Order flow basics for Croatian futures traders on NT8

Open the SuperDOM on the MNQ at half past three in the afternoon, Croatian time, and the first impression is that something is broken. Numbers on both sides of the price rewrite themselves several times a second. A block of 400 contracts appears at a level, sits there for two seconds, and vanishes without a single trade going through it. The last traded size column scrolls faster than you can read it.

If you learned to look at markets on the Zagreb exchange, or on a CFD platform where the "depth" was a picture your provider drew for you, none of this has any precedent in your experience. There is nothing to compare it to. And almost everything written about order flow in Croatian is either a translated marketing page or a course sales pitch, which is why the same handful of questions circulate in Croatian trading groups year after year without ever getting a straight answer.

This is the straight answer. What the data actually is, which NinjaTrader 8 windows show which part of it, what you need to subscribe to before any of it works, the four things worth learning first, and the two things everybody believes about depth that are wrong.

Why Croatian traders start from zero here

It is not a knowledge gap, it is a structural one. The domestic market simply does not produce the raw material that order flow reading is built from.

The Zagreb Stock Exchange is a small market by design and by size of economy. On a typical name outside the handful of blue chips, the book has a few orders on each side, the spread is wide enough to matter, and whole minutes pass with no prints at all. You cannot learn to read aggression against resting liquidity in a market where there is almost no resting liquidity and very little aggression. That is not a criticism of the exchange, it is arithmetic.

Then the second layer. A great many Croatian retail traders got their first screen time on CFD platforms, where what is labelled market depth is generated by the provider and reflects the provider's own quoting, not a central limit order book with real participants queued in it. Habits formed there transfer badly, because they were formed against a simulation.

So when you connect NinjaTrader 8 to a CME feed, you are seeing a genuine central limit order book for the first time. The E mini and Micro E mini S&P and Nasdaq contracts are among the most heavily traded futures in the world, and their books are deep, fast and continuously contested. Everything in this article assumes that specific environment.

One practical note on your side of things before we get into the data. Croatia has used the euro since 1 January 2023, so your bank account, your salary and your costs are in euro, while the CME contracts and the market data subscriptions are priced in dollars. The Croatian National Bank publishes euro reference rates and the national statistics that go with them on its website, where the August 2026 annual HICP inflation reading for Croatia was 3.7 per cent. Keep that dual currency reality in the back of your mind: your fixed monthly costs for doing this are dollar costs.

Order flow is three data objects, not one

The word "order flow" gets used as if it were a single indicator. It is not. It is three separate streams, and confusing them is the source of most bad analysis.

One: time and sales, also called the tape. Every executed trade, with price, size, timestamp and, crucially, which side was the aggressor. When a trade prints at the offer, a buyer crossed the spread and took liquidity. When it prints at the bid, a seller did. This is fact. It already happened, it cannot be cancelled, and it is the only part of the three that is beyond dispute.

Two: market depth, the book. The resting limit orders at each price above and below the market, aggregated by size. This is intention, not fact. Every one of those orders can be cancelled in a microsecond and a large proportion of them are. Depth tells you what is currently offered, not what will be honoured.

Three: derived measures. Everything computed from the first two. Delta is buy aggression minus sell aggression over a period. Cumulative delta is the running total. A footprint chart is the bid and ask volume shown inside each bar at each price. Volume profile is traded volume distributed by price rather than by time. None of these are extra information, they are rearrangements of the tape, and that is a feature: the rearrangement is what makes patterns visible.

Hold on to the distinction between the first two. Fact versus intention. Nearly every rookie mistake in this discipline comes from treating the book as if it were the tape.

Which NinjaTrader 8 window shows what

NT8 splits the three streams across separate tools, and knowing which is which saves a lot of confusion.

