Futures spread trading

Why are you guys giving up your ideas on this blog? As soon as you give someone a hint of what you are doing, you're losing your edge and income.

There are plenty of books and research these past few years on spread trading that it will only get harder to make a living.

Smart people don't talk, they research and apply to make money.

I know plenty of people who divulged their strategies and income to make a name for themselves; they are no longer around.
 
Quote from HTmarket:

Why are you guys giving up your ideas on this blog? As soon as you give someone a hint of what you are doing, you're losing your edge and income.

There are plenty of books and research these past few years on spread trading that it will only get harder to make a living.

Smart people don't talk, they research and apply to make money.

I know plenty of people who divulged their strategies and income to make a name for themselves; they are no longer around.
More...
The devil's in the detail.

Spread trading is not easy, and there's none of that "holy grail" bullshit to be assumed here.

In addition to the ideas on this thread, there's still a good amount of real quant work that needs to be done in order to trade this stuff with any edge at all.

And .. none of that has been discussed here.

Check out the propeller-heads on wilmott if you're into that kind of thing. Occasionally they blert out something worth more than nothing.

And the posts are shorter. Mostly.
 
Knowing how things are done isn't edge, it's just standard, tried and true statistical methods (Hamilton's "Time Series Analysis" is a brick of a book with very good reason). Edge is the stuff above and beyond that truly allows you to distinguish the signals generated from white noise. Well, that and risk management. I speak (at least in part) as one of the propellerheads on Wilmott.
 
Quote from Martinghoul:

Knowing how things are done isn't edge, it's just standard, tried and true statistical methods (Hamilton's "Time Series Analysis" is a brick of a book with very good reason). Edge is the stuff above and beyond that truly allows you to distinguish the signals generated from white noise. Well, that and risk management. I speak (at least in part) as one of the propellerheads on Wilmott.
More...

See, I said the posts were shorter. Mostly. :p
 
Actually, in terms of "edge", competently automating shorter term spread strategies is a significant part of the puzzle, as is refining it intelligently once built. - The devil in the details.

This is not a job for a traditional trader: Not enough tech.

Nor is it for an IT dude: Not enough markets n math.

And quants might not have the patience to build out the fiddly bits, nor monitor the damn thing all day.

So .. you need a trader/IT/quant skill set in one, which is hard to find, but I think the ideal candidate tends to have an electrical engineering background. Applied signal processing, and some pragmaitc coding skills.

Something many of the IBs struggle to find, and struggle even more in justifying the hire.
 
I would disagree that an "edge" in futures spread trading should necessarily require shorter timeframes for sampling and shorter position holding periods.

I do believe that there is "edge" in the spread combinations and construction, and there is "edge" in a good trade entry model and position management strategy. And there is "edge" in developing said models and strategies specifically for spread markets.

Generally speaking, a well-constructed spread combination will almost always "behave" better and model better than the analog singular flat price instrument.

You simply do not have to operate as an arbitrageur in order to successfully spread trade futures. In fact, in terms of my own personal trading I can acheive Sharpe Ratios above 5.0 and max monthly drawdowns averaging less than 40% of an average monthly net return swing trading futures spreads. And that ain't half bad. The SPAN overnight margin credits makes it the cheapest futures leverage going these days.

I used to think up until a few years ago that I needed to to trade 400K RTs per month on a HF basis in order to make money - and now I know that emphatically not to be the case.

And for alot of folks willing to give it a sincere effort, it sure as shit is alot easier and more consistent and less stressful than scalping ES or CL or 6E all day long. At the very least, you would be expanding your capabilities and giving yourself more flexibility. The bottom line is in fact the bottom line - does your bank account really care if you made it scalping versus spread trading ?
 
I have a couple of questions:

1. Since spread trading is relying on the widening or narrowing of a spread, stop losses on each leg are not practical? Rather one should exit with a loss manually if the spread widens too much when we are betting that it will narrow?

2. What time-frames are usually used? Daily? Minute?

3. How does one find spreads? First you have to find highly correlated markets? Do you chart them both and look for divergence?

4. Do you take profits when they go back to normal? Is it called 'fair value'?

5. What is balancing legs? If I were to spread EURO vs GBP, since EURO is 12.50 and GBP is 6.25, would I use 1 euro to 2 GBP? Or since GBP moves more I would leave it at 1 to 1?

6. I was doing some research and it seems that Agricultures are heavily spread traded but FX is not. Why is this so?

7. In the picture, would I go long the top one and go short the bottom one? Is this how you find spreads, look for divergences?

attachment.php
 

Attachments

You're thinking about some of the the right problems ;o)

Edit:

1. Yeah, but this can be automated.
2. What timeframes are the combinations reverting in?
3. May need more math than that if you want to win.
4. Yes, & loosely.
5. Depends if you want to be dollar or vol (volatility) neutral.
6. FX is spread traded, although those smelly fx guys might not talk about it in those terms.
7. Short the leader, long the lagger for convergence. I can’t comment on divergence trading.
 
Quote from Rationalize:


3. May need more math than that if you want to win.
More...

What do you mean by may need more math? You mean finding value? How do guys do that, I hear a lot about finding relative value or "fair" value, but everyone seems to do it a little differently....which then makes it not value(if they're all different?).
 
You must log in or register to reply here.
Back
Top