Inter-exchange arbitrage is not about predicting the market, but about the mathematics of liquidity. The same asset is priced differently on different platforms, and this difference is a source of income. But there is a nuance: the classic spread rarely exceeds 0.5%, and it lives for only a few seconds. Manual trading loses here before it even starts—a person simply cannot react in time to changes in the order book.
The Arbitron platform solves this problem radically: it takes over the entire cycle, from monitoring to execution. The service tracks quotes and funding rates on 20 exchanges and automatically places both sides of the trade (the so-called "legs"). At the same time, funds remain on the trader's exchange accounts, eliminating custodial risks.
Why the spread is not always profit
The raw price difference shown by any scanner is deceptive. Four factors "eat" it before real profit: taker fees on both legs, insufficient order book depth, slippage during execution, and the natural decay of the spread itself. For example, a full two-leg trade cycle involves four market orders, and platform fees decide everything here. Choosing the right exchange pairing is already half the success.
The key issue is speed. The spread lives exactly until other arbitrageurs close it. A scanner updating every 10–15 seconds already shows history. That is why the platform bets on infrastructure: each account gets a dedicated AWS server with a static IP in the region closest to the exchange engines. Measurements show that a request to Binance from Tokyo takes about 23 ms, while from Singapore it is already 206 ms. During that time, the order book of a liquid futures contract fully refreshes, and the order arrives at a book that no longer exists.
Backtest instead of a showcase
Instead of the usual screener with "pretty" percentages that are impossible to reach, Arbitron offers a scanner based on backtesting. The platform runs the last 8 hours of recorded quotes through the same strategy used by trading cards and subtracts fees. The 5-minute delay is a deliberate filter that weeds out illusory opportunities. Each row in the table is a simulation that accounts for the trader's personal VIP level and discounts for the native token.
Special attention is paid to reliability assessment. A rating from 0 to 100 considers the number of closed cycles, profit margin, and market depth. If the score is below 50, it is a clear "yellow flag," and I would recommend treating such opportunities with extreme caution.
Trading cards and protection from chaos
The work revolves around trading cards, each linking a pair of exchanges, a coin, and two spread thresholds. The platform handles divergence in either direction: if the spread moves up, one configuration opens; if it moves down, a mirrored one. This is an important advantage, as many bots are rigidly tied to a single scheme and miss half of the movements.
The protection system is thought out to the smallest detail. Three levels of safeguards—from a hard stop to delisting. The latter, by the way, is critically important: when an exchange removes a contract, one leg disappears, and the position is left without a hedge. The platform tracks delisting announcements through official APIs and notifies the trader but does not close positions automatically—so as not to mistake an API failure for an instrument removal and sell at a random price.
Keys and fees
Security here is built on the "envelope" principle: API keys are encrypted with AES-256-GCM, each user has their own key, and withdrawal rights are not requested at all. This is a non-custodial model where access is limited to trading only.
The platform's financial model is also unconventional: a fee is charged only on realized profit, and even then, once a week. Losses are carried forward, so a week without income costs nothing. For me, this is the most honest approach on the market—the service earns only when the trader earns.
My verdict
Arbitron is a tool for those ready to delegate routine to algorithms but still understand the basic principles of futures trading. A starting capital of $3000 is a reasonable minimum, and setup from half an hour to several hours pays off with stability. In a world where speed decides everything, such platforms become not a luxury but a necessity for systematic trading.