Going offshore to escape the CFTC

Broker Discussion Forex Brokers
5364

I have traded on velocity and citifx and they have the option of aggregating the prices for you based on your order size, FCIntl is also a very good broker but they have always favoured QIB’s and institutions. Never traded with them because I felt that I will be realtively too small in comparison with their other clients and can possibly get not treated as well.

One way of doing that (which you have already figured out, is to slice your order and spread across multiple brokers. At the end of the day every broker / entity / liquidity pool has a tag number or identifier attached to it and all the limits (as per the best of my knowledge) is attached to these specific ID’s. So LP’s treat them independent of eachother.

In this specific case, I believe you will only get diversity of LP ID’s (because all broker’s clients, technically would be a sub-id or same ID as that of the broker). I dont specifically see how it will help you in getting better fills. As far as I see their liquidity is very similar (if not exactly the same) and if you are hitting the markets on both of them independently, its going back to the same set of LP (at best their agregators will give you some internal routing diversification, but not significant). So technically, you are splitting your orders only to be recombined again (atleast partially) at the LP level. Hope that I’m making sense :slight_smile:

Essentially, if you look at Advance Markets (not an offshore broker but a kangaroo broker) you can do size with them but their spreads are not the best ones in industry.

Coming back to your point, I think you will have to both slice your orders and add delays to your orders (say 50 MS or so) if you want true diversification, this will in a way fool the LP’s that quickly move spreads for few MS and come back when they see significant sizes. But, if you are doing a hit-and-run strategy, you will have to be slightly more creative than that.

hope that helps

5365

@MrInvisible
I’d like to hear more about your strategy, but this looks like
lots that far outstrip anything I’m able to consider… or that I
would expect a typical retail brokerage to fulfill, especially
the U.S. friendly ones we have on our list.

Just, in general, is this a scalping approach; where you go
for a handful of pips ? I can understand why you are concerned
about liquidity and slippage; or perhaps even LP’s "backing away"
from their Bids/Offers if these are to be achieved in a single
transaction…

[EDIT] I mean that in my trading a single transaction rarely exceeds
0.1M or a “standard lot” and I rely on multiple fills within that size
range, so I have less of a worry on any single transaction, and
not much concern about slippage…

hyperscalper

5366

Weird… for some reason I can’t quote or reply properly on here.

@whywescalp

You answered my question in detail, thanks for your help.

I also traded Citi Velocity as well as the single bank platforms UBS Fx trader, Deutche autobahn & Barclays Barx. I was a small fry @ FC, but my volume was decent @ >5 yards a month. I can’t complain, they always treated me well. That was back when 150-250K could get you in the door through one of their Introducing brokers compared to 10M now.

Do you think these smaller offshore brokers would be able to setup custom direct/aggregated price feeds with their LPs in larger amounts similar to larger retail brokers?

@Hyperscalper

My target is anywhere from 3-7 pips, some consider that scalping and some don’t.

It’s not an algorithmic arbitrage system that is adversarial to brokers and their LPs by taking advantage of discrepancies/lag in their price feeds. Thankfully, backing away from quotes has not been an issue. Ironically, it was a problem with the much larger Gain Capital.

5367

RETAIL MT4/5 CONNECTION LIMITATIONS

@MrInvisible well, I’ve just standardized on the ubiquitous and cheap
MT4/5 broker connections. We know that they will typically “hold” your
orders, aggregate them, possibly for periods of at least 50 msecs
or longer. This gives them plenty of time to “optimize” fills which are
slightly in their favor.

For lot sizes I trade, I get really excellent precision, but I’m not counting
on any single fill significantly affecting the overall result and, as I said, I
am not using lot sizes which would be expected to generate significant
slippage, plus I don’t “chase price” with Market orders, which is where
"adverse slippage" comes in…

So my question would be, why are you (or are you?) looking for
individual fills of 10, 50, 100 lots in one transaction, or perhaps you’re
not; rather than to distribute those entries across multiple transactions??

hyperscalper

5368

It’s not a high frequency system. It only produces 2-3 trades per day, so when opportunities do occur I need to take advantage. I enter with the full ticket size across multiple brokers and get out. Other then during the lower volume JPY session & around news releases, filling 4-6M with decent spreads has never been a problem with most of the retail brokers I have experienced.

5369

I think you should ask their sales guys to get you in touch with their senior manager or so, and with some negotiations and persuasion you can get direct or aggregated feeds and some preferential treatment (its all baked in the commissions).

