High-Frequency Trading Orderbook

How do you make money from a stock without having any opinion about the stock?

Every other kind of trader has a view: the company is undervalued, the chart is about to break out, the internet is about to notice. High-frequency trading has no view. It trades the plumbing, EG: the queue of orders waiting at each price, the price difference between exchanges in milliseconds, the fact that someone, somewhere, is about to trade and you can be standing in front of them or out of their way.

This page simulates that plumbing. $STONK represents our fictional ticker on one exchange, at one-millisecond resolution. You observe a market maker -- the participant whose entire job is to be on both sides of the book at once.

Simulation

• One session shows twenty seconds of the book up close.
• The verdict displays 3 different makers on the exact same order flow and their Monte Carlo PnL.
• Pro screen is the whole session the way a desk would look at it.

Every row is a price, every column a moment, and the red/green brightness is how many shares were resting there, waiting. Most P&L charts show you what you made; this shows the thing you were standing in.

Every trade has a nice colour to label who took it -- noise, informed, sniper -- because the simulation is omniscient. Real flow is never labelled; how a maker tells the two apart after the fact is in How HFT makes money.

Trading strategies here

Description

Performance

• Human Scalper (Uncompetitive): Too slow to react. Its outdated quotes are frequently hit by faster market participants, forcing it to pay accidental "taker" fees. It performs poorly overall, especially during news events.
• Fast, Naive Maker (The Worst Performer): Because it is fast, it lands at the front of every order queue. However, without safety rules, it gets repeatedly hit by informed traders right before the price moves against it (adverse selection).
• Fast, Careful Maker (Consistent Winner): It succeeds by using two key defensive rules:

1. Inventory Skew: Pulls back bids or offers when holding too many shares, avoiding over-exposure.
2. News Pause: Steps out of the market for 150 ms during news events to dodge sudden price shocks.

Who Is the "Sniper"?

The Sniper is an ultra-fast automated algorithm located directly inside the exchange's data center.

• How it works: It sits idle until a public news event occurs. The instant news breaks, it sweeps up to 30 lots of stale quotes still sitting at the old price.
• The threat: If your algorithm cannot cancel outdated quotes faster than the Sniper can execute against them, the Sniper will take your money every time news hits.

Reading the PnL (panel 7)

Understanding a HFT

Every trader you have heard of holds a view on value, and they differ mostly by how far ahead they look and what evidence they use.

HFT is the only trader that trades flow instead of value. It makes the same money whether $STONK is a great company or a fraud, because it never holds $STONK long enough for that to matter.

WSB and HFT are the two ends of one axis: one holds forever on no information, the other holds for milliseconds on nothing but information.

"HFT" names a speed tier, not a strategy. Inside it live at least three different businesses -- market making, latency arbitrage, and very-short-horizon prediction -- and they make money in different ways.

How did HFT happened?

NASDAQ started as an electronic quote board in 1971 when humans still traded by phone. After the 1987 crash, the Small Order Execution System was made mandatory so retail could always get filled, and a group of traders promptly learned to exploit it faster than the human dealers could update their quotes.

The "SOES bandits" are the first people to make a living from being quicker than the other side's screen. Electronic crossing networks followed in the 1990s, and the SEC wrote them a rulebook of their own in 1998.

Two regulatory changes then built the modern market almost by accident. Decimalization in 2001 (NYSE in January, NASDAQ in April) which replaced fractions with pennies, and the bid-ask spread -- the human market maker's whole income -- collapsed from a sixteenth of a dollar to a cent. The only way to make a living on a one-cent spread is volume, and volume means machines.

Then Regulation NMS (adopted 2005, phased in through 2007) required orders to be routed to whichever venue showed the best price, which split US trading across a dozen exchanges and handed a permanent arbitrage to whoever could see all of them fastest.

By the end of that decade, machines were roughly half of US equity volume -- the SEC's own 2010 review put the estimates at "50% or more".