• SuperDOM. The ladder. Price down the middle, resting bid size on one side, ask size on the other, your working orders shown at their level, with a volume column. This is your depth window and your order entry at the same time. Learn the order handling here first, because clicking the wrong cell in a fast market is how beginners get a position they did not intend.
• Level II window. The same depth data as a flat table. Useful on a second monitor when you want the numbers without the ladder geometry.
• Time and Sales window. The raw tape, colour coded by aggressor side. Set a size filter so that only trades above a chosen size show, otherwise on the ES it is unreadable noise.
• Order Flow charts and tools. Depending on your NinjaTrader licence tier, you get the footprint style bid and ask chart types, volume profile, cumulative delta, a depth map that renders the book as a heatmap over time, and a VWAP with standard deviation bands. The free tier gives you SuperDOM, Level II and Time and Sales; the footprint and profile tooling sits in the paid tiers. Check what your licence actually includes before you build a method around a window you cannot open.
• Market Replay. Downloaded historical days you can play back tick by tick at any speed. This is the single most valuable feature on the list for someone learning, and it is the one people skip.

What you have to pay for before any of it works

Here is where people lose a week. You can install NinjaTrader 8, connect to a simulated feed and see charts without paying anybody. You cannot see real market depth that way, and every order flow technique in this article needs real depth.

Three separate things have to be in place:

• A connection that carries depth. The routing and data layer behind your broker, Continuum, Rithmic or CQG, has to be provisioned for market depth on your account. Ask the broker explicitly. "Do I have CME depth enabled" is a one line support ticket.
• A CME market data subscription. Exchange data is licensed by CME Group, billed monthly in dollars, and passed through by your broker. There is a non professional rate for individuals trading their own money, and a much higher professional rate. The current amounts are published by CME Group and change from time to time, so take the number from the exchange or your broker's fee schedule rather than from a forum.
• Tick Replay or Market Replay data. To study past sessions properly, NT8 needs the historical tick data downloaded and, for certain calculations on historical bars, Tick Replay enabled on the chart. Without it, indicators that need every tick will show you numbers that never existed in real time.

Budget for this honestly. Depth data, a decent data connection and possibly a VPS are a fixed monthly cost in dollars that arrives whether you trade or not. For someone starting out in Croatia, that cost is often the thing that decides whether this is a serious project or an expensive hobby.

The four things worth learning first

Ignore the encyclopaedia of patterns. These four carry most of the value and each one can be verified in replay.

Absorption. Aggressive market orders keep hitting one side, the tape shows steady volume at the bid, and price does not go down. Somebody is absorbing that selling with resting buy orders and refilling as they get hit. The tell is the mismatch: heavy one sided volume, no price movement. Absorption is interesting because it is the point where the aggressive side is spending money and getting nothing, which is often where a move ends.

Imbalance. On a footprint chart you compare the volume traded at the ask of one price against the volume traded at the bid of the price below it, diagonally. When one side is several times larger, that is an imbalance. Two or three stacked imbalances in the same direction at consecutive prices mark a level where one side pushed with real force, and those levels often matter when price returns to them.

Iceberg and refill. A price level shows 50 contracts on the bid. Two hundred trade there and it still shows 50. That is a hidden order refreshing its visible portion, and it is the clearest possible signal that a real participant with size wants that price defended. Unlike a big displayed order, an iceberg cannot be faked cheaply, because it only reveals itself by actually filling.

Exhaustion. Price pushes to a new session extreme, the tape goes heavy in the direction of the push, and delta keeps rising while price stops making progress. The last buyers are buying at the worst price of the day. Combine this with the absorption reading and you have most of the useful content of a reversal signal.

Learn them in this order, one at a time, in replay, on a single instrument. The MES or MNQ are the right choices to start: liquid enough to have a proper book, small enough per tick that a mistake during the learning phase costs a coffee rather than a week's pay.

Two things everybody gets wrong about depth

The first: a big number in the book is not support. A 900 lot bid five ticks below the market feels like a wall. It is a limit order, and limit orders are cancellable at no cost. It can be pulled in the time it takes price to travel one tick. Resting size that disappears the moment it is approached was never liquidity, it was display. The only size you can trust is size that has actually traded, which brings you back to the tape. Watch how the level behaves when price reaches it, not how it looks while price is far away.

The second: the book you see is not the whole book. Your feed shows a set number of aggregated price levels, typically ten on each side for CME equity index futures. Beyond that there is more, and you cannot see it. Iceberg orders hide their true size by design. Some participants never rest an order at all and only ever take. Treat the ladder as a partial and deliberately incomplete picture, because that is exactly what it is.

