In crypto trading, speed and accuracy often determine success. Every millisecond can be decisive — it determines whether a trader can gain an advantage over their competitors. For market makers whose strategies are based on narrow spreads and high transaction frequency, colocation and sub-accounts have become critical tools. Not only do they reduce order execution times, but these tools also increase trading stability, optimize strategy efficiency, and enable risk management with maximum precision.
Instant Execution, Smarter Strategies: How Traders Use Colocation and Subaccounts
Colocation is the placement of a trader's servers as close as possible to the exchange's servers. The shorter the physical and network distance between them, the faster the order reaches the exchange and is executed. In high-frequency trading (HFT), even 1–2 milliseconds can be decisive in making a profit or a loss. But colocation is not only important for speed: it ensures the stability of strategies. For arbitrage between exchanges or market making with narrow spreads, it is critical that signals are processed instantly. Any instability in execution can lead to losses or missed opportunities.
Subaccounts are additional accounts within the main exchange profile, each with its own balance and order book. They allow you to:
separate strategies and portfolios,
limit risks in a certain direction,
test new strategies without affecting the main account.
A market maker can simultaneously conduct arbitrage and strategies based on order book imbalances, with each strategy operating autonomously and not affecting the others.
How Cutting 15–20 ms of Latency Can Add 8–10% to Your Strategy’s Profitability
A few months ago, I decided to take a closer look at how to reduce latency when trading on spot and futures markets, especially where high volatility leaves no room for error. In high-frequency strategies, even an extra 5–10 ms can eat into the spread or completely change the final outcome of a transaction. Experience quickly showed that standard API solutions, which usually seem “sufficient,” begin to lag behind at peak moments. And this lag does not apply to numbers in logs, but to real losses.
To understand which tools really give an advantage in the market, I compared the market making programs of leading exchanges. The conditions vary, but the trend is clear: rebates, flexible commissions, and technical capabilities are becoming the new competitive currency for professional traders.
Bybit offers discounts based on trading volume: for spot trading, from -0.001% to -0.0075%, and for futures trading, from 0.0028% to -0.0125%. There is also a separate program for connecting market makers to projects that need liquidity.
Gate.io offers discounts of up to -0.015%, access to the Market Maker Protection (MMP) system, and the opportunity to get an interest-free loan of up to 400,000 USDT to scale your strategy. Plus, 24/7 personal technical support.
Binance offers discounts on trading fees: the best participant can get a discount of up to 0.005%, and the next four — up to 0.002%. In addition, selected pairs are offered a 0% maker fee and increased API limits.
Bitget has a discount structure with maker fees ranging from -0.005% to -0.015% depending on monthly trading volume. Additionally, there are extended limits on sub-accounts and APIs, as well as dedicated technical support.
But what caught my attention the most was the WhiteBIT exchange. It offers low fees and discounts of up to -0.012% on both spot and margin trading, as well as something that is extremely important for HFT: colocation, sub-accounts for separating strategies, high-speed API with WebSocket, FIX, Webhook integration, and 24/7 personal support.
After doing the math, I realized that access to colocation reduces the average latency by approximately 15–20 ms. And these aren't just numbers in the logs — it's a noticeable difference in the speed of receiving data from the order book and the ability to take the first positions in the execution queue.
Add to this the ability to structure your work using sub-accounts: isolate independent algorithms, distribute risks between models, and test new approaches without unnecessary operational noise. Plus, with stable WebSocket, FIX, and Webhook channels, the problem of data delays simply disappears.
According to my calculations, the results can be quite significant: the profitability of the strategy will increase by 8–10% over the same period without increasing the risk profile. Spreads will become more stable, and execution will be almost instantaneous.
As a result,
this study has once again confirmed that in market making, it is not the one who comes up with the most complex strategy who wins, but the one who optimizes the infrastructure to the level where every millisecond works for them. Colocation and sub-accounts are not “additional options,” but fundamental tools that determine competitiveness today.
