What is market depth?

Open the depth window on any actively traded instrument and you see a ladder split down the middle. On one side sit the bids, resting buy orders, each saying "I will buy this quantity at this price." On the other sit the asks, resting sell orders saying the same in reverse. Nothing on the ladder has traded yet: every row is a statement of intent, waiting in a queue. Most Indian platforms lay the two sides out together, bids on the left and asks on the right, and for each price they show three things, the price itself, the total quantity resting there, and the number of orders that make up that quantity. A level showing a large quantity spread across many small orders behaves very differently from the same quantity sitting in a single block, which is why the order count is worth a glance.

The two rows that matter most are the ones closest to the middle. The best bid is the highest price a buyer is currently willing to pay; the best ask is the lowest price a seller is currently willing to accept. The gap between them is the bid-ask spread, and it is the immediate cost of crossing from one side to the other for a small trade. Everything below the best bid and above the best ask is depth: the reserves of size that get consumed only if a bigger order arrives. Read the ladder from the middle out and its shape, tight or wide, heavy or thin near the touch, is the clearest picture you can get of how easily the instrument can be traded, which is exactly what a broad guide to liquidity in trading is about at the level of a single name.

The window you see is deliberately shallow. The standard NSE feed shows the best five price levels on each side, a view called market-by-price at five levels, or MBP-5, and it arrives as a snapshot rather than a stream of every individual order. That is the depth ladder most retail terminals display. A deeper twenty level view exists and is redistributed by a few brokers at a premium, and above that sits a full order-by-order feed that streams every order added, modified and cancelled, a colocation-grade product carried on dedicated lines and built for latency-sensitive firms rather than a normal retail terminal (all as of 17 July 2026; the exchange revises its data products from time to time, so verify the current tiers at source). For almost every retail decision the five level window is enough, provided you remember it is the shallow end of a much larger book.

One detail on the window repays a closer look: the order count beside each quantity. A level showing 6,000 shares across twenty-nine orders is a crowd of small participants, and it tends to be stickier, because no single trader can pull all of it at once. The same 6,000 shares held in two orders is a pair of large players who can both step away in an instant, so that level is more fragile than its size alone suggests. Behind every aggregated number sits a queue ordered by price first and then by time, so that an order resting earlier at a given price is filled before a later one that joined the same level. That price-time priority is the plumbing beneath the ladder, and the finer machinery of the aggressor, the resting side and the printed tape belongs to the guide on order flow a level deeper; here it is enough to know that each row is a queue of many intentions, not a single solid lump.

Here is the mechanism that turns the ladder from decoration into a tool. A market order does not fill at one price. It fills against the resting sizes on the book, best price first, and if it is larger than the quantity waiting there it consumes that level and walks to the next, worse, price, and the next, until it is complete. Your average fill is the size-weighted blend of every level it touched, and the gap between that blended price and the fair price you started from is the cost your size imposed on itself. Traders call that gap the impact cost, and it is the single most useful number the depth window can give you, because it is the part of the cost of trading that the headline spread never shows.

Take the liquid book above and send a market buy for 12,000 shares (illustrative). The best ask is 500.10 with 4,000 resting, so 4,000 fill there. That level is now exhausted, so the next 5,600 walk up to 500.20 and fill. You still need 2,400 more, which come from the 6,400 resting at 500.30. Your order is complete, having touched three price levels. The volume-weighted average fill is (4,000 at 500.10 plus 5,600 at 500.20 plus 2,400 at 500.30) divided by 12,000, which works out to 500.19 a share. Measured against the arrival mid-price of 500.05, the fair price halfway between the best bid and best ask when you clicked, your impact is about 0.14 a share, which on 12,000 shares is roughly 1,640 rupees, or about 2.7 basis points of the traded value. In a book this deep that is a rounding error, and that is precisely the point: the ladder let you know it would be trivial before you sent the order.

It is worth being exact about the two reference prices, because they answer different questions. Measured against the best ask you saw, 500.10, the slippage on this order is about 0.09 a share: that is the pure penalty for being large enough to walk past the touch. Measured against the mid-price, 500.05, the impact is about 0.14 a share, which also folds in the half-tick it always costs to cross from the mid to the ask. The mid-price version is the one exchanges use, and it is the fairer measure of the total friction, because even a one-share order pays the half-spread. Either way, the discipline is the same: the spread tells you the cost of a token trade, and impact cost tells you the cost of your trade. For anything past a token size, the second number is the one that decides whether the trade is worth doing.

