How do YOU handle full size and mini contracts?

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
jklatt
Roundtable Fellow
Roundtable Fellow
Posts: 79
Joined: Sat Feb 17, 2007 11:51 am

How do YOU handle full size and mini contracts?

Post by jklatt » Mon Jul 12, 2010 5:25 pm

Now that the bulk of my data mining is done, I'm building my portfolio in CSI and I've run across the decision of having to choose between using both the full size and mini contracts in my portfolio or choosing one or the other.

Knowing that I plan to start with a relatively small account size ($75kish maybe more), I'm leaning toward wanting to trade the mini contracts when available (provided they meet my liquidity requirements).

Each contract type seems to have its own benefits:

1) Full sized contracts tend to have a longer history to test with and can be more liquid than their mini counterpart.

2) Mini contracts' size are probably better suited when trading a small account but they typically have far less historical data to test with and can be far less liquid at times.

What do some of you guys do? I think ideally I'd like to include both the full size contract and the mini contract in the portfolio and have Blox give priority to the mini contract if entries are generated for both instruments at the same time. Maybe create a switch to turn mini trading on and off so when the account grows in size, it switches to full size contracts? Is this possible?

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Mon Jul 12, 2010 7:53 pm

In some cases the mini contract is so much more liquid than the fullsize, that most market participants only trade the mini. Even the big boys only trade the mini. Three classic examples are the nasdaq, the midcap-400, and the russell-2k. Some would argue the mini S&P is in this category too.

In other cases, the mini is illiquid and so everybody trades only the fullsize, such as the "miNY" contracts for NYMEX products.

Finally, there are some situations where both are reasonably liquid. In those cases, I try to trade both.

Chris67
Roundtable Knight
Roundtable Knight
Posts: 1046
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 » Tue Jul 13, 2010 7:43 am

I trade a mixture of both depending on - as Sluggo correctly identifies - liquidity
Remember also that if a full size contract has a longer history then you can alsways backtest using the fullsize and generate orders / trade using the mini - the results will be pretty similar - although NOT identical - but this at least gives you a good indication of how your mini is likely to have tested out over a longer period of time
It is quite facinating how some mini's trade so much more than the fullsize - Mini Dow , and sp400 Midcap spring to mind -

trader20
Senior Member
Senior Member
Posts: 37
Joined: Fri Mar 26, 2010 10:57 pm

Post by trader20 » Fri Jul 23, 2010 1:54 am

I do what Chris67 suggests, test and run daily order generation on fullsize contract (with corresponding adjustments for big point val) but execute on mini.

I actually trade QH and QU (miNY) that sluggo mentioned. Usually there is NO(!) bids or offers. However, I have discovered that if I submit my bid or offer at about $100-$150 away from where the fullsize is currently at, I can usually coax the arbitrageurs to take my trade. For example, if I want to buy 1 QH, and HO is offered at 20240, I can get someone to hit my bid on QH at around 20300 ($126 above HO's best offer). It might take some time but so far, I've always been able to get a fill, eventually. That's on a 1-lot. The situation is even better when trading a 2-lot, since 2 QH = 1 HO. It's an instant profit to the arbs. In that case, I might only have to pay $50 or so (20260 in the above example).

This would not work if you need to use "in the market" stops. Worth considering if you want/need the granularity, typically only trade on the RTH open or close, infrequently, and you can afford $150 slip. Note that the miNY's are cash settled, so no delivery worries.

The other mini's I actively trade are QM, QG, J7, E7, YG, YI, QC. No problem on any of these - all have reliable reasonable bid/ask spreads, IMO.

nqwr
Contributor
Contributor
Posts: 1
Joined: Thu May 06, 2010 9:01 am

Post by nqwr » Tue Jan 18, 2011 6:50 pm

Here is the approach I have taken with the Big vs Mini contracts in equities. Happy for anyone to point out any flaws.

