Ceasar2 signal

[timo]

Thanks Nick,

Good to know its randomly allocated. That way sounds the fastest. And means I don't have to change my name to '0x20'.

And thanks Will,

It sounds like ICM book potentially isn't deep enough to support both Tomas's private clients and the ST clients.
The top of book pricing is excellent, but the private clients are soaking it up too quickly, and leaving the ST clients the deeper pricing, which isn't so great.

Maybe it's actually a disadvantage sending our orders to ICM so quickly
Because its' throwing us at the tail end of the private client book queue.
I wonder what would happen if we sent the orders a few seconds later, after the book had been cleared out and pricing restored?

Anyway, looks like ICM is going to be like this for a while, unless they get access to more liquidity, or some of Tomas's clients leave.

Does cTrader use a different liquidity pool, since it's in UK? Does it use LMAX pool?

[WillT]

And let me point out it literally will be such a small fraction that in real terms I doubt it even matters when you have thousands of orders try and close if you are first or 100th i nthe queue. Fact is at that moment in time, the flood of orders will hammer liqudity.

Last week has been good on these closures though - still waiting for a biggen to slip!

[Nick]

So with ICM & C2 it's all about what position we are in the queue.

Seems like ICM have enough liquidity to fill all of Tomas's private clients, and a small % of ST clients without too much slippage.

Nick - what determines a ST clients position in the queue?

Say there are 10 ST clients using C2 and ICM. On a basket of 10 trades. For arguments sake let say your 2 accounts are #1 and #2, and my account is #10.

In this situation you would likely get very little slippage, and I would get considerable slippage.
Even though the time between sending the orders is next to nothing, my trades would be hitting the ICM book 90 orders (9 clients x10 orders) below your orders.

Are the ST client positions randomly determined? Or based on signal join date? Or alphabetical? I'll change my name if it is ROFL ;D

It's completely random. Our copier cycles in tiny loops as your MT4 communicates with our system, so if you happen to be at the front of the loop when the order is placed then you'll get filled a fraction before the others.

[ralf]

..... Or alphabetical? I'll change my name if it is ROFL ;D

If the sorting order is based on ASCII then a single space would be an excellent choice.

[timo]

So with ICM & C2 it's all about what position we are in the queue.

Seems like ICM have enough liquidity to fill all of Tomas's private clients, and a small % of ST clients without too much slippage.

Nick - what determines a ST clients position in the queue?

Say there are 10 ST clients using C2 and ICM. On a basket of 10 trades. For arguments sake let say your 2 accounts are #1 and #2, and my account is #10.

In this situation you would likely get very little slippage, and I would get considerable slippage.
Even though the time between sending the orders is next to nothing, my trades would be hitting the ICM book 90 orders (9 clients x10 orders) below your orders.

Are the ST client positions randomly determined? Or based on signal join date? Or alphabetical? I'll change my name if it is ROFL ;D

[Nick]

The "problem" with ICM is they have generally an excellent spread in EURUSD.

The problem here is that you're looking at the top of book pricing.

There might only be 10 or 20 lots available at that price, as soon as 100 clients try and push through a combined 1000 lots, orders will be pushed down the book.

That's what results in slippage.

[timo]

The "problem" with ICM is they have generally an excellent spread in EURUSD.

I think that's right. As the master is on ICM, he is getting excellent fills with low spread, so that already sets the bar really high to copy.
Whereas when some signal providers use a demo account, or brokers with wider spreads, it makes it easier to beat or match the master.

Still, I'll be watching Nicks 2 test accounts at ICM with interest. Because in theory they should do the same as any of our accounts at ICM, since we are all in the ST boat.
If there is still a big difference between those 2 test accounts, and any of ours, then the only variable then is the actual server @ ICM, or the position in the copier queue.

Nick, do you know what server your test account is on? The MT4 one that is ...

[SpiderX]

The "problem" with ICM is they have generally an excellent spread in EURUSD.

To my understanding, Armada spread on EU is around 0.1-0.2 pips most of the time, plus they promote themselves to be good for news trading.
So should not be that different.

[ralf]

From experience...on CNS/Armada, the slippage is worse than Beeks/ICM.
This is despite the fact that execution times are better on Armada.

The "problem" with ICM is they have generally an excellent spread in EURUSD.

[SpiderX]

I would say it's within the threshold of users that ICM can fill. They're also getting in before everyone on SimpleTrader (as much as I hate to admit it).

It would be interesting to test it on someone like Armada/LMAX. I believe they use different sources of liquidity.

From experience...on CNS/Armada, the slippage is worse than Beeks/ICM.
This is despite the fact that execution times are better on Armada.

Cheers

[Nick]

I understand and it makes sense... But Nick, then why Tomas's direct clients on his server, everyone being on the same broker (IC Markets), do always get equal or even better fills than the master?

I would say it's within the threshold of users that ICM can fill. They're also getting in before everyone on SimpleTrader (as much as I hate to admit it).

It would be interesting to test it on someone like Armada/LMAX. I believe they use different sources of liquidity.

[VictorTous]

Just wanted to update further on this. All baskets on our 2 test MT4 and cTrader are working fine.

I know you all want me to say I've found something, but there is no way we are getting delays in the copier itself. If there were delays, I'd be able to replicate them easily and I simply can't. I need a big basket, during high volatility to close out and create a large time delay, as that is when I can look at the test accounts and work out where the delay is coming from.

Until that actually happens, I'm just sat waiting I'm afraid.

