EXPOSING How Banks Hide their Trades

ELI5/TLDR

When a big institution wants to buy a huge amount of something, it can’t just place one giant order — that would scare everyone and move the price against it. So it slices the order into tiny pieces and feeds them in one sliver at a time, hiding the bulk like the underwater part of an iceberg. Most trading screens can’t see this. But with a particular kind of expensive data feed, you can tell that a string of small orders all share the same ID, which means they’re really one big hidden order — and that lets you guess where the big money is quietly piling in.

The Full Story

Liquidity is the thing under the price

The whole video hangs on one claim: the most important thing to understand as a trader is liquidity — who is willing to buy and sell, at what price, and how much. The host argues that the popular retail idea of liquidity (stop-loss clusters above highs and below lows that price “hunts”) is only a sliver of the real picture.

That’s just like seeing not even half the picture. This is like, uh, let’s say 10% of the picture.

The fuller picture splits into two halves. Passive liquidity is the resting orders sitting in the book — people offering to buy or sell, but only at their chosen price, waiting. Aggressive liquidity is the side that crosses the spread to trade right now. In the trade’s own vocabulary: market makers post (passive), market takers hit (aggressive).

The order book, in plain terms

An order book is just a ledger of intentions. Above the current price sit the asks — sellers offering, e.g., 33 contracts at 101, 42 at 102, and so on. Below sit the bids — buyers offering, e.g., 41 at 99, 33 at 98. The best ask is the lowest price a seller will accept; the best bid is the highest a buyer will pay.

If you want to buy, you have two moves. Post a buy limit and wait (passive — you control the price but not whether you get filled). Or accept a seller’s offer right now (aggressive — you get filled instantly but pay the spread). The host’s recurring metaphor is a butcher’s counter: do you offer meat, or take the meat.

When a big aggressive buyer arrives and there isn’t enough at one price, the order eats through several levels — some at 101, the rest at 102 and 103. That climbing-the-ladder effect is slippage: the order is too big to fill in one place.

Why size is a problem only the giants have

A retail trader buying five contracts never notices any of this — the order vanishes into the market instantly. But imagine BlackRock needing five million contracts. Two bad options: place one enormous visible order (everyone front-runs you, assuming you know something), or sweep the market aggressively (you slip the price up violently and pay through the nose).

So the quants invented a third way. Chop the giant “meta order” into millions of small ones and pour them in slowly, mixing passive and aggressive fills to minimize cost. This is where the named execution algorithms come from — VWAP, TWAP, percentage-of-volume, implementation shortfall — all variations on optimizing how a big order is dribbled into the market.

Typically bullish order flow will lead to more bullish order flow.

That’s the practical hook: because a visible burst of buying is usually just a slice of a much larger hidden program, order flow has a measure of predictability. The big buyer still has thousands of contracts left to fill.

The heat map and the hidden part

A liquidity heat map colors the order book by how much size is resting at each price — hot where there’s a lot, cold where there’s little. Good software also shows how that liquidity moves and gets eaten over time. Price tends to get pulled toward the hot zones, because the big players need that liquidity to fill their orders. The host shows price rocketing up to consume a 465-contract resting order.

But the genuinely interesting part is what the heat map can’t show. An institution that doesn’t want to advertise will display only, say, 8 contracts. Each time a counterparty eats those 8, a fresh 8 instantly reappears, then another, and another — until a huge order is quietly filled with no visible wall ever forming. That’s the iceberg order: a tiny tip on the surface, the mass hidden beneath.

The data-feed plumbing (the actual mechanism)

This is the technical core, and it’s where the title’s “exposing” lives. Whether you can see icebergs depends entirely on how granular your data feed is.

Level 1 (top of book): only the best bid and best ask, plus executed trades. Cheap, sometimes free. Enough for footprint charts, volume profile, and delta — but you cannot build a real heat map from it.
Level 2 / full market depth (MBP — Market By Price): every price level in the book, not just the top. Now you can see the whole resting ladder. But MBP only tells you the total size at a price — 78 contracts could be one institution or 78 retail traders each with one. You can’t tell.
MBO — Market By Order (some call it Level 3): shows each individual order with its own unique order ID, on both the passive and aggressive side. This is the most granular feed money can buy.

And here is the punchline. An iceberg’s many small refills all carry the same order ID. So with an MBO feed, you can see that a run of 8-contract orders is really one player — one hidden whale — and infer where the institution is accumulating.