ES (CSI #487) vs. SP (CSI # 290)
- ES started trading on 9-9-1997, but the data in CSI goes back to 4-21-1982, which is the first day the SP contract started trading
- It appears that CSI just prepended the SP data to ES, however there is no footnote to this effect on CSI's Futures Fact Sheet
- Running the two contracts side by side, they track each other very closely, except for 6-16-2000 where the closes are 23 index points apart
- So I decided to use ES for both backtesting and order generation

DJ (CSI #496) vs. YM (#699)
- YM started trading on 4-8-2002, but data goes back to 11-07-1997, which is a weird date as it is neither the start of the DD contract (3-20-2006) or the DJ contract (10-6-1997)
- YM data also gets squirrely prior to 4-8-2002, as has been pointed out here viewtopic.php?t=7713.
- Again, there is no foot note in CSI's Futures Fact Sheet to explain what is going on
- After 4-8-2002, YM and DJ closes track ok.
- So I decided for back testing, to use DJ, adjusted so that it has the Big Point Value and margin of YM
- For order generation, I use YM, as I don't want to trade "X-number" of DJ contracts by accident because I forgot that DJ really equals YM.

The pre 4-8-2002 YM data has been discussed on this forum, but I have not been able to find any forum postings that have addressed this with CSI. I think Sluggo once stated something to the effect that data more than 3 months old is probably correct because other eyes have reviewed it ....That sounds very reasonable, but in this case I'm not sure. Given this is the YM contract, I thought I would have found a forum posting with the definitive explanation. I'll wait a few days to see if anyone points out what should have been obvious, or how use properly use the forum search function, and I don't hear anything I'll follow up with CSI.

ND3 (CSI #2395) vs. NQ (CSI #497)
- NQ started trading on 6-21-1999, but the data goes back to 4-10-1996 when the ND contract started trading
- Again, It appears that CSI just prepended the ND3 data to NQ, however there is no footnote to this effect on CSI's Futures Fact Sheet
- Running the two contracts side by side, the closes track each other very closely, except for 8-23-1999 where the closes are 100 index points apart, and on 1-6-2000 where the closes are 32.5 index points apart
- So I decided to use NQ for both backtesting and order generation

MD (CSI# 104) vs. EMD (CSI #684)
- EMD started trading on 1-28-2002, but the data goes back to 2-132-2002 (start of MD contract....)
- Looks like the old prepend without a footnote again......
- The closes track well, so...
- I decided to use MD for both backtesting and order generation

And finally.....

HSI (CSI# 119) vs. HMH (CSI#649)
- I don't seem to be able to find when HMH started trading, data goes back to 10-9-2000
- HSI data goes back to 12-29-1989, again can't find first date of trading
- In this case, CSI does not appear to be prepending HSI data to HMH. Makes me wonder if there is a reason for this or not......
- But HSI and HMH closes track well except for 9-13-2001 where they are 240 index points apart. Given the date, this seems feasible...
- So I decided for back testing, to use HSI, adjusted so that it has the Big Point Value and margin of HMH
- For order generation, I use HMH

I have not been using any of the Russell 2000 contracts, as that seems like a mess to figure out, and other things have always been more pressing.

I am using big contracts for energies and currencies, but might take another look at the mini EUR/USD, EX (CSI# 499) instead of CU (CSI#524), perhaps might take a look at mini soybeans and corn as well.

Mats
Roundtable Knight
Roundtable Knight
Posts: 133
Joined: Thu Feb 28, 2008 2:59 am
Location: Stockholm

You must look at each mini contract..

Post by Mats » Thu Jan 20, 2011 5:31 am

-- to see if there is enough liquidity, especially in NYMEX some of the minis does not trade at all.

I had the same problem as you in the beginning and i traded both minis and full contracts on the same commodity depending on the value of my portfolio. It works quite well but the main rule as stated before is important = "look at liquidity".

Good luck..

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Thu Jan 20, 2011 8:29 am

I agree with Mats, it's crucial to include liquidity as one of your criteria in choosing what to trade and what not to trade.

For example, you chose the fullsize Midcap SP400 (CSI "MD") rather than the eMini (CSI "EMD") -- yet the volume on this contract is extremely low, see image. Maybe low volume is an indication of poor liquidity.

A few trading systems use ONLY the Close price when making their trading decisions and placing their orders; for these systems, your observation "the closes track well" is sufficient. However, the vast majority of trading systems use Stop orders or MarketOnOpen orders, so they require accurate values for the High and Low (Stop orders) and/or for the Open (M.O.O.). Additionally, many systems use "Average True Range" as part of their calculations; this requires accurate values of the daily High and Low prices.
Attachments
Midcap_400_Fullsize.png
Midcap_400_Fullsize.png (28.57 KiB) Viewed 2711 times

Post Reply