Everything so far assumed the book you read is the book you trade against. It usually is not, at least not exactly, and this is the single most important caveat in the whole subject. A resting order is an intention, not a commitment. The trader who placed it can cancel or move it in a fraction of a second, and in a modern market the overwhelming majority of orders are cancelled rather than filled. The ladder you are looking at is a photograph of a crowd that is constantly walking in and out of the frame, and by the time your order arrives the crowd has already shifted. That does not make the book useless, but it does mean you must read it as a snapshot of what is resting now, not a guarantee of what will be there when you need it.

This is why a large resting order is not the same as support. A conspicuous block of buy orders sitting just below the price looks like a floor, and it is often narrated as one. But the very visibility that makes it look like support is also why it frequently vanishes: a genuine buyer who truly wanted the stock would rarely advertise the full size and invite others to trade ahead of them, and a great deal of large, visible resting size is placed to be seen rather than to be filled. When the price actually drifts down toward that block, it is common to watch it evaporate a tick before it would have traded, because the order was never a commitment to buy in the first place. The book gives you a place the price may pause; it never gives you a wall you can lean on.

It helps to understand why the crowd shifts so fast. Most resting orders in a modern electronic market are never filled at all; they are placed, repriced and cancelled as conditions change, and automated participants update their quotes many times a second. A great deal of visible depth is therefore genuinely meant, but only for now and only at this price: it is liquidity that will follow the market up or down and quietly step aside when a large order comes hunting for it. This is what traders mean by fleeting liquidity, and it is not sinister. A quote is simply a standing offer the maker is free to withdraw, and the faster the venue, the shorter that stand tends to be. The practical consequence is that depth two or three levels away from the touch is softer than it looks, because it has the most time to move before your order could ever reach it, which is another reason the near levels deserve more of your trust than the far ones.

The impact cost of an order is not a property of the order alone. It is a property of the order and the book together, and the same size can be trivial in one instrument and ruinous in another. This is the difference between a deep book and a thin one, and it is where the depth window earns its keep, because it is the one place you can see the difference before you commit. A liquid large-cap carries thousands of shares at every near level, so an order walks a tick or two and stops. An illiquid small-cap or a thinly traded microcap carries a few hundred shares at each level and wide gaps between them, so the same order sweeps the whole visible book and keeps going.

Send the identical 12,000-share market buy into a thin book (illustrative) and the arithmetic turns brutal. Suppose the best ask is 500.30 with only 1,200 resting, the next level is 501.10 with 900, then 502.40 with 800, 504.00 with 1,100 and 506.50 with 700. Walk the order and it takes the whole visible ask side, all 4,700 shares across five levels, at a volume-weighted average of 502.60, an impact of roughly 52 basis points against a 500.00 mid, close to 12,220 rupees on the part that fills. And it is only a part: 7,300 shares, more than sixty percent of the order, cannot fill at all inside the visible book. They would either push the price into the exchange's daily price band and stall against the upper circuit, or sit unfilled while the trader stares at a half-done order. In the deep book the same 12,000 cost 2.7 basis points and completed; in the thin one it cost nearly twenty times as much per share and could not even finish.

Even within your five levels, the number you see at a price may not be the whole order. The exchange offers a legal way to hide size called a disclosed-quantity order, the exchange-native iceberg: a trader can choose to display only a fraction of a large order in the depth, and once that visible slice fills, the system automatically reveals the next slice, and so on, so the market never sees the full weight at once. In the cash segment the disclosed portion cannot be less than 10 percent of the total order quantity, and the facility is not permitted in the futures and options segment (as of 17 July 2026; verify the current specification at the exchange). A large order there is instead sliced on the client side by the broker or a trading algorithm. Either way, the consequence for a depth reader is the same: a level can absorb far more than the ladder implied, because most of the order was never shown.

The visual tell is a level that refuses to die. You watch a modest quantity at a price get filled, expect it to disappear, and instead it reappears at the same size, again and again, as if the sellers were infinite. What you are seeing is one large iceberg refreshing its visible tip. The same effect is produced deliberately by execution algorithms, which take a big parent order and drip it into the market as a stream of small children timed to blend with normal flow, whether evenly through the session or in proportion to volume. Both techniques exist for a sound reason, to move size without announcing it and inviting others to trade ahead, and both mean the honest reading of a quiet, ordinary-looking level is that it might be shallow or might be the tip of something large. You cannot tell from the visible number alone.

In practice you rarely know for certain that a level is an iceberg; you infer it. The tell is a price that trades heavily yet refuses to move, with the same modest quantity reloading at the touch again and again while the tape prints far more volume than the visible size could ever explain. That is genuinely useful information, because it marks a price at which a large, patient participant is willing to transact, which can be a better place to work your own order than an empty level next door. But treat the inference with humility. The reload could equally be several independent traders refreshing their quotes, and because you cannot see the reserve you never truly know how much is left behind the tip. Read a refill as a sign of hidden willingness to trade at that price, not as a measurement of how much willingness, and never as a promise the supply will still be there for your next click.