The rest is the arms race becoming physical: a straight-line fiber cable dug from Chicago to New Jersey (2010, about $300 million, three milliseconds saved), then microwave towers because fiber was too slow, then a book (Flash Boys, 2014) and an exchange built around a deliberate 350-microsecond speed bump (IEX, approved 2016). Along the way HFT firms simply became the market makers that exchange specialists used to be. The casualties:

How does HFT work?

For each price, the exchange keeps a queue of orders waiting to trade there -- bids below the current price, asks above it -- first come, first served.

A limit order joins the back of the queue at its price. A market order sweeps the best queue first, then the next, until it is filled. Whoever is at the front of a queue when a market order arrives gets the trade. This is price-time priority.

Queue position is an asset. Two orders at the same price are not equal. The one at the front fills on the next uninformed order that comes through. The one at the back fills only when everyone ahead of it has been taken. Being last in line means you get filled precisely when you would rather not be.

The spread is not a fee; it is a price. The gap between the best bid and the best ask is what the market pays a maker to stand there. It is set by competition among makers, and it goes exactly as low as adverse selection and inventory risk allow.

How HFT makes money (ordered by impact)

1. Market making: sell the spread, many thousands of times

Quote a bid and an ask around the mid; when a buyer arrives you sell at the ask, when a seller arrives you buy at the bid, and the difference is yours.

For $STONK, that difference is a cent. Then, the exchange adds a rebate for providing the liquidity that got taken. Multiply by hundreds of thousands of fills a day and the cents add up to become a business.

$0.002 rebate for the maker and a $0.003 fee for the taker per share are the US convention: Rule 610(c) of Regulation NMS has capped the fee for taking a displayed quote at $0.003 per share since 2005, and exchanges set their rebates just under it (see any exchange price list).

In the meantime, market makers watch out for adverse selection. Defined by price at the moment of the fill, subtracted by the midprice one second later.

When an uninformed trader hits you, the price wanders and you keep the spread. Low risk money.

When an informed trader hits you -- someone who knows where the price is going -- the price keeps going through your quote, and the fill was not income; it was the bad news arriving. A fill is evidence that you might have been wrong, and the market maker's real skill is telling the two kinds of fills apart and pricing the second into the spread.

This is also why the biggest HFT firms pay retail brokers for their customers' orders (PFOF). Retail flow is uninformed, so its fills low risk.

How do you tell informed flow from noise? You do not -- not at the moment of the fill. The simulation labels every trade because it is omniscient.

In real life, a maker has to count thousands of fills and sorts them by everything it can observe -- which broker the order came through, its size, how soon after a news print it arrived, which venue -- then "toxic flow" is labeled after the maker recognizes fills that kept losing.

2. Latency arbitrage: be first when the world moves

The same stock trades on many exchanges. An ETF and its basket are the same thing. A future and the spot it settles to are the same thing. When one of them moves, the others are momentarily wrong, and whoever reaches them first collects the difference.

In the simulation, anyone (not including you) who moves this fast is the sniper. When news moves $STONK's true value, every quote left on the wrong side is stale, and the sniper takes all of them -- at the old prices -- before their owners can cancel.

Whether your quote is among them is decided by one comparison: your latency against theirs.

3. Short-horizon prediction

The queue itself leaks information. Suppose a best bid has 10 times the size of the best ask, it is normal to assume the next tick is more likely to be up.

Strategies that trade that milliseconds-ahead statistics are like trying to catch a wave before it arrives.

Isn't this pennies in front of a steamroller?

By definition, yes: tiny gains, and maybe a large loss. How does the market maker not get flattened?

The trades that do get flattened -- short volatility, leveraged carry, the funds that blew up in 1998 and 2018 -- share one property HFT does not have: commitment. IE: they hold a position they cannot sell when the steamroller arrives.

