A modern matching engine can accept, match and acknowledge an order in a few microseconds. The message that carries that order to the engine travels down a cross-connect cable measured in metres, because the firm that sent it rents rack space in the same building. Everything about high-frequency trading follows from those two facts: the decision has to be made inside the time it takes light to cover a short distance, and the only way to compete is to shorten the distance.
For a retail trader this world is mostly invisible, and the popular version of it is wrong. HFT firms are not sitting on your stop loss. They are competing with each other over fractions of a spread, thousands of times a day, on instruments where the tick is worth a few cents and the edge per trade is smaller than a rounding error.
What high-frequency actually describes
High-frequency trading is not a strategy. It is a constraint on how a family of strategies is implemented. The common ones are electronic market making, where a firm quotes both sides and earns the spread while hedging its inventory, and statistical arbitrage between correlated instruments or between the same instrument on different venues. Neither idea is new. Floor market makers did the first one with hand signals. What changed is that the profit per trade shrank and the number of trades exploded, so the whole thing became an engineering problem.
The engineering is where the money goes. A serious desk pays for colocation inside the exchange data centre, direct raw market data feeds rather than consolidated ones, microwave or hollow-core fibre links between financial centres, and increasingly order logic burned into FPGAs so a decision never has to touch an operating system. That is a fixed cost running into the millions per year before a single trade is placed. It is why the field consolidated into a handful of firms rather than spreading out.
How this reaches the retail order book
In spot FX and CFDs there is no central exchange. Prices come from a set of banks and non-bank liquidity providers streaming quotes into aggregators, and your broker sits on top of that stack. Many of the largest non-bank providers are, functionally, high-frequency firms: they quote continuously, hold inventory for seconds, and hedge it against other venues. Read how liquidity providers work with brokers and you are reading a description of speed-sensitive market making with the label removed.
The practical result for a retail trader is a tighter quoted spread in liquid hours and a quote that can be withdrawn very quickly when conditions change. Both come from the same source. A firm willing to quote a half-pip spread in EUR/USD can only do so because it can cancel and requote faster than the market can move against it. Take away the speed and the same firm quotes wider, or does not quote at all.
Faster quoting does not mean deeper quoting. A tight spread on a small displayed size is a different thing from a tight spread you can trade in size, and the difference shows up exactly when you need it, in the seconds around a data release.
The part that does affect you
Where speed genuinely reaches a retail account is in the handling of your own order after you press the button. That chain has several hops: your device to the broker's gateway, the gateway to the aggregator, the aggregator to whichever provider is quoting the best price, then an acknowledgement back. Each hop adds time, and price can move during it. This is the ordinary origin of slippage and, on platforms that still use them, of requotes.
Nothing in that chain is improved by a faster strategy. It is improved by a shorter physical path and fewer intermediate steps, which is why traders running automated systems put them on a VPS near the broker's servers rather than on a laptop at home. It is also why a broker's technology choices matter more to fill quality than the trader's do. A platform that batches, queues or round-trips orders through an extra layer adds tens of milliseconds that no trader can recover.
Why copying the model does not work
People occasionally try to build a retail version: an expert advisor that scalps a one-pip target hundreds of times a day, run on a low-latency VPS. It fails for structural reasons rather than bad luck. The strategy earns less than one spread per trade in gross terms, so the trading cost consumes it before anything else happens. Volume-based rebates, which are a real part of professional market making revenue, are not available at retail size. And brokers running such flow will widen or restrict the account, because a book of one-pip scalps is unhedgeable at their cost base.
If you want the honest comparison, work through the arithmetic in spreads, markups and commissions. Any strategy whose expected edge per trade is smaller than the round-trip cost is not a strategy, it is a subscription to your broker. High-frequency firms solve that by having a cost base that retail traders cannot reach, not by trading better.
What to take from it
The useful lesson from HFT is about market structure rather than tactics. Displayed liquidity is conditional. Spreads that look free in the middle of a London session are being priced by machines that can step away in microseconds, and they do step away around scheduled news, at rollover, and when a venue's book thins out. Position sizing that assumes a stable spread is making an assumption the market never agreed to.
That is also why serious execution analysis at firm level looks at fill quality distribution rather than headline spread. A broker choosing between providers should be measuring rejection rates, hold times and the size actually available at the top of book. Retail traders can do a rough version of the same thing by logging their own fills against their intended prices and looking at the pattern over a few hundred trades rather than the last three bad ones.
Leveraged trading carries a high risk of loss regardless of how fast anyone else is. Speed is not the variable that decides a retail outcome. Cost per trade, position size and the discipline behind them are, and those are entirely inside your control.
"Retail traders worry about being picked off by HFT. They almost never are. The speed race happens between the firms quoting prices, and the trader just inherits the tighter spread it produces."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- HFT is a category of latency-sensitive market making and arbitrage, not a strategy retail can copy with a faster broker.
- The speed competition happens between quoting firms and venues, not between an HFT desk and an individual retail order.
- Colocation, direct feeds and hardware-level order handling are the real cost of entry, and they run into the millions per year.
- Retail execution quality depends on broker routing and liquidity relationships, not on shaving milliseconds off a home internet line.
Frequently Asked Questions
Can a retail trader do high-frequency trading?
Not in any meaningful sense. The strategies depend on colocated servers next to the matching engine, direct exchange feeds and custom hardware, which costs far more per year than a retail account holds. A retail trader with a VPS is reducing latency for order delivery, which helps execution, but it is not the same business.
Does HFT trade against my orders?
In FX and CFDs your counterparty is your broker or the liquidity provider behind it, not an HFT fund reading your stop. Speed-sensitive firms make money quoting and hedging at scale, and a single retail order is too small to be worth targeting.
Is high-frequency trading bad for ordinary traders?
The evidence goes both ways and depends on the venue. Fast market making has generally narrowed quoted spreads in liquid instruments. It has also made displayed liquidity less durable, so a large order can find that the price it saw was gone before it arrived.
About the Author
Alex Onta is an Executive Director at SINGUARD. He built eTrader, the terminal, the mobile apps, eTrader Broker, Copytrading, Business and Community, along with the worldwide clustered-server infrastructure it all runs on, with his brother Roman Onta helping on the design, and he leads that division today. Together with Roman he builds the Prop Firm CRM, the Broker CRM, Scalegram and CopySignals, and the two of them carry worldwide compliance, payment processing and international business structuring side by side. He lives and works in Dubai for most of the year. Meet the executive duo leading Singuard's five divisions.