Spreads Trading

[FTSE Beater]

Hi all

I've recently been asked to start a thread regarding spreads trading.

To give you a bit of background, I attended the Joe Ross seminar back in late 2003 where he talked through his method of trading it. Ever since it's been something that I constantly return to. My hesitancy with spreads trading is that it does require a large amount of capital and the draw downs can be very large indeed.

The reason for setting this thread up is to discuss anything and everything to do with spreads, from what looks good at the moment to how to enter a trade.

So here goes.
Like Joe Ross and most other Spreads traders, I'm using MRCI to scan the possible combinations and highlight any potential trades.

The best chart I can find at the moment is this

Chart courtesy of MRCI​

The chart looked very promising until the recent dip. Having said that it still looks like it has a way to go.

Any thoughts or questions are more than welcome

 

[jimbo57]

I dont know anything about lumber, seasonality etc, but as an oil trader I trade time spreads and inter market spreads all the time. The key factor is not the correlation it seems to me, but convergence. But trading a long time strip, j/u for example, you are implicitly trading j/k, k/m..etc. If the front spread is weak, you should be selling it, if its strong you should be buying it - simple. Convergence in commodity spreads is the over-riding principle at work.

Look at the wti spreads- oil at $55++ and bullish, yet time spread convergence is down and has been for many months.

 

[freestyle]

Hi Ftse beater
I believe in the past that you have said you entered spreads leg by leg rather than as a spread.
I have tended to enter as a spread with Refco Express. This allows control on price but doesn't always seem to get the best price.
I would prefer to go in on each leg but the broker can take hours to report a fill on pit based contracts. Quite often the fill doesn't come till after the end of the day.
Whilst I do get live quotes this doesn;t necessarily indicate a fill at those quotes.
Any ideas please on how to tackle this problem?

 

[gene]

How 'bout limit/or better orders? Also, since you have live quotes, once you get your fill report, you can see if you got your fill where you wanted it.

 

[ERA]

hi freestyle,

You sound like a patient man, you should fire your broker for fills like that.

But some practical advice:

Try and stay away from the open and close of markets, I like to put spreads on when the market is quieter mid session, the fills are quicker . I look for quiet side ways moving market to leg in.

I send the more illiquid order ticket first ( ie back months) , while I prepare the more liquid order. I normally will go with the last trade prices as quoted. If I feel doubtful of a fill I will sometimes improve the order by a tick on the last trade. Usually does the trick. This works for me, again I think everyone needs to fine their own style to suit the broker etc.

If I ever was unhappy with my broker I would seriously consider 'openecry' with it's chat window where you can give a spread order to the broker and ask him to leg in, if at any time you feel the fill isn't coming back quick enough you can always say' yeh! what's up with my fill', like broker asist 'on tap'.

 

[FTSE Beater]

Ah ok peeps, I'm confused already

---------------------------------------------------------

Hi jimbo57

I dont know anything about lumber, seasonality etc, but as an oil trader I trade time spreads and inter market spreads all the time. The key factor is not the correlation it seems to me, but convergence. But trading a long time strip, j/u for example, you are implicitly trading j/k, k/m..etc. If the front spread is weak, you should be selling it, if its strong you should be buying it - simple. Convergence in commodity spreads is the over-riding principle at work.

Look at the wti spreads- oil at $55++ and bullish, yet time spread convergence is down and has been for many months.

I think I understand what you mean. You take 2 contracts far apart from each other and over time the prices will converge on each other?

Am I correct?
I might be running before I can walk but.....

Would I be correct in saying these charts are what your looking for?

....and

Which is excellent, just short the chart and away you go, however you get this chart:

...and the drawdowns could kill you.

----------------------------------------------------------

Hi Freestyle

This is one of the early problems I had. I found that by entering one leg at a time, your filled so much quicker.
The fills will take a time as your dealing with the pit, but Refco should still come back to you with a fill fairly quickly. Not sure about how you get an accurate fill though.

-----------------------------------------------------------

Hi Gene

The problem with limit orders is that if you enter the spread in one go, if you don't get filled the first time they look at the ticket, then it can take a lot lot longer to get a fill than normal.
The only other option is to enter the legs separately and treat each side as a separate trade

This is just my limited experience.


PS: For any other mods reading this, yes I do know that I mentioned legs separating....but it was purely trading talk.....honest :cheesy: :cheesy:

 

[Markus_H]

Check the volume!