Put the honest reading in one place. Depth is strong on the questions of liquidity and execution cost and weak, to the point of useless, on the question of direction. Keeping those two columns straight is most of what separates a trader who uses the ladder well from one it quietly misleads into trading the book instead of a plan. It can tell you whether an instrument is cheap or expensive to trade and how far your size will walk; it cannot tell you which way the price will break, nor guarantee that the size it shows will still be there when you act.

Notice that every entry in the left column is about cost and feasibility, and every entry in the right column is about the future or the truth of the display. That is the line to hold. The book is a precise instrument for the narrow question of what it will cost you to trade a given size right now, and a poor instrument for anything else. Traders get into trouble the moment they promote it from the first job to the second, reading a heavy bid stack as a forecast of a rally rather than as a note about where liquidity happens to be sitting this second.

Used honestly, market depth belongs to the execution layer of trading, the thin slice between deciding to act and getting filled. Its first and best use is sizing: before you send an order, the ladder tells you whether the resting quantity can absorb it near the touch or whether it will walk the levels and cost you, which is how you judge that you can enter and, just as importantly, exit your intended size without paying to move the price against yourself. Its second use is timing around obvious levels: a conspicuous resting order can mark a price where the market may briefly slow, so you place your own order with that in mind rather than blindly into it, while remembering it can be pulled. Both uses are refinements of a plan you already have, not a substitute for one.

Depth is not a standalone signal. It does not find the trade, it does not tell you which way price will break, and it does not set your stop or your size, which come from structure and a risk budget decided in advance. The ladder is the finest layer of detail on the screen, and it is the last thing to consult, after the level worth trading, the point that would prove the idea wrong, and the size the risk allows are already settled. Consulted in that order it sharpens execution; consulted first it tempts you to trade the book instead of the plan, which is how a tool for measuring cost becomes a generator of impulses. That upstream discipline, deciding what and how much before you ever open the depth window and treating the book only as the final check on cost, is exactly the sequencing that the method we teach is built around.

The book answers one question with real precision: what will it cost me to trade this size, right now? Promote it to any larger question and it stops being an instrument and starts being a temptation.

Strip everything back and the depth window measures one thing precisely: the cost of impatience. A market order says "fill me now, at whatever the book charges," and the book charges the spread plus the impact of your size walking the levels. In a deep instrument that toll is a rounding error and impatience is nearly free. In a thin one it is a heavy tax, and more than half your order may not fill at any sane price. The ladder does not remove that toll; it lets you see it before you pay it, which is the entire practical value of learning to read depth.

That reframing also points at the only two real cures, and neither is a cleverer read of the same book. If your size is walking the levels, you can make it smaller, so it fits inside the near liquidity and pays almost nothing in impact. Or you can be patient, using a limit order that refuses to pay worse than a price you set and waits for the market to come to you, accepting that it may fill slowly or not at all. Both trade urgency for cost. What does not work is staring harder at a thin ladder hoping to find a way through, because the problem is never your reading, it is that the liquidity to absorb your order is simply not there. Read from the middle out, size for the book that is actually in front of you, treat every large number as a snapshot that can vanish, and use the window for what it is: the last, sharpest check on the cost of a decision you have already made.

• NSE real-time market-data tiers. NSE differentiates the best bid and ask, the best five level market-by-price window (MBP-5) that most retail platforms show, a deeper twenty level view, and a full order-by-order feed carried on dedicated lines for latency-sensitive firms; the deeper feeds are premium or colocation-grade rather than the standard retail depth window. Verify current data products at source, as they change. nseindia.com
• Impact cost as the exchange measure of liquidity. NSE and its index arm use impact cost, the cost of executing a representative transaction relative to the ideal mid-price, as a liquidity measure and a criterion in index construction. This is the formal version of the volume-weighted walk computed in this guide. niftyindices.com
• Disclosed-quantity (iceberg) order specification. In the cash segment the disclosed portion of an order cannot be less than 10 percent of the total order quantity; the next tranche is revealed automatically once the visible slice fills, and the facility is not permitted in the futures and options segment, where a large order is instead sliced on the client side. This is the exchange-native iceberg that lets visible depth understate true size. Confirm the current specification at the exchange.
• SEBI PFUTP Regulations, 2003. Spoofing, placing orders with no intent to execute to create a false impression of demand or supply and cancelling them, is prohibited market manipulation, dealt with under Section 12A of the SEBI Act, 1992, read with the SEBI Prohibition of Fraudulent and Unfair Trade Practices Regulations, 2003 (as of 17 July 2026). sebi.gov.in