Add a third, smaller one while we are here: latency changes what you see. From Croatia, your packets travel to the matching engine in Chicago and back. That round trip is a physical constraint and no amount of money fixes it entirely. It matters very little for a discretionary trader reading absorption over ten seconds, and a great deal for anything trying to react inside one second. Know which of the two you are.

The Croatian clock, and when to actually watch

Croatia is on CET in winter and CEST in summer, the same as Germany and Austria. For most of the year that puts the New York cash open, 9:30 Eastern, at 15:30 Croatian time, and the cash close at 22:00.

The exception is worth writing on a sticky note. The United States switches to daylight time on the second Sunday in March and back on the first Sunday in November; the European Union switches on the last Sunday in March and the last Sunday in October. In the gaps, roughly two weeks in March and one week at the turn of October into November, everything shifts an hour and the cash open lands at 14:30 local. Two or three times every year somebody misses the open because of this.

For order flow specifically, the hours that matter to you are the first ninety minutes after 15:30 and the last hour before the close. That is when participation is highest, the book is deepest and the patterns above are cleanest. The European morning session on the ES has thin depth, wide effective spreads and a much higher proportion of noise, which makes it the worst possible environment to learn in. If your job means you can only sit down at nine in the morning, use that time for replay of yesterday's US session rather than for live trading.

A thirty day plan that actually works

Structure beats enthusiasm. This is a realistic sequence for someone in Croatia starting from nothing.

Week one: mechanics only. No analysis. Download five recent sessions of Market Replay data. Practise order entry and modification in the SuperDOM until placing, moving and cancelling a bracket is automatic. Learn what happens to your stop when you flip a position. This is the boring week and skipping it is how people lose money to the platform rather than to the market.

Week two: the tape alone. Time and Sales with a size filter, SuperDOM beside it, nothing else. Watch one hour of replay a day and write down, in a notebook, every time you see aggressive volume fail to move price. You are training your eye to detect absorption and nothing else.

Week three: add the footprint. Same hour, now with a footprint chart. Mark the stacked imbalances. Then check what price did when it came back to those levels later in the session. You are building a personal evidence base instead of accepting someone else's claims.

Week four: one instrument, live, minimum size. One contract on the MES. Two hours a day, 15:30 to 17:30 Croatian time. The goal is not profit, it is to find out which of the things you saw in replay you can actually act on when your own money is in the trade. Keep the journal going, because the gap between replay you and live you is the real subject of this month.

After that you will know whether order flow suits how your head works. Some very good traders read the tape well and some equally good ones find it a distraction and do better with structure and levels. Both answers are legitimate, and finding out costs you a month and a data subscription rather than an account.

Who supervises any of this in Croatia

Short answer: nobody supervises your CME trading, and it is worth being clear about that.

HANFA, the Croatian Financial Services Supervisory Agency, supervises and regulates the Croatian non bank financial market, including investment firms, funds, pension and insurance, and it publishes consumer education material. It does not supervise the Chicago Mercantile Exchange, and it does not supervise a US futures broker that has accepted you as a client under its own rules. If you deal with a firm outside the EU, your recourse is that firm's home regulator, not HANFA.

Where HANFA is genuinely useful to you is the other direction: if a firm approaches you claiming to be authorised in Croatia or in the EU, checking that claim against the supervisory registers is quick and it is the cheapest fraud protection available.

On the tax side, profit from trading is taxable income for Croatian residents and the rules on which category applies, how losses are treated and what has to be reported are set out by the Tax Administration. The current rates, thresholds and forms are published by Porezna uprava, and since the local income tax surcharge was reformed the applicable percentages depend on your municipality, so take the number from the source rather than from an older article. Keep every statement, in dollars, with the conversion you used and the date, from your first trade.

This article is general information about market structure and platform tooling, not tax, legal or investment advice. Tax rules and market data terms change, and your own circumstances may differ, so check the current position with Porezna uprava, HANFA or a qualified Croatian adviser before you act on anything here. Leveraged futures can lose you more than the amount you deposit.

order flow croatian traders

Educational content, not investment advice. Trading futures involves substantial risk of loss and is not suitable for every investor. Any reference to Algomyth results is hypothetical, simulated over historical data, and shown in full on the performance page; past performance does not guarantee future results.