Q on when others roll new old crop to new crop

General discussions about futures.
Post Reply
MarkS
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Wed Dec 15, 2010 12:10 pm
Location: Chicago, IL

Q on when others roll new old crop to new crop

Post by MarkS »

Just got a roll alert based on my settings to go from July corn to Dec. But there is almost an 80 cent disparity in prices, and I think there is a potential for the trend to split along new/old crops....

Should I override the system and force the CSI data to stay in July or roll to Sept, or should I take the signal and roll as specified? Anyone trading corn now -- what are you doing/planning on doing as July comes to a close -- Sept or Dec?
Chris67
Roundtable Knight
Roundtable Knight
Posts: 1052
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 »

(what I do is) FOLLOW THE SYSTEM
sluggo
Roundtable Knight
Roundtable Knight
Posts: 2987
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo »

An earlier post of yours discussed grad school projects studying commodity price correlations. How did you handle rollovers in school? Can you do the same thing now, in real trading?

You also mention
MarkS wrote: ... can happen to spread traders. Ask calendar-spread traders in the wheat pit from 2-3 years ago... You won't find them in the pits, because after 20+ years of trading the spread the same way, things changed. From the story, as I was told from a floor trader that stood in the pit, three of these gentlemen lost over a cumulative $50M+ and all "blew out" as they kept putting on larger and larger positions while they waited for a mean reversion that did not come.
Those people probably don't trade via computerized mechanical systems, but yet they may be able to offer you some general advice about rollovers.

And finally, as mentioned just above, why not trade what you test and test what you trade? Roll the same way with real money trades, that you rolled in historical simulation?

Or maybe you are starting to doubt the way you rolled in historical simulation. In which case, you'll want to (A) figure out a rollover plan that is comfortable for you; (B) change your historical continuous contracts for testing, so they roll over this new way; (C) run all your backtests anew, with the new continuous contracts + new rollovers.
MarkS
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Wed Dec 15, 2010 12:10 pm
Location: Chicago, IL

Post by MarkS »

Thanks guys. My issue isn't on how to do the rollovers but on whether I should over-ride my system this one instance due to unique fundamental issues as opposed to technical ones; fundamental issues which may continue to drive a bull market in July/Sept contracts but not in the December into which I am to roll on Monday.

But I see also what both of you were saying: my historical tests have always rolled me forward in this manner, and I am happy with those historical results, so I should do the same here. I've always been a discretionary trader; I find the shift to the systematic approach is easy 90% of the time (as Chris says, "follow the system"), but psychologically very difficult the other 10% when the voice in your head is telling you to do something else (e.g., this corn situation, or not taking profits in silver when it went parabolic last month and my stop was a full $8 lower!)

As for those wheat guys, I had/have friends in the pits, but not those guys in particular. Would have no idea how to contact them at this point.
ecritt
Roundtable Knight
Roundtable Knight
Posts: 134
Joined: Sat Aug 28, 2004 3:44 am
Location: Phoenix, AZ
Contact:

Post by ecritt »

This is why I don't use "rolled" contracts. The Dec and Jul contracts represent different products with different supply/demand realities. Why must they be treated as if they are mutually exclusive? You could spread your risk across both of them. In fact, you could spread your risk across all the active corn contracts...trade the whole futures curve. Of course, over time this will give you very different results from always cramming everything into the contract with the highest concentration of speculators.
sluggo
Roundtable Knight
Roundtable Knight
Posts: 2987
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo »

ecritt wrote:In fact, you could spread your risk across all the active corn contracts...trade the whole futures curve. Of course, over time this will give you very different results from always cramming everything into the contract with the highest concentration of speculators.
You could do this.

It may be useful to go through your portfolio of markets and check to see what fraction of them can possibly be traded this way. For some of them (e.g. interest rates, stock indices, currencies) there is little or no old crop / new crop distinction between delivery months. For others, there isn't sufficient volume in the back months to trade them comfortably. The CSI UA feature "Nth Nearest" might be a handy tool to help in this investigation. You could compare front month volume/OI vs 2nd nearest vs 3rd nearest vs. (etc).

You may discover that 60% of the markets in your portfolio can "trade the whole futures curve". Or you may discover that only 15% of the markets can be traded this way. Knowing the number may influence your "is it worth going through the effort" decision.
Attachments
nearest.png
nearest.png (38.39 KiB) Viewed 7167 times
rgd
Roundtable Knight
Roundtable Knight
Posts: 396
Joined: Thu Aug 10, 2006 5:21 pm

Post by rgd »

I much prefer rolling in a backwardated market. Theoretically, you get paid for being long.

Absent that, there seems to be a certain security in following the system, no over-analyzing, no second-guessing. Over time you realize that the "wisdom" of the system prevails more often than not. If you really feel there is something there, test it and find out for sure.

This all brings back memories of rolling Minneapolis Wheat in 2008 when the Mar/May was at $1.50 inverse, only to watch it go to $4.00 a couple days later. Of course, all this was near the top anyway.
MarkS
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Wed Dec 15, 2010 12:10 pm
Location: Chicago, IL

Post by MarkS »

Eric -- very novel (to me) way of approaching it. Almost like going long the cash or spot and storing it versus playing different crops/delivery dates.

Sluggo, you make a great point of effort versus contracts. The majority of contracts I trade are financial. And for some, I have no idea on differences -- for example, is there a main "slaughter month" in hogs or is it just done consistently. I know for a fact wheat, corn, and beans all have old/new crop distinctions. Something I am going to put on the "to-do" list in time.

rgd -- going with the system is what I ended up doing. I am managing the money of a few friends, and they invested in the system and not my ability to second-guess farm fundamentals; I poised the question to them last night and let them make the choice. The system won out, so I rolled to December this morning.
Chris67
Roundtable Knight
Roundtable Knight
Posts: 1052
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 »

THIS weeks price action in July and Dec corn has probably answered the question on what to do when the system says roll from July to Dec
The spread is irrelevant ?
AceofAce
Senior Member
Senior Member
Posts: 49
Joined: Thu Sep 30, 2010 5:06 am

Post by AceofAce »

How about rolling the turtles way - Roll to the strongest contract month for the longs and the weakest month for the shorts. Dont roll but close the position if roll would result in a position that would not have opened anyway.

There are a number of systematic ways to measure strength (compare number of Ns in each contract month, visual examination, % change since signal was generated etc)
sluggo
Roundtable Knight
Roundtable Knight
Posts: 2987
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo »

Pretty cool graphs from SGX, which they call Pace of Roll (link)

Wouldn't it be great if every exchange provided these for every contract?
Attachments
MSCI_Singapore_Stock_Index_Futures.png
MSCI_Singapore_Stock_Index_Futures.png (152.27 KiB) Viewed 6529 times
Post Reply