Algorithmic trading firms — the forex market makers

A decade ago, the price on EUR/USD was set almost entirely by large banks. Today a sizeable share of the quotes on the currency market comes from firms the average trader has never heard of: XTX Markets, Citadel Securities, Jump Trading and Virtu Financial. These are not banks but specialised electronic market makers that compete on algorithms and reaction time rather than on the size of a balance sheet. This article explains who they are, how they post prices and what their presence means for spreads, liquidity and your own account.

Who the non-bank market makers are

For years, making markets in currencies was the banks' domain. To quote on both sides and carry the risk yourself, you needed an enormous balance sheet, and only the largest institutions had one. Over the past decade that picture has cracked. A group of firms appeared that do not lend money, do not hold deposits and do not run branches — they do one thing: they continuously post bid and ask prices across dozens of instruments at once, earning on the difference between them.

The loudest of them is XTX Markets — a London firm that for several years has ranked at or near the top of the global tables for share of foreign-exchange turnover, regularly placing ahead of most banks. Alongside it work Citadel Securities, Jump Trading and Virtu Financial — firms that grew out of the US equity and futures markets and carried their technology over to currencies. What they share is that they are non-bank liquidity providers and that their edge is not capital but engineering.

One thing is worth settling straight away: these firms did not replace the banks, they stepped in alongside them. The hierarchy I describe under tier-1 dealers still holds — except that at the top, next to the dozen-or-so banks, a handful of new players now appear, supplying much of the liquidity in the most liquid pairs.

How they make markets with algorithms and speed

The mechanism fits in one sentence: a firm posts a bid price and an ask price at the same time, and earns on the spread multiplied by a vast number of trades. The profit on any single one is microscopic — a fraction of a pip — but repeated millions of times a day it adds up to real revenue. The whole craft lies in updating that price faster and more accurately than the competition, and in never being left with an unbalanced position when the market suddenly moves.

Hence the two pillars of the business. The first is co-location — physically placing your own servers in the same data centre that houses the engine matching the orders. Every metre of cable is microseconds of delay, and in this game microseconds decide who gets to refresh a quote first. The second pillar is internalisation: instead of immediately sending an order out to the market, the firm matches its own clients' opposing orders in-house. If one client is buying euros while another is selling at the same moment, the market maker nets it internally and never has to step outside at all.

"The growth of execution algorithms has changed the way liquidity is provided to, and taken from, the foreign exchange market." — BIS Markets Committee, FX execution algorithms and market functioning, 2020.

The competition, then, is not over who has the bigger balance sheet but over who has the better model and the shorter delay. That is an entirely different logic from a classic bank — and it is precisely why a firm employing a few hundred engineers and physicists can compete for market share with an institution that has hundreds of thousands of staff.

What it changes for spreads and liquidity

The most easily noticed effect works in the trader's favour: in normal conditions, spreads in liquid pairs are tighter today than in the days when banks made the market alone. More entities posting prices means sharper competition for every order, and sharper competition usually narrows the gap between the bid and the ask. You pay less just to get into and out of a position.

There is, however, a flip side, and it has to be understood. These firms manage risk dynamically — their algorithms constantly compute how much position they can hold. When volatility spikes, the model does what it is meant to do: it pulls back its quotes or widens the spread sharply to limit the risk. The result is that liquidity which looks deep on a quiet day can thin out at exactly the worst moment — during a major data release or a sudden shock.

This is not a charge against these firms — a tier-1 bank behaves much the same under stress, because no one wants to stand on the wrong side of an avalanche. The mechanism is worth remembering, though, because it explains something every trader who has traded a payrolls release or a central-bank decision already knows: the very instrument that cost a fraction of a pip in the morning can carry a spread many times wider the second the news lands.

What this means for an individual trader

The most important message is this: from the vantage point of your own account, the presence of these firms is essentially invisible. You do not trade with XTX Markets or Citadel Securities directly — you trade through a broker, and whether it reaches for non-bank liquidity you will see only indirectly: in how tight your spread is and how good your execution feels, especially during the busiest hours of the market.

From that follows a practical conclusion. You stand no chance of racing these firms on their own ground — on fractions of a second and on infrastructure worth tens of millions. Exactly the same distance separates an individual trader from the large players in other segments of the market; I wrote about it when comparing the retail and institutional sides. Your edge lies elsewhere: on higher timeframes, in patience, and in the analysis of fundamentals that no latency algorithm will price for you. It is also worth knowing that non-bank market makers are not the same as funds aiming to profit from a move in price — I explain the difference in goals and method under hedge funds in forex.

If you want to see how the various participants fit together in one place, the market participants overview on ForexMechanics maps out the full liquidity chain — from banks to funds and algorithmic trading firms.

What to do after closing this article

1. Check your broker's execution model. Go to its website and look for information on its liquidity providers and on whether it runs an A-book (orders routed out to the market) or a B-book (netted internally). This is one of the most important things to know about the place where you keep your money, and one that most beginners never check.
2. Measure the EUR/USD spread at two different moments. Note its value once in the quiet middle of the London session, and once in the exact minute of a major macro release — the US labour-market data, for example. You will see with your own eyes how the liquidity supplied by market makers can retreat as risk rises.
3. Match your trading style to your real edge. Instead of fighting over seconds on a one-minute chart, move your analysis up to the hourly timeframe or higher and base decisions on fundamentals and on market structure. That is the ground where an algorithm's reaction time stops mattering and what you can think through starts to count.