Thanks to MBO data feeds, we can see that these small orders of eight contracts have the same order ID, so they’re likely part of one big order.

The demo and the sales pitch

The rest is a screen-share of the host’s own product, Deep Dome (an add-on to his Deep Charts platform, pitched at ~$300 lifetime, with a bundled “liquidity bootcamp”). He flips a data source from MBP to MBO, turns on an “iceberg tracker,” and walks through replays: a 125-contract iceberg seller capping a session high; a 160-contract iceberg marking a session low; absorption walls in the London session and pre-market that “plot” turning points. He even runs a paper-trade, scaling into longs alongside a 1,000-contract buy iceberg near a low-volume node.

He’s careful, to his credit, to say it doesn’t work every time, and that you must add context — volume-profile value areas, opening-range breakouts, key highs — and backtest it rather than trade icebergs blindly.

Of course, they’re not going to plot the high every single time. I wish it was like that, but if it was like that, we would all be rich.

Key Takeaways

Iceberg order: a large order that displays only a small visible quantity; each time the visible slice is filled, it silently refills from a hidden reserve, concealing the true size.
Two kinds of liquidity: passive (resting limit orders, posted by market makers — price control, no fill guarantee) and aggressive (orders that cross the spread to execute now — fill guarantee, but pay the spread).
Stop-hunt “liquidity” is only part of it. The retail framing of liquidity as stop clusters above/below swing points is a small slice; the resting bid/ask book is the larger reality.
Slippage happens when an order is too big to fill at one price and eats through successive book levels, getting progressively worse fills.
Meta orders: institutions split one giant order into millions of small child orders and feed them in gradually to avoid moving price and avoid front-running.
Execution algorithms — VWAP, TWAP, percentage-of-volume, implementation shortfall — exist to optimize how a meta order is dribbled into the market (balancing spread cost vs. market impact).
Order-flow persistence: a visible burst of institutional buying is usually one slice of a much bigger program, so bullish flow tends to beget more bullish flow — a source of short-term predictability.
Liquidity heat map: colors resting size by price level; price tends to be drawn toward hot (high-liquidity) zones because big players need that liquidity to fill.
Data-feed hierarchy: Level 1 = top of book only; Level 2 / MBP = full depth but only total size per price; MBO (“Level 3”) = every order with a unique order ID.
The detection mechanism: iceberg refills all share the same order ID, so only an MBO (Market-By-Order) feed can unmask an iceberg — MBP cannot, because it aggregates.
Feed providers differ: CQG typically delivers Level 2 as MBP; Rithmic and DXFeed deliver MBO.
Absorption: a hidden iceberg soaking up all incoming pressure builds a wall buyers/sellers can’t pierce, often marking session highs and lows; reportedly strongest in London and pre-market sessions.

Claude’s Take

The title is clickbait, but the underlying microstructure is real and mostly described accurately. Iceberg (or “reserve”/“hidden”) orders genuinely exist on most major exchanges and are a standard, openly-documented order type — there’s nothing illicit about banks “hiding trades”; it’s just sensible execution to avoid moving the market against yourself. The book-mechanics, slippage, passive/aggressive split, meta orders, and the VWAP/TWAP/implementation-shortfall family are all textbook-correct. The Level 1 / MBP / MBO distinction is the most useful and accurate part of the video, and it’s the actual answer to “how do you spot a hidden order”: you need order-level data with persistent IDs, because aggregate price-level data mathematically cannot distinguish one whale from a crowd.

Where to keep your guard up: this is, structurally, a 12-minute infomercial for the host’s own software. The detection isn’t as clean as the replays suggest. Many venues randomize iceberg refill behavior and don’t always keep a stable order ID precisely to defeat this kind of reverse-engineering, so MBO unmasking is probabilistic, not certain — the host quietly says “likely part of one big order,” which is the honest hedge. The cherry-picked replays where an iceberg “plots the high” are survivorship theater; he admits it doesn’t work every time, but every example shown happens to work. And spotting a large absorbing order tells you size is present, not direction or outcome — absorption can precede a reversal or get steamrolled.

Net: a clear, genuinely educational explanation of order-book microstructure and data feeds, wrapped around a product pitch with the usual edge-implied-but-not-proven framing. Worth it for the MBP-vs-MBO concept alone, which is the part most retail explanations skip. Score 6 — solid pedagogy, real mechanics, but discount the implied tradability and remember who’s selling the shovel.