Based on what I read from your posts so far, I’m assuming you are a liquidity taker and are sensitive to time. Since you are only trading a few times a day and are putting big lots when you get the opportunity and getting out with 3-7 pips (sort of like a breakout trading or news trading :wink: ) may be not technically so, but in a way similar flow. Some brokers will put that as toxic flow. I would actually merge that kind of a strategy with some simple random cross trading (buy with one broker and sell with another) to conceal your trading pattern

To a large extent, I agree with @HyperScalper because that is a very efficient and proven way to trading bigger size with retail brokers. Slice your transactions over multiple brokers, vary them by few 10 ms to let LP’s feel they are different trades and take liquidity across multiple LP’s and brokers. Essentially you can automate to send 100’s of trades in less than 100 MS to over 10 different brokers and achieve your target size with little slippage. I do that as well when I have to put size on my trade. Although Im not nearly as technically advance as @HyperScalper but I can manage with my little experience in tech and flow management.

I would recommend get multiple brokers and dont ask for FIX connections, as that makes them (atleast internally) qualify you as potentially toxic flow. Put simple automation on MT4 with multiple instances running on a large local dedicated server (VPS might not be enough in your case) and slice out your trades into smaller tranches.

Hope that helps !

5370

MT4 ORDER OPERATIONS ELAPSED TIMES

@MrInvisible sorry, I just can’t get close to your ability to enter
orders quickly, say with 10 millisecond resolution… These are
current elapsed times I’m getting (round-trip msecs) times for MT4
into FinProTrading.com, for example.

My Amsterdam server has ping time of 1 millisecond from FinProTrading’s
MT4 server… This is not some VPS; but a dedicated Xeon server
with 8 processor threads… ?

These are not really “high performance” brokers… ?

[20180817-01:55:33.544(GMT)] OrderOp Buy Limit elapsed: 79
[20180817-02:16:18.148(GMT)] OrderOp Close Position elapsed: 98
[20180817-02:16:18.148(GMT)] OrderOp Close Position elapsed: 98
[20180817-02:16:18.148(GMT)] OrderOp Close Position elapsed: 98
[20180817-02:16:18.258(GMT)] OrderOp Close Position elapsed: 210
[20180817-08:08:59.004(GMT)] OrderOp Sell Limit elapsed: 71
[20180817-08:09:00.632(GMT)] OrderOp Sell Limit elapsed: 63
[20180817-08:09:01.529(GMT)] OrderOp Sell Limit elapsed: 62
[20180817-08:09:02.046(GMT)] OrderOp Sell Limit elapsed: 72
[20180817-08:09:02.573(GMT)] OrderOp Sell Limit elapsed: 92
[20180817-08:09:17.313(GMT)] OrderOp Close Position elapsed: 139
[20180817-08:10:18.043(GMT)] OrderOp Cancel Order elapsed: 88
[20180817-08:10:18.043(GMT)] OrderOp Close Position elapsed: 88
[20180817-08:12:39.531(GMT)] OrderOp Sell Limit elapsed: 132
[20180817-08:12:39.968(GMT)] OrderOp Sell Limit elapsed: 63
[20180817-08:12:40.525(GMT)] OrderOp Sell Limit elapsed: 111
[20180817-08:13:52.805(GMT)] OrderOp Cancel Order elapsed: 78
[20180817-08:13:52.835(GMT)] OrderOp Close Position elapsed: 110
[20180817-08:15:25.725(GMT)] OrderOp Cancel Order elapsed: 81
[20180817-08:15:25.725(GMT)] OrderOp Cancel Order elapsed: 81
[20180817-08:15:25.795(GMT)] OrderOp Close Position elapsed: 149
[20180817-11:48:21.859(GMT)] Buy Market OrderOp elapsed: 140
[20180817-11:49:10.849(GMT)] OrderOp Buy Limit elapsed: 72
[20180817-11:49:14.559(GMT)] OrderOp Buy Limit elapsed: 71
[20180817-11:49:16.409(GMT)] OrderOp Buy Limit elapsed: 69

hyperscalper

5371

MT4 AGGREGATION PHASE

I had a situation with an 800 millisecond order operation time (round trip) and
requested to have the “aggregation” hold phase removed, which brought
times down into the 200-250 millisecond range. The broker explained that this
was now a “direct connection to liquidity providers”… It certainly was faster.

These long aggregation times are for the broker’s benefit, of course, and
disadvantage scalpers…

But then I found much faster brokers, and have not yet pushed to request
that from them, as they are acceptably fast out of the starting gate…

hyperscalper

5372

@whywescalp

Entry is time sensitive, however I’m definitely not news trading or entering during fast moving market periods. In the past, I only used the method of entry you’re recommending for orders that could not be filled within the top 2 levels of an order book… do you really think it is necessary in my case?