Will, you have developed one of the best (if not the best) signal-copying technologies available for retail traders... So don't worry, I understand that not every aspect can be resolved from your side.
Thanks for your enormous constant effort to keep improving things

The tricky thing about the C2 & C2.1 signals also is that Tomas already had an extensive client base before coming to ST.
So he has many customers he runs the EA for on his private VPS's. So in terms of execution, they will always get filled long before any of us will here at ST

Then my conclusion is that liquidity is not a big issue with these many closing orders in the EURUSD.

I agree on the liquidity matter... It must be something related with real time execution/fills at specific times, and different slippage at every Broker.

- Using the same broker as the master account will give you worse slippage. I know it seems counter intuitive, however you'll get better fills if you are accessing different liquidity than the masses. The majority of clients think the best execution is obtained by matching the master account, but this is not the case.

- The volume of orders makes a big difference People always assume that the FX market is a bottomless pit of counter-parties ready to fill their order at the desired price, this isn't the case. When Caesar closed the basket in question, there were 3,500+ orders that were instantly processed by our copier and distributed to what is only a handful brokers. When that volume of orders is processed at the same moment, liquidity providers have trouble filling them quickly, so it results in a delay. For confirmation on this please call your broker.
[/B]

I understand and it makes sense... But Nick, then why Tomas's direct clients on his server, everyone being on the same broker (IC Markets), do always get equal or even better fills than the master?

[timo]

Thanks Nick,

That should be on a sticky somewhere

The same broker theory makes complete sense. More liquidity will always equal better fills. I'll let Tomas know about this, so he can advise accordingly, as he recommends ICM as the best broker for C2, which I guess is not a good idea. Personally, I'd like to move C2 over to cTrader or Axi.

I think it is his local clients that may be spoiling the party for the ST clients. He has close to 100 (could be more, I can't be arsed counting - http://www.myfxbook.com/members/togr) clients locally, and I assume all these guys get filled before we even get a look in here at ST. And if he is pushing ICMarkets as his preferred broker, then that would logically create a hole in liquidity for any ST clients using ICM. Makes complete sense now put that together with what you said.

So I think the answer for me is simple ..move C2 away from ICM

And you are spot on - you do have the best trade copier in the world, hands down ...

p.s. I won't bother tracking my pip performance with ICM any longer, but if anyone wants to compare pips / performance with me just ping me

[Nick]

@ whiteape - agreed. Anything over 15% slippage is too high for me, and > 30% is getting into outrageous territory :P

The tricky thing about the C2 & C2.1 signals also is that Tomas already had an extensive client base before coming to ST.
So he has many customers he runs the EA for on his private VPS's. So in terms of execution, they will always get filled long before any of us will here at ST - a similar situation to the Viper / Trusted MAMs.

However, considering the results of his customers on his VPSs are excellent, and some of the subs results here are also excellent, I still feel there is room to increase the performance here. Especially when we're using the identical broker and VPS to Tomas, and the only variable is the ST copier. It doesn't seem to be black and white, e.g. NOT all ST customers have massive slippage. It's mixed, with unpredictable and seemingly random results.

I know Will is working hard on the investigation, and with any luck he might find some gremlin in the copier that can be tweaked or adjusted to deliver us more pips and less slippage.

Hi Timo,

We have been supporting clients on our trade copier for over 2 years. During that time we've dealt with 5 different signals that have attracted 300+ clients. As a result we've complied a huge amount of information on how the market/brokers/technology works because slippage always becomes an issue once a signal reaches this level of popularity.


What I've learned during this time:

- Using the same broker as the master account will give you worse slippage. I know it seems counter intuitive, however you'll get better fills if you are accessing different liquidity than the masses. The majority of clients think the best execution is obtained by matching the master account, but this is not the case.

- Slippage is highest during low liquidity. If you try and close a trade during the quietest period of the day, or during high volatility (when liquidity is also thin) you are going to get slipped. This is because there aren't as many counter-parties to your order to fill it at the desired price.

- The difference in VPS latency is negligible. The difference between 1ms and 5ms means virtually nothing in real terms. Obviously a cheap VPS on the other side of the world with 100ms latency is going to have an effect, but the constant debate betwen broker/VPS latency is almost pointless in my view. The broker is 99% of the equation.

- Local copying is faster than remote copying. Running a local copier will give you marginally faster execution than a remote copier because of the additional link in the chain. So it doesn't surprise me that accounts running on Tomas's VPS are getting good execution. However if you were to try and load 100+ accounts onto this setup we'd be back where we started with widening slippage.

- The volume of orders makes a big difference People always assume that the FX market is a bottomless pit of counter-parties ready to fill their order at the desired price, this isn't the case. When Caesar closed the basket in question, there were 3,500+ orders that were instantly processed by our copier and distributed to what is only a handful brokers. When that volume of orders is processed at the same moment, liquidity providers have trouble filling them quickly, so it results in a delay. For confirmation on this please call your broker.

We will continue to work diligently on improving our technology to give the fastest execution possible, however I can assure you there isn't a faster, or more stable remote trade copier in the world.

[ralf]

..... Tomas already had an extensive client base before coming to ST. So he has many customers he runs the EA for ..... the results of his customers on his VPSs are excellent .....

Then my conclusion is that liquidity is not a big issue with these many closing orders in the EURUSD.

I have seen that Caesar2 closes baskets repeatedly when sudden price moves happen. And the system opens and closes baskets during news events too. During these fast moves the system seems to close the baskets if an (possibly fast) retracement against its position happens. That is my impression when comparing the closing prices including the time stamps of the system and my two subscriptions.
So, every second counts in such an situation.