart approach,

try using PCI-e hard drive, you will be amazed! 🙂

http://www.newegg.com/Product/Product.aspx?Item=9SIA3ER1842500

there are cheaper versions too,

but when i saw this one i took it,

a very fast beast, still waiting for SQ4 but well prepared :), i was using a rental machine, but now i’m building one for my self little by little..

For sure, thats smart approach,

try using PCI-e hard drive, you will be amazed! 🙂

http://www.newegg.com/Product/Product.aspx?Item=9SIA3ER1842500

 

there are cheaper versions too,

but when i saw this one i took it,

a very fast beast, still waiting for SQ4 but well prepared  :), i was using a rental machine, but now i’m building one for my self little by little..

Wow, 800GB SSD is like SSD on steroid. Too bad mine is a laptop, can’t install a PCI card.

Sounds like you’re building a hardcore machine!

Trying too XD,

you can always rent a machine if you interested,

SQ4 should be much faster then the current version, so you dont have to have a beast of a machine to run it, but it will help :),

i’m more of a scalping guy so i like faster results and quick testes, thats why im building this machine, i have different sources of tick data and just 2 sources of M1 data so thats why i needed that 800GB PCIe HD.

Trying too XD,

you can always rent a machine if you interested,

SQ4 should be much faster then the current version, so you dont have to have a beast of a machine to run it, but it will help :),

i’m more of a scalping guy so i like faster results and quick testes, thats why im building this machine, i have different sources of tick data and just 2 sources of M1 data so thats why i needed that 800GB PCIe HD.

I tried to look around for server rental but it is too costly to get a high specs server, so I ended up buying a gaming laptop instead LOL…

By the way, talking about scalping, I find it pretty hard to generate scalping strategies with high trade volume in SQ.

https://contabo.com/?show=servers

https://contabo.com/?show=servers

Am generating using selected timeframe, retesting with M1 and final tests (WFM/some robustness) with tick. I think it’s important to use tick for accuracy. I use symbolic link clones now to save space on my main ssd and have all the data on the secondary ssd. ‘Link Shell Extension’ is a good program for creating these.

Sent from my iPhone using Tapatalk

Am generating using selected timeframe, retesting with M1 and final tests (WFM/some robustness) with tick. I think it’s important to use tick for accuracy. I use symbolic link clones now to save space on my main ssd and have all the data on the secondary ssd. ‘Link Shell Extension’ is a good program for creating these.
Sent from my iPhone using Tapatalk

I am using Link Shell Estension as well, but using junction folder. I initally didn’t use symbolic link because it appears to be a shortcut rather than a folder, so I chose junction instead. So assumed symbolic link does the job as well?
I am using junction to mimic the “bin” folder of SQ, and also to mimic the hst and fxt folders for MT4.
It works like a charm, and this saved huge amount of space and hence allow me to put all my SQ instances on my SSD.

Junction folder and symlinks are similar if you are using it on a local machine. I am using them the same as you; to mirror the bin folder in SQ and placing the data on a larger ssd drive. With symlinks you can create links from folders and directories outside your local machine, ie on a network share whereas with junction points, they are only limited to your local drives.Junction folder and symlinks are similar if you are using it on a local machine.

Yep only using on local machine, so junction works fine for me.
Anyway I was thinking it will be great if SQ4 can allow user to specify a central depository folder where all tick data are stored and shared by many instances of SQ. So whenever u install a new instance of SQ, you can simply point to the central depository folder during the installation and that will save the need of using junction or symlinks.