LME new electronic contracts

TWI

Senior member
Messages
2,552
Likes
263
On 21 August 2006, the LME announced its intention to introduce LMEminis, which
are smaller-sized, cash-settled and electronically traded LME contracts.

The Exchange has now confirmed that LMEmini contracts for Primary Aluminium,
Copper Grade A and Special High Grade Zinc will be launched on Monday 20
November 2006.

There will be an LME and LCH.Clearnet LMEmini fee holiday up to, and including,
29 December 2006. LME and LCH.Clearnet fees for 2007 will be announced towards
the end of the year.
 
Hi TW

Do you have links to any more information?

The LME website is really poor.
 
I have no more info than this. I already trade 3month electronically these days , never ever call the brokers. It is offered through marex and sempra via select. Thing is after a few weeks you end up with a card full of positions on different dates that you then have to send to a broker to clear off incuring additional cost.
I had a load of Dec options and was hedging the gamma in 3 months for a while, did not keep close eye on spread and ended up costing me a bundle. These new contracts will be monthly expiry, just like US normal futures market. It is going to make playing the LME/Comex arb so much easier.
Will let you know if I hear more.
 
Sounds like you have a lot of stuff on the go. Must be a nightmare to manage.

As soon as something gets easy to arb electronically then a stat arb shop will set up a black box
and routing that is absolutely optimised for latency and costs. Hence, rather than becoming
easier it becomes harder unless you are willing to stand the costs and work involved in being
the most efficient at doing the arb.
 
a stat arb shop will set up a black box
and routing that is absolutely optimised for latency and costs

Yes, I hope this will be me.
I have made good money playing this arb. when trading copper always just look which is cheapest or more expensive at times of opening/closing and then end up with spread book.

I do cover a lot of markets but some are fully automated, a lot are manually executed but signal automated and some are common sense relative value. I thank Reuters 3000 link with excel for keeping things in order, without which, it would be a complex nightmare.
 
Good luck.
Stat arb is execeptionally competitive and there are groups that are very skilled but then I am sure you know that.

I have a similar mix but without any discretionary part and probably far fewer markets.
However, I find the hand executed stuff is a real pain compared to the fully automated.
Are you also doing risk control and portfolio allocation in Excel?
 
Most of the risk control is handled in TS via EL but I use Rina to get a correlation matrix to define cross product risk distribution and so define % risk per product or system. Rebalance every month.
You are right about the manual execution but in a lot of the commodities it is the only option with current liquidity. Stop Limit orders have no guarantees of a fill and Stop Market ends in excessive slippage. Very hard to code the manual executions where I can see depth and place a 50lot in 2lot clips. Overall the manual stuff has better net slippage over time than the automated. Problem is trying to keep up with it all and with things running 24/7 it does require some dedication to the task. Slippage remains my major concern at all times and is what really makes or breaks the PnL.
 
I tend to use the Rina correlation matrix just to make sure everything is not particularly correlated.
Have given up using it as part of the weighting strategy as this just leads to over-fitting IMO.
If I were doing a lot of LTTF then I probably would start it into account.

I can see the problem using TS and EL as this is very restrictive. I coded my own execution platform so
providing there is electronic access I can knock up an algorithm to get into the market with minimal slip.
For example with stop-limits, I have written a few execution classes that place native exchange orders but will then run a sequence to make sure any unfilled or part filled orders are completed within a set time and at the best possible price. WIth good algos you should be able to improve on manual execution.
 
Your stuff sound interesting and certainly more advanced than what I have been able to accomplish with TS. would like to talk about this at some point as it is an area I need to explore.
As for using the matrix, it is no problem once I dump it into a spreadsheet and update one variable in each product. May be curve fitting but it is hard to know what is and what is not whenever anything is defined based on historical data. I tend to er on the side of knowing why I am doing something and so long as I have a reason that sounds reasonable then I think i am doing better than picking an arbitrary value. In this case it is done every month so it is fairly dynamic curve fit at worst and appears to work better than finger in the wind. with this many products I have not yet figured out a better way.
 
The problems of Markowitz have become too clear to me this year.
As a result I have come up with a quite radically different approach and conceptual framework
for allocating to a portfolio of trading systems. The resulting weights are not too different but
the approach is much more flexible, robust and seems to give improved real trading performance.
However, it is still a work in progress.
 
Sounds good, or it would do if I had the faintest idea what you were on about ;) I'd hazard a guess that "stat arb" means statistical arbitrage but apart from that no idea.

Anyway, I saw the news item about these new "e-Minis" from the LME but I was wondering about brokers who offer online access to LME Select. Any ideas?

Thanks :)
 
I have a single platform through Marex which give access to LME, ICE, LIFFE, CME, eCBOT, eNymex.
LME full size contracts can be traded on "select" platform which is available through a lot of brokers. Last time I put a base metal order through floor was a long time ago.
 
Top