Order Flow Trading Reborn: Live Data Via OrderFlowFX

As many of you know it is not too common that I get excited about certain things, especially when it comes to any kind of product that gets released for this industry. The first question I ask myself when it comes to talking with other people in relation to the site, is “Would I want to use it?” If the answer is no, then it’s obviously a no-go. As I already don’t need to tell you, there are heaps of junk that hits the market and finding something that provides any kind of clarity is a rare exception – this is one of the reasons I heavily revised the resource section.

So here is the deal: NBT as a web entity is basically a financial liability, and so I hit the pavement and went on a hunt to find some new partners. And it is very, very rare that we do this but after some Google hunting I found one that was in their infancy stages, and the more we talked, the more we clicked. They are also based right around the corner from me, so major points there.

I found out about OrderFlowFX a few weeks ago. As I rarely have issues expressing my opinion, the FX market is a basically a den of thieves. I have had trouble in the past dealing with the conduct of some of the people I am forced to work with sometimes. This is also probably why I like what these guys are doing. They offer transparency in a market that is blurred out by even the guys that seem the most ethical.

Essentially they offer order flow in two different forms. These 2 order flow feeds spread out across a range of indicators: a retail and an aggregate institutional. The retail feed is sourced from the books of several retail brokers. The institutional is sourced from a liquidity pool used by a range of market makers. So here we have a couple different ways to look at depth for the over the counter, interbank FX market.

No retail broker offers live order flows that you can store in a history. Some have depth on ECN platforms (such as white labeled Currenex platforms), but they only give them to clients with large amounts of money and the feed itself can be mostly deemed utterly useless because you only have a small, live snippet and nothing more. Volume profile is available elsewhere but buy and sell intent is segregated. I have one of these fees on my current platform and I never even use it.

Brokers don’t want their clients having the info for a few reasons relating to one thing: its bad for business. First, if they are a market maker, they don’t want their clients doing what they do. Second, I can only guess that no broker wants their clients to be constantly reminded about how horrible they actually trade. It’s a dark cloud, understandably, and also bad for business.
Here is a little more about these types of models, and feed that gets displayed:

The Market Making Model

The “market making” model that a huge chunk of retail brokers exercise (prime brokers of course do this as well though in virtually any market) earns them money in a couple different ways.

1. First, they are able to capture a marked up spread. They claim this is better for the client because they are able to better control spread fluctuations, etc. This is basically garbage. I have traded on tons of institutional ECN platforms now and prefer using them over any fixed, market making model any day of the week. The costs are dirt cheap compared to retail brokers and I can see spread fluctuations as a function of volatility, not to mention you can trade inside the spread, etc. Lots of advantages on these platforms.

2. They trade against client positions. This is where they roll in the extra bank. Risk controls are obviously still used and programs employed to detect predatory trading or anything else that could be detrimental to their P&L.

And so for a number of different reasons, OTC (over the counter, interbank) Forex volume is obviously very different than anything that gets passed through an exchange (such as FX futures). You’re talking apples and oranges, and in the case of OrderFlowFX it’s buy and sell intent. Everyone on the futures market knows that the participants of this centralized pool has access to their data and so trading reflects it. This isn’t even to mention the nature of the product, and its obvious and strict correlation to interbank rates. Tape traders that know what they’re doing can make good sense of this, and it is a trained skill like any other, such as support and resistance trading.

The Retail Feed

The retail feed what what they started with and displays live buying and selling of retail clients in a number of different ways. They have indicators for cumulative or net positions total, segregate by open, etc. What is alarming is the overwhelming number of losses accrued by the retail crowd. Sometimes you just need to see it to believe it. Either way, a number of different strategies could be used to trade against these. I am going to be working with the founder, Tim, over the next few weeks and help flesh out the remainder of what him and his friends have put together. He plans on posting the results over his new blog, which just started yesterday.

The best results we are finding so far come in times of high volatility. Retail traders tend to stick with the mantra “buy low sell high” a little too seriously, and end up accruing and insane amount of drawdown in the process. These positions are usually left open in hopes that they will return, either to feel the pinch of a margin call or close in agony.

We are also finding that different results come pair-specific. Certain pairs exhibit different types of trading behavior, and drag in a different crowd that trades them.

The Institutional Feed

The value of this feed is very different from the retail feed in that you can’t simply buy or sell against the positions due to the nature of market participants using it. Currently, OrderFlowFX has one indicator, and working on several others, that uses this feed properly. Again, you can’t simply look at buys and sells and say “these guys are wrong” (in any circumstance). The current indicator in production creates a band around price that demonstrates extreme bid and ask outcomes during times of low volatility. These levels are used as support and resistance in the future.

Clumped orders that exhibit large deviations tend to move price in a certain direction. These can precede the movement itself. This is what is being worked on currently, so they have apparently been spending some time in the late night working on ways to optimize this.

Optimizing Old and New Strategies

There are a number of different strategies that can be applied using longer-term order flows. As with anything else, here it is important to note that a strategy is a strategy no matter what it uses as a primary input. Understanding/optimization and good risk reward is required. Volume is a tool in any case, but I know we have a lot of people out there curious about more depth for this decentralized market.

I have never seen these types of flows offered before, so to me at least, this is new territory. But the concept should be relatively simple: take advantage of scenarios in which a greater majority of the trading public struggles, as twisted as it might sound. Retail traders can and will be correct at certain times, so you need to isolate those times in which they are not.

Tim has already done some of this work for us, and explains it in a few words on the site:

• Buying or selling into a non-event spike, attempting to latch onto (what he/she views) as a trend, when in fact price is exhausting and reversing. These are typically larger price bars not found within the context of a news release.

• Trading against a large, event-based spike that does not reverse. Traders mistakenly identify a news event as an opportunity to buy low or sell high. Price continues, accruing a loss.

• Trading into a large heavy area of support or resistance with the expectation of a breakout or fade. Price travels the opposite path.

• Leaving drawdown exposure open and conversely, closing profitable exposure rapidly.

So in a number of different obvious ways explained here, there is opportunity. In many ways this feed is well suited for systematic traders, but the discretionary value is there as well. And without wasting too much time here digging into strategies I found one very common pattern: high volume led to the best returns.

And so that’s it for today. Just something of interest in which I ended up getting invested for the site. Tim and his coworkers are working on updates at virtually all times, and just released one for Metatrader’s new build, with work on other platforms. Open code too, so you can fiddle away.

And clicks through here end up supporting this site, but please use at your own discretion and understand what it is. For most of you I don’t need to explain this but for others, the disclaimer is there just in case. Onward.