5373

I don’t think I really know enough about your case, to be able to say anything. In Retail Forex, the Order Book is hidden. I get a feed from the Dukascopy SWFX exchange and so can estimate the profile of the Order Book. But most Forex operations have no access to Market Depth, nor to any Time and Sales from which Inventory Analysis could be done. I do both of these things, but based upon the Dukascopy feed which contains surprisingly revealing information, when analyzed intensively by fast algorithms… Into the weeds, and off topic, especially since U.S. persons cannot have Dukascopy accounts… I have to work through associates for that piece of the puzzle…

hyperscalper

5374

@hyperscalper

C-Trader on Trader’s Way offered level II market depth with time of sales, but US trader’s were blocked from using that particular platform last year. 2-5M was typical within the top 2-3 levels on each side during the UK/US session. This was very similar to the ECN order book’s of FXCM active trader, MBTrading navigator & Interactive Brokers TSW.

5375

Well, I don’t use CTrader, but you may know MT5 is offering “market depth”. In both of
these cases, I suspect, those Market Depths are either 1) bogus, or 2) useless. They are
almost certainly NOT representing true liquidity… But setting that aside, let’s talk about the
"sparseness" of Market Depth. This is a challenge I’ve beaten with special analysis
techniques. Sparseness involves “missing” size at various prices, at various times.

Now, the usefulness of the Dukascopy feed is that it ALWAYS gives 10 Levels of Depth
on either side of the market. That doesn’t sound like much but, it is ALWAYS 10 levels.
What I’ve seen from MT5 “anemic” displays of depth (I haven’t looked at CTrader)
is that often you’ll see only a few sizes, due to the fundamental sparseness of
these things. So the 10 Levels from Dukas on both sides of the market, are MUCH
more useful then that would suggest initially. 10 levels above the market could
represent ONE PIP. Or, due to Sparseness, it would “span” several PIPs in total
Price “distance”.

Personally I’d love to get a useful Depth; but, for me, over the years, the Dukascopy
depth is the best “free” Depth feed on which I’ve build algorithms for Trend, as well
as Trade/Order Flow through Time and Sales virtualization…

We know Forex is a “decentralized” market, but what is also true is that they all
tend to “synch” together, which enhances the value of any specific ECN’s Depth
information, provided it has some quality to it… It’s an odd thing; how much information
can be gleaned from only 10 Levels, especially if they “spike outwards” to cover
several PIPs of actual price distance from the inside market… TMI, sure, but just
to get an idea of how I approach Depth, etc… which is THE most useful short
term trend indicator possible… Valuable if Forex traders can get a handle on it !!

[EDIT] to simplify the basic Rule, it is that “Price moves to Size” or, rather, Market
Makers or Liquidity Providers move size against the Inside Market when they
wish to “transact” with the Retail population more strongly; and they simply
"pull" the inside Bid or Offer and “magically” the Price bumps right into their
advertized size, and the transaction takes place…

hyperscalper

5376

Ok, went into a panic this morning when I lost access to my LQDFX account. Turns out they made a no-notice change to their liquidity provider. Traders need to log into the backoffice and click the link to update their password then download MT4 again.

I went through those steps and was able to access MT4 again. FX offering looks pretty much the same, but spot/CFD’s are totally different now. Depending on what you trade you may or may not like the changes.

One cool thing was LQDFX now offers gold and silver trading across eight major currencies – USD, JPY, EUR, GBP, CHF, AUD, NZD, SGD. That’s the most I’ve seen anywhere except Oanda, and of course they no longer offer spot metals to US residents.

5377

At Federal Court Orders Estonian Forex Dealer to Pay over $10 Million in Sanctions for Defrauding U.S. Customers and Orders Introducing Broker that Solicited for Dealer to Pay $85,000 Penalty | U.S. COMMODITY FUTURES TRADING COMMISSION , the US CFTC says any US resident who has ever lost money while trading with Tallinex is now eligible for a restitution of the loss: “The Default Order requires Tallinex to pay $10,289,391 in restitution to U.S. customers”. You can contact the CFTC for details at [email protected]

5378

Yup. — We’ve been watching that situation develop — HERE and HERE.

5379

So I’ve switched to TurnkeyFX. Support has been great, commissions seemed much better then other brokers, and they promised really REALLY low spreads. Have been live trading for over a month now and have found some issues.

