Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
I suspect that this is a fools errand but I don't know enough to be sure that it is..
I watch the calendar spread between ES contracts and I see that the spread fluctuates constantly.
I assume someone is making money when this spread gets just a bit too far out of range and subsequently mean reverts.
*Is* someone making money? Who? How much are they making? Why can't it be me?
It's frustrating because this is the kind of strategy that performs perfectly in a backtest, but fails in realtime because it's all about timing..
Anyway if it's not too boring for the veterans, I'd like to know the story here... I have considered jumping through all the hoops to figure out which colocated or VPS server provider to buy, which data feed to purchase, which broker to work with, etc, but I suspect after all of that, any time my algorithm detects the "buy" signal and sends in an order, the position I end up with will just be randomly located on either side of the mean, and then when exiting, the position I exit at will be similarly random.
So, please tell me what I need to know so I don't waste time trying =)
Lots of people trade calendar spreads, more so in the grains and energies, where supply/demand shocks frequently change the prices of one month as opposed to another.
The CME offers "exchange traded spreads" as a tradable symbol, which allows you to simultaneously sell June ES and buy Sep ES for example.
Different fundamentals drive the differences in prices between contract months. Interest rates, cost of carry, etc. Stat arb algos typically keep these differences in check.
If you are backtesting the spread symbol, if you apply proper slippage and commissions (commissions are double), you might have a way forward.
If you are backtesting the individual component symbols, you might just be seeing data timing differences, and not anything tradable.
I will say if you see a "perfect" backtest, that sounds too good to be true. And 99 times out of 100, it usually is.
First off I apologize for the "REMOVED" links below, as I am so new here it won't let me post links (even to threads on futures.io). Oh well. Sorry about that.
Great feedback, thank you. Perhaps the backtest I imagine is perfect because I am ignoring the important thing you pointed out: slippage. Oops.
I signed up to CME Group to look for the ES calendar spread symbols but they must be buried somewhere, not coming up in my initial search. I did find energy spreads and other types of spreads
Anyway your comments have interested me in exploring this idea further - but, as I'm relatively new to all of this, it's not clear which platform(s) would be a good place to explore. This old thread - REMOVED - seems to be pointing me in the direction of TradeStation or MultiCharts.. and some other research I did is pointing me towards Rithmic VPS/dedicated servers... and a broker?
I'm also kind of interested in writing my own trading software, I have decades of experience w/C++ and I sense that every millisecond matters in a trade like this.. this thread - REMOVED - has me thinking about running SierraCharts..
I wish I could just cut through the noise and get a recommendation for a dedicated or VPS linux machine in the CME Aurora data center with 0 hops between my software and the live order book feeds. I'd need to leverage a C++ library to deal with connecting and parsing the data and sending orders... and I'd need a broker.. and probably other things I'm not considering.