Cause & Effect Trading! Cause & Effect Methods of Analysis! | Traders Hideout


futures.io - futures trading strategies, market news, trading charts and platforms


Traders Hideout


Discuss day trading practices and futures trading strategies on this forum for all markets. This forum is also for discussing and reviews for brokers, data feeds, and commercial or third party add-ons




 

Cause & Effect Trading! Cause & Effect Methods of Analysis!

  #5 (permalink)

charlotte nc
 
Trading Experience: Master
Platform: ninjatrader
Broker/Data: NinjaTrader
Favorite Futures: Emini (ES, YM, NQ, ect.)
 
Posts: 388 since Jan 2015
Thanks: 71 given, 965 received

Hi AlexMill,

Sorry to hear that you have had recent financial troubles. In terms of framing your question, I interpret this "Cause & Effect" that you mentioned as a sort of means to an end to rationalize the market. I think you are looking for a deeper understanding of how and why things work the way they do and most importantly you are looking for some way to quantify and determine when x occurs, do y. This is my take on what you are looking for....

So really what you need is a correlation study to tell you what if anything is related. You need this study done to quantify the independent variables (Market does x) then the dependent variables will react (market does y based on x).

If this is what you are solving for, then you need to pursue a correlation study. You will need to pick a hypothesis as a starting point and then get the raw data to analyze to see if your hypothesis is right and offers a betting edge.

While I can assure you that there are no easy buttons here.... there is a process that you can undertake that may help you solve this. So to kick things off, I will give you an example of a hypothesis and show you how you could go about conducting a test to see how correlated something is. This will get you thinking and maybe even help...

Hypothesis: If the Ask volume on the DOM is > than the Bid volume on the DOM by 50% at least, and is higher on at least 75% of all the levels (Not just super heavy on one price level) then it is a reasonable assumption to believe that there is enough resistance on the Ask side to drive the price down. This hypothesis has the following assumptions.
1. The market orders are fairly even and are not significantly skewed to one side or the other
2. The DOM levels are stable and are not just a 10 second snapshot where volume temporarily spiked.
3. The imbalance favoring the Ask side is based on several levels not just 1 or 2 with large orders that may be from a single market participant that might cancel.

Bet: Your specific bet might be that over the next 10 minutes the market will shift lower by at least 5 ticks. Your binary outcome will be if the market moves 5 or more ticks down you classify this as confirming your hypothesis, if the market moves up 5 ticks or more then it runs counter to your hypothesis.

Sample Size: In order to quantify if this works you will need at least 100 trials of this particular flavor. This should be randomly sampled over different time frames, periods, etc.

Success Criteria: In order for this to be a viable edge, you will need far more than just a 51% win rate on your bet. Considering commissions you will likely need to be right closer to 65% - 75% of the time.

How do you test this?

You would need to extract the level 2 data, set thresholds to search for the scenario with the settings I have described above, then flag specific time stamps as meeting this criteria, and mark the price as your starting reference price. I would recommend testing this using market orders not limit order to make sure you can get filled. From here you will need to track how the price develops over the next series of ticks (Does it move down 5 ticks or more before it moves up 5 ticks or more).

I recommend programming to solve for the data extraction, and analysis.... Though you could do a few of these manually looking at charts, it would be hard to get to enough of a sample size that this would be statistically significant.

But this is the type of thing that you are looking for, and the type of method / process and analysis you will need to undertake to see if you can find an edge. *** I am not saying that the specific example scenario I described even works, I am just showing an example of the process...

You should think about trading from the viewpoint of betting. Every trade you enter you are placing a bet. If you don't have a strong opinion about the outcome of that bet with statistical evidence to show you a betting line you can exploit you shouldn't take that bet. In trading most people are literally just saying: "Hey there is a team with red jerseys playing a team with blue jerseys... I think the blue team will win today... Why? I don't know, I just have a gut feeling.

But wouldn't you feel better betting 100 bucks that the Golden State Warriors would beat a college basketball team?

Anyway, that's the process you need to take to be successful. No free lunch, no shortcuts.

Best of luck!

Ian

In the analytical world there is no such thing as art, there is only the science you know and the science you don't know. Characterizing the science you don't know as "art" is a fools game.

Last edited by iantg; June 7th, 2018 at 04:15 PM.
Visit my futures io Trade Journal Reply With Quote
The following 3 users say Thank You to iantg for this post: