Triangular Arbitrage in a Zero-Transaction-Cost Currency Contest

Summary

The document describes a simulated real-time currency trading competition in which teams could request current prices and submit buy or sell orders through a server. Participants began with equal holdings across six currencies, and trades incurred no transaction costs. The author’s team built a bot that checked three-currency conversion cycles for inconsistent prices and traded when it found an arbitrage opportunity. The team placed fifth among 35 participants, while the top teams reportedly earned far more.

The text asks what strategy might have performed best under these contest conditions, but does not explain the leading teams’ methods or provide enough price, timing, or order-handling data to identify a superior strategy. Its account offers a concrete example of triangular arbitrage in a simplified setting, while leaving open whether other opportunities, faster execution, or different trading logic drove the rankings. The result should not be treated as evidence that any particular strategy will generalize to currency markets with transaction costs or real execution constraints.

Key ideas

• The competition allowed immediate currency trades without transaction costs.
• The author’s bot searched three-currency cycles for inconsistent exchange prices and traded on detected opportunities.
• The team ranked fifth among 35 participants, while the leaders reportedly earned substantially more.
• The document does not reveal the winning strategies or explain the performance gap.
• Results from a cost-free simulation may not carry over to markets with execution constraints.

Tags

# Best strategy to maximize Profit if no transaction cost?


# Best strategy to maximize Profit if no transaction cost?












I was recently in a competition which simulated real time currency trading. Teams were supposed to build bots that could request current prices of currencies, buy, or sell currencies using HTTP request. The objective was at the end of the competition, teams that gained the most profit wins. Each team was given an initial 500 units from each currency (Total of 6). Also, transactions incur no cost. (Basically once the server receives your request to buy/sell, it performs it right away if valid).

Me and my team developed a bot that used triangular Arbitrage. We request the prices, then try all possible three ways, if any is non consistent, we arbitrage it.

Now, me and my team were relatively successful and we ranked fifth (Out of 35 teams). However, I couldn't help but see that the top three teams generated ridiculous profits (15 times of magnitude as us). So my natural question is, what strategy would've been the best in this sort of scenario?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.