The market maker is a penny-picker with no commitment, and the business is engineered around that fact:

1. Risk is position × time, and they hold neither. A crash takes seconds to minutes to develop; a maker's exposure per fill is a few hundred shares for a few hundred milliseconds, and it does not keep what it buys: there is a hard cap (ten lots here). When it is long, its quotes shift so the next customer is more likely to buy from it than sell to it, so the position drains away on its own.
2. The spread is the steamroller premium. They are literally paid to stand in front of a steamroller and get out of the way before it hits. If no one dares to stand, the spread gets wider until it starts to be worth it.
3. They have the option to not be there. When news hits, or volatility jumps, a maker can quote more carefully: wider, smaller, or not at all. If enough makers do that, the spread widens until some other maker is greedy enough to take the risk at the new price. The 2010 flash crash was every maker stepping back at the same moment -- with nobody left to take the other side, prices fell through the hole until buyers were found far below. Their safety is the market's fragility.
4. Speed reassigns the steamroller to someone slower. When news hits, the fastest firm can cancel its quotes in a millisecond; the same event then runs over whoever was still quoting. If you are fast enough, you can play hot potato with anyone without burning your own hand.
5. The law of large numbers reduces their risk. Trade hundreds of thousands times, and your P&L stabilizes to a range you can accept.

What actually runs them over is operational, not directional: a deployment bug that kept firing orders cost one firm $440 million in 45 minutes in 2012; bad data feeds, exchange outages, or a risk module that fails launch.

The classic penny-picker is a man who picks up slowly from the road. The market maker is someone who runs fast enough to pick up any penny, and the only thing that reliably kills him is his own shoelaces.

The engineering

Latency is the time from something happening in the world to your order arriving at the matching engine.

It consists of multiple components, including the feed reaching you, your decision, and your order.

It matters because after news, every quote resting at the old price is free money for whoever reaches it first. Your quote survives only if your cancel lands before the fastest predator's order.

The cost of being slow has a shape:

Only the last factor depends on speed, and it is a cliff, not a slope: faster than the sniper and we are that sniper, slower and we are sniped at every news event.

The cliff decides whether we pay the news tax, not whether we can still be profitable. A sniped fill costs us (one lot times the jump, minus the spread we were paid), and whether that tax exceeds our income depends on how often news comes and how big the jumps are, against how often we get filled and at what spread.

Since ranking, not speed, determines the winner, the race has no finish line. Every expense on tryign to win the latency game is a deadweight tax on everyone who trades -- the formal version is Budish, Cramton & Shim (2015), and the races have since been measured on exchange message data by Aquilina, Budish & O'Neill (2022):

1. (Left) There's a sniper with a fixed speed. If you're faster than him, he never catches your quotes — you lose nothing to him. If you're slower, you lose all quotes to him.

2. Past the step you're already getting caught every time. So why does revenue keep going down as you get slower? Because slow quotes are wrong quotes — the price moved and yours didn't. And why does it flatten out near zero instead of going to minus infinity? Because a really slow quoter barely trades at all. Fewer trades = less to lose. You're not winning, you're just not playing.

3. The grey line (20 ms sniper) starts to fall way before 20 ms. This is because the sniper isn't the only guy hunting you. The informed trader may hit stale quotes, or reacts a few milliseconds after news at random times, not on a fixed clock. With the slow sniper, the informed guy now becomes the fastest hunter.

4. (Right). A naive maker never generates revenue at any speed because he isn't cautious for informed flows. His worst spot is 5 ms: still fast enough to get filled constantly, too slow to be right about the price. Past 5 ms his line "recovers", not because it starts to make money, but because he quits.

How much does it cost to be fast?

The simulation lets you pick a latency with a slider. In real life, the slider is expensive. This is how HFT becomes a high CAPEX business instead of being smarter than your rivals and why individuals should give up day trading.

Why do people buy radio towers to solve a math problem?

Because in the late game, the race is decided by who arrives first, and arriving first is a matter of distance and hardware, not cleverness.