FTSEBeater,

Lumber is way too illiquid to trade:
The daily volume in the May contract is below 1,000 contracts, and yesterday only 35 (!!!) contracts were traded in the Sep contract.

Even if the pit traders would let you enter this spread, they would never let you out again, at least not with a profit!

I would stay away from spreads that trade with less than 800 - 1,000 contracts per side.

Markus

 

[jimbo57]

Hello FTSE Beater

Re the charts you have put up, the first is WTI and demonstrates the convergence principle very well. The second two charts are heating oil charts which displays a huge seasonality (summer/winter) as well as sharp reactions to near term weather forecasts - I would not recommend trading HO spreads. The seasonality/weather reaction can override convergence.

On the crude side, the main point I was trying to get accoss, perhaps not very well is that you must trade in the direction of convergence (of the spreads). Crude (WTI) is a great example. Flat price it has been very bullish for a couple of years, but convergence in the spreads is most definitely down and continues to be this way. Yesterday is a prime example, a huge up day in the market, but whilst the front month CLJ05 was up $1.40, CLZ05 was up $1.60.

Many think of spreads as a low risk flat price trade (the correlation between the months taking out much of the risks) - this is definitely not true, it is a different beast altogether.

best regards

 

[Beach Runner]

FTSE Beater wrote:
My hesitancy with spreads trading is that it does require a large amount of capital

Hi Ftse Beater,
A good thread. Much needed. And already some interesting posts. Thanks for starting it. I don't quite understand the above sentence as spreads require far less margin? Not so?

 

[oatman]

There was always more opportunity to spread on the opg and closes as differentials were often distorted.This is the time to put spreads on and take them off when the market settles and the diff returns to "normal". I wouldn't recommend "legging" a spread.The whole point is to trade in a safer manner. If your order is entered as a spread, the price is irrelevant. You're only concerned with the diff. This is the method for daytrading.

 

[FTSE Beater]

Hi Markus

I hope your keeping well
I hadn't looked into the volume on the Lumber, but now you've pointed it out - OUCH!!!
That could have been risky.

----------------------------------------------------------------------------------------------

Hi Jimbo

That's excellent. Thanks for the explanation.
Apart from the WTI, is there anything else worth trading :?:

-----------------------------------------------------------------------------------------------

Hi BeachRunner

Thanks for your kind comments, so far my knowledge has jumped up
If you use the margin on the commodities you 're asking for trouble. From what I've seen (and hopefully someone will correct me if I'm wrong), the margin you can get is higher than what it should be, and a little loss can be huge!!!

------------------------------------------------------------------------------------------------
Hi Oatman

That sounds good for day-trading, however if you're trying to take advantage of the seasonal trend then obviously day-trading isn't any good.
Having said that, even if you're entering a long-term position you could use what you mentioned to get a head start on the price.

 

[Edster]

Hi,
Could someone please explain (or post a link to explain) what exactly IS spread trading? Something that will explain it to a newbie.

Thanks,

Edster

 

[ERA]

Hi Edster,

Try www.spread-trading.com , that's probably a good place to start.

cheers

 

[Beach Runner]

FTSE Beater,

I'm afraid I don't quite understand your reply. You write:

If you use the margin on the commodities you 're asking for trouble. From what I've seen (and hopefully someone will correct me if I'm wrong), the margin you can get is higher than what it should be, and a little loss can be huge!!!

The margin on a spread trade is considerably lower than on an outright. Or are we talking at cross purposes here?

 

[fastnet]

MRCI is an excellent source of information and historic spread charts. They also allow a two week free trial to their subscriber newsletter if you ask them nicely.

MRCI’s site run hypothetical portfolio of certain spreads dating back 15 years. It's very convincing stuff - full of stats inferring 85-95% sometimes 100% repetition of certain specific spreads over this period. However this website is chocabloc with legalese for a reason. The portfolio results are hypothetical. Nothing wrong with that as long as their up front about it. As FTSE-B mentioned it's the drawdowns that could get you into trouble.

Included are figures showing the max unrealised profit and loss at any time whilst the trade was open. Very often the worst unrealised loss was much more than the best unrealised profit. Furthermore the realised profit, ie the result of the trade, is often a small fraction of the worst unrealised loss.

These trades almost always (approx 85-95% of the years) result in some sort of profit rather than a loss. There is a high expectancy with these trades but you'll need a strong stomach to sit through the troughs.