  • SPREADS: Yes, they typically have low spreads (less then a pip), but most news events or when liquidity eases off (beginning of Asian session), those spreads go through the roof (+10 pips). 8h30 (8:30am) on a Friday during a CAD news event the spread increased to +35 pips. Granted, I haven’t monitored spread a lot at my other brokers, but that shocked me!!

  • SWAPS: So another trader was kind enough to walk through the spreads today and compare them to a few other brokers he uses. End result is that Turnkey really (like REALLY!!) sucks on most swaps. Unless you’re in-and-out during the day, those swaps really eat into your profits quickly. It’s something I hadn’t paid a lot of attention to in the past, but now that I hold some trades for 2 days - 2 weeks, it’s starting to make a huge difference. Turnkey better true up it’s swaps soon or just go out of business.

So that’s my thoughts at this point. Hopefully we can negotiate some better swaps with them, but I’m going to begin looking at some of the other brokers in the above list. Right now it sucks to trade FX as a US citizen (EU is starting to feel our pain with their recent regulatory changes).

Cheers,
FW

5380

Yes, there are specific times when TurnkeyForex or FinProTrading have terrible spreads, mainly Rollover time or certain News Events where there is lower liquidity or higher uncertainty. I would describe their Tick feed as “highly reactive” and feeding close to 5x-10x as many “deltas” or ticks as the 3rd tier brokerages… However…

Execution is lightning fast, and accurate; and I’ve previously posted the Order Operations round trip latencies which are very stable, and on the fast end of the sprectrum for these kinds of “cheap” brokers… (Unfortunately, we U.S. persons do not have the full range of choices of the world’s best, but these two brokers are among the best we have…)

SWAPS? I don’t know about that topic.

Here’s a screenshot of their spreads NOW, just 5 minutes away from Rollover…

FinProTradingSpreads-just-prior-to-rollover.PNG313×520 3.69 KB

[EDIT] by comparison, here we are NOW, 21:15 GMT immediately after rollover.

AfterRollover.PNG355×535 4.63 KB

[EDIT2] If you use close brokerage STOPs these Spread anomalies can cause
triggering of STOPs… I recommend not using broker stops; or keeping them far
away as a “safety net” only, partially because of this effect…

hyperscalper

1 Like
5381

SWAPS are typically a function of interest rate differential. Based on my analysis there are two types of brokers. One that focus on commissions and let the swaps be calculated theoretically and another types who focus on swaps and cover that up in their commissions. Technically, I prefer the first type but it also depend on your strategy. If you are a low frequency trader, you should focus more on swaps and not as much on commissions. While if you trade more than a couple of times a day, you should focus more on commissions (this , mathematically is irespective of the size of the trade but the average size of the trade)

Here is how theoretically SWAPS should be calculated in a “no-arbitrage” pricing model.

(Interest Rate of Numerator - Interest Rate of Denominator ) / 365 (or 252 based on frequency of swap posting)

So for example, if you are long BRLUSD and assuming the interst rate on BRL is 13% and that on USD is 1% then your swap charges is 12% per annum. So for one lot of BRLUSD you will pay 12000 USD per year or $32 per standard lot per day. And if you are short BRLUSD, you will earn $32 per standard lot per day.

Now to make it precise, you will have to add interest rate factor of your account into prespective. This is a hidden area and every broker has his own interest factor. But this in theory is to cover the 400X leverage that you are utilising and the cost of providing you with that leverage. Lets say if the interest cost if 1% and you are using 400X leverage, then this cost is 4% per year for the broker and your swap rates for Long BRLUSD will be 16% (13% - 1% + 4%) and Short BRLUSD will be -8% (13% - 1% - 4%) which means that you will pay $43 per standard lot per day and if you are short BRLUSD you will get $22 per standard lot per day.

Also note, there are only 2 variable factors in this One) What is internal leverage cost of the broker and Two) is the broker baking in swaps in the commission or not (lower commission versus higher commissions)

Now these examples are in no way reflection of Finpro or turnkeyforex or turnkeyfx but a general an example that is used by each and every broker in the industry. I know it so precisely because I used to do swap arbitrage and this is the basic calcuation that needs to go into the model to find arbitrage opportunities. And if you have an interest in SWAP arbitrage, you should try and find islamic accounts and see how you can play around that (sorry, no more details on a public forum about it :slight_smile: )

Hope that helps

5382

If you are looking for an offshore broker then I can highly recommend Hugosway.

1 Like
5383

Yeah it’s those times when spreads just go wild that get me nervous. Yes typically they’re tight, but there’s quite a few situations where they’ve widened to +5 pips.