Light has a speed. In glass it travels at about two-thirds of its speed in air. Chicago's futures exchange and New Jersey's stock exchanges are 1,179 km apart, so light in a vacuum needs 3.93 ms one way, about 7.9 ms for the round trip.

Microwave towers on the straight line got within 0.1 ms of the physical limit by 2012, about 8 round trip, at the cost of low bandwidth and outages in rain (Laughlin, Aguirre & Grundfest 2014, who measured the drop in the timestamps of the trades themselves).

A microsecond saved in a chip is worth nothing next to a millisecond saved by a shorter path, so the money goes into tower leases and radio licenses.

Software can be copied; a tower on the straight line cannot. There is even a physics paper computing the best places on Earth to put trading nodes given the speed of light (open-access copy).

Simulation details

What the simulation ignores

What each of our parameters do

Glossary

References

Models the simulation is built on:

• Glosten, L. & Milgrom, P. (1985). Bid, ask and transaction prices in a specialist market with heterogeneously informed traders. Journal of Financial Economics 14(1), 71-100. The adverse-selection logic behind the fill microscope.
• Avellaneda, M. & Stoikov, S. (2008). High-frequency trading in a limit order book. Quantitative Finance 8(3), 217-224. The inventory-skew rule the careful maker uses.
• Menkveld, A. (2013). High frequency trading and the new market makers. Journal of Financial Markets 16(4), 712-740.
• Budish, E., Cramton, P. & Shim, J. (2015). The high-frequency trading arms race: frequent batch auctions as a market design response. Quarterly Journal of Economics 130(4), 1547-1621. Why the race is a tax.
• Aquilina, M., Budish, E. & O'Neill, P. (2022). Quantifying the high-frequency trading "arms race". Quarterly Journal of Economics 137(1), 493-564. Latency-arbitrage races measured on exchange message data: about one per minute per stock, most decided in 5-10 microseconds.

History and figures:

• Library of Congress, Wall Street and the Stock Exchanges: historical resources -- NASDAQ's 1971 launch as an electronic quotation system.
• Harris, J. & Schultz, P. (1998). The trading profits of SOES bandits. Journal of Financial Economics 50(1), 39-62.
• SEC (1998). Regulation of Exchanges and Alternative Trading Systems, Release 34-40760.
• SEC (2000). Order directing the exchanges and Nasdaq to phase in decimal pricing; NYSE completed the switch on 29 January 2001, NASDAQ on 9 April 2001.
• SEC (2005). Regulation NMS, Release 34-51808; Rule 610(c) access-fee cap, 17 CFR 242.610; 2024 amendments reducing the cap to $0.001; compliance deferred to November 2026 by SEC Release 34-104172 (31 October 2025) and to November 2027 by the order of June 2026, 91 FR 36022.
• SEC (2010). Concept Release on Equity Market Structure, Release 34-61358 -- HFT share-of-volume estimates "typically exceeded 50%".
• CFTC & SEC staff (2010). Findings Regarding the Market Events of May 6, 2010.
• Spread Networks (2010), the Chicago-New Jersey fiber; Laughlin, G., Aguirre, A. & Grundfest, J. (2014). Information transmission between financial markets in Chicago and New York. Financial Review 49(2), 283-312 -- the fiber-then-microwave latency drop, measured.
• Knight Capital Group (2012). Press release on the 1 August 2012 trading loss -- roughly $440 million in about 45 minutes.
• SEC (2016). Approval of IEX as a national securities exchange, with its 350-microsecond speed bump.
• Lewis, M. (2014). Flash Boys: A Wall Street Revolt. W. W. Norton.

Open-source references used for the screens:

• Databento, How to build an order book (DOM) visualization -- the price-ladder idiom.
• OrderFlowMap -- a Bookmap-style liquidity heatmap in one HTML file.
• hftbacktest -- queue-position and latency models for backtesting market making.
• ABIDES -- agent-based market simulation with noise, value and market-maker agents.