There is no account made for commission nor slippage. A wide range of spreads are included some of which include illiquid contracts such as lumber. MRCI make the short-comings of this hypothetical portfolio very clear. I have nothing but praise for their site.

Reading around this subject over the past few weeks I have learned that a common use of these charts is to scan for likely set-ups. However the charts of the spreads are then traded using TA type techniques to enter, exit and place a stop loss. If this does not increase the overall profitability it will almost certainly dramatically decrease the chances of a catastrophic loss which, as LTCM proved, is more likely than the stats would have you believe.

Even a 99% probability trade is unsuccessful 1% of the time. However the peace of mind these stats brings assumes that the 99 profits are at least no worse than the single loss. MRCI’s stats tell a different story - this small probability loss could be much larger than any single profit which evens out the overall expectancy of the portfolio.

In my mind (and after just three weeks so please bear with me) spread trading is like writing naked options. The probability of each trade is very high and the risk is fairly small in any single trade. However, across the portfolio as a whole there is the risk of a catastrophic loss which will wipe out all gains and then some.

MRCI’s advice is to use their charts to scan for trades but not to enter trades on this basis alone.

Thanks for stating this thread FTSE-B. Spread trading is an area of growing interest on this site if only thanks to Mr Ross and his freebie infomercial.

 

[fastnet]

Okay - considering the post above here's a gentler and more consistent trade.

Buy July 05 Soybean Meal (SMN05) Sell Dec 05 Soybean Meal (SMZ05) both listed with CBOT.

It's gentler because the tick size and volatility are less. This trade would have been profitable in each of the last 15 years.

Seasonal trade entering 7th march and exiting 10th may - again approx.

Close on Friday gave a spread of 3.4 so shouldn't be too much harm done in entering a little later. My stop would be a spread less than -10. Target around 20. last year chart went crazy and spread reached 120 in June /July.

I have no idea how difficult these trades would be to place (liquidity) , how much commission you would pay or whether a broker could enter as a reduced margin spread.

Anyone any thoughts on this trade? Comments on possible drawbacks? ERA? H20? FTSE-B?

Info taken from MRCI - didn't want to post the chart but will send them an email to ask if they would mind my doing do in the future. After all the one thing these threads so always have in common praise for MRCI and their service. No such thing as bad publicity right?

Cheers, FN

 

[ERA]

Hi FN,

Great posts there, I can see you are making good progess with your insights. Keeping things ticking over without information overload, a few comments on this spread.

This spread is a old crop/new crop spread, as harvest is in the Fall. The corresponding exchange minimum margin for this spread is $338 and maintainence is $250, if the spread were an old/old contract month like july/ Aug it would be $203/$150 margin. If we look at comms typically, $10-15 RT x2 , so 20 to 30 buck to place the trade. The outright is SM future is $1114, so 1/3 of the outright margin.

I will comment a lot more on some of your interesting points a bit later and this trade, also interested to hear what other have to say.

cheers
ERA

 

[jimbo57]

Close on Friday gave a spread of 3.4 so shouldn't be too much harm done in entering a little later. My stop would be a spread less than -10. Target around 20. last year chart went crazy and spread reached 120 in June /July.

Cheers, FN

Hi,

I posted earlier re crude, and again I profer only ignorance as I did with lumber re your soya suggestion.

As I read it, and conventions vary betweem market, you suggest purchasing a spread (ie buy near/sell deferred in my convention) and paying somewhat over 3. Your stop is at -10 (ie a possible loss of greater than 13c) and your profit at 20c(ie a gain max 17c). Forgive me if I have misunderstood, but this looks close to a 50/50 trade given the targets - why?

 

[fastnet]

Hi Jimbo - all I can say is that if you saw the chart (that I don't have the bottle to post here - it doesn't belong to me) then you might draw the same conclusion. Maybe I have been a little unambitious with the target. As mentioned last year the spread reached 120+ although this isn't usual by any means.

To be honest I'm not really a believer in targets being fixed multiple's of stops. This logic, if applied would certainly result in a net gain. The problem is that the market rarely lets you decide how to trade. I do use this type of thinking but in a more flexible way. I also move stops to b/e as quickly as possible (just outside a volatility band) once I'm in profit.

Very simply, if the spread dropped to -10 then something would be very wrong and I'd want to be out.

FN

 

[jimbo57]

Many thanks FN for the explanation - I understand now.