Tradingview Public Library

Well, I actually try to predict what the red line is going to do. This is because the machines will not allow the price to drift from the red line. The basket will trade one one side of the line and this will start to pull the line up or down, then it reverses very strongly.

It's easy to see at 10:45 trading above line stops and yellow crosses over the blue line signal end of rally (sell). Later at 15:10 (the red line has flattened) the value of the basket rally from under line to above and yellow crosses blue signalling start of new rally.

The price is being traded by robots that predict the path of red line and try to lead it. When the robots are 'sure' about the direction of the red line they will buy or sell aggressively. I think it is because they know that trend is slow to turn around and it will be safe to trade.
 
Last edited:
Real Money said:
Tradingview is cool, you know I might just try to get something going with that because they have intra-day spreads, but (it's not free) you have to subscribe to the data feeds and the website.

ThinkorSwim thinkScript is a free programming language for building studies and is very powerful. You can download free studies online very easily. There is a big open source community providing code and scripts for ToS users.

Here is an example of one that I made. It builds and charts a polynomial regression on the target instrument, then below the chart displays the same instrument detrended against the regression average. Finally, there are a pair of staggered (overlapping sample) moving averages applied to that and viola! You have a denoised, momentum measured, price process that is easy to interpret visually.
More...
Image is of a trend up, then down. How does it do in chop because most squiggly line indicators handle trends well enough but in chop ... they get "chopped" up?
 
Guys, I think you should really carefully backtest whatever strategy you are trading. Just because something works for now doesn't mean it is working tomorrow. Ofcourse the past is not necessary indicating how the future might be but I like to believe that a strategy should at least have worked in the past (without being optimized).
TV is a really good charting package but anything else...
 
SunTrader said:
line indicators handle trends well enough but in chop ...
More...

That chart shows the cash value of a basket of futures (long,short,short,long)

Dow Jones Index Futures
Ultra Long Treasury Bond Futures
Russell 2000 Futures
15-25 Year T-Bond Futures

Technically, this is probably called a rate index fly. This is a heavily spread basket.

A long short basket with those futures is not going to chop. This is because it takes a massive amount of money to move that basket. People who have enough money to move that basket are not dumb enough to try and go against it.

So, this indicator is well suited for this. Spreads like this are very strongly mean reverting.

It's not something you would just put on ES to trade off.
 
I am not sure if I can follow you. Is that a "strategy" you trade or just some indicator you look at to make up your mind what to do next?
If it is a strategy, did you test it?
 
It's just an example of a spread dude. It happens to be a favorite among the big money prop firms because of the risk profile (hedged in equities and hedged in rates market).

My "strategy" is that I measure HFT index arbitrage and that gives me an informational advantage. Then I trade using discretion. Rate and index futures and spreads.

I have a math degree and am partial to martingale betting since the informational advantage gives me a very high probability of success. I'm not a systems trader and I don't use back tested strategies. I prefer forward testing.
 
Real Money said:
I have a math degree and am partial to martingale betting since the informational advantage gives me a very high probability of success.
More...
I don't mean to start the martingale discussion here, just have a question. How do you position size with martingale? So let's say you get that rare outlier event which gives you very high prob of success. However, you cannot load the boat because you know that you may have to double up several times. So you end up with relatively small position on an event that happens very infrequently.
 
I do a couple things.

First I take small profits for a few times because I have high probability.

Then I make a slightly larger bet. If it goes against me I wait......then I make larger bet.

When trading martingale the average price is what is important, because you are increasing your chances of success.

Sometimes better to triple size or more even than double (this can move your average price further from initial trade and closer to the market).

You can just trade small enough in the beginning that you can take more heat before adding. Or you can hedge the first small bet and add after the losing part.
 
qlai said:
let's say you get that rare outlier event which gives you very high prob of success. However, you cannot load the boat because you know that you may have to double up several times.
More...

Well, I think it is better to just have positions long and short. That way when the outlier event happens you have an out sized profit to take, as well as a large loss that is likely to roll down from traders using your strategy (buy/sell when outlier happens).
 
I do not have a problem with Martingale when trend trading with an "always in" system. The basis for reentry during the set should be according to the signal and not some mathematical ladder. The flow is always biased eventually...you just got to be there when it breaks. Jack Hershey/ Grob / Bubba is the only one I have read that talks about wash trades and such..

If Martingale is used to ramp up after a loss in trend trading then you cannot have any breaks...your trades must be consecutive and truly "always in".

ES
 
Last edited:
You must log in or register to reply here.
Back
Top