DEUS Journal
Published in

DEUS Journal

Redefining on-chain FX trading

Sourcing FX liquidity where cheap and abundant, and bridging it on-chain for the benefit of traders and xDEUS holders.


Before the Gains network, there was virtually no FX trading on-chain. Their innovative approach, trading FX on virtual liquidity, changed the landscape.

Virtual liquidity means that all trades settle at oracle prices rather than at an internal price, and all trades happen inside a closed system. This technique can thus be called unhedged virtual-AMM, hereafter FX-vAMM.

When the FX-vAMM books are balanced (long open interest = short open interest), traders are betting one against another; when the books are unbalanced (e.g. most of the traders are long), they bet against the protocol itself (more precisely against a user-provided DAI vault, collateral to all trades).

FX-vAMM has revolutionized FX on-chain trading, but it has strong limitations: since unbalanced order books are risky for the protocols, positions cannot scale, and holding positions becomes costly for traders; trading might even be halted when deemed risky for the protocol, for instance around news releases.

DEUS bridges cheap and abundant liquidity from traditional financial markets into crypto markets. The RFQ (request for quote) technology allows users to keep trading no matter the market volatility. Trades are executed with true market impact, which matters for any trader who wants to react to news or “trade the news”.

Virtual liquidity vs market liquidity

While the concept of virtual liquidity is elegant and has great cost advantages since it does not require trading infrastructure, it also has limitations that the RFQ (request for quote) system introduced by DEUS aims to solve.

DEUS RFQ grants access to market-wide liquidity, while FX-vAMM virtual liquidity models are limited by the collateral available in the system. RFQs and market-wide liquidity enable true market impact and unlimited open interest whenever this matters.

Virtual liquidity works as long as there is overall no open interest (OI) in the books. Traders just bet one against another and the protocol takes a fee to facilitate bets.

When books are unbalanced (more long OI than short OI), traders as a whole take a large position against the protocol, which becomes increasingly risky for the protocol itself. In an effort to reduce the risk, the protocol must then charge hefty funding and rollover fees so that net open interest goes back to zero.

The risk for the protocol is even higher when the market is volatile, when liquidity is thin, or when news may induce price jumps.

  • Volatility means a greater risk of loss on unbalanced books, so borrowing and funding rates tend to increase with ex-post measured volatility, and can potentially lead to the liquidation of a profitable position.
  • News releases potentially introduce jumps, akin to local extreme volatility, so unhedged virtual AMMs either charge a discouraging trading fee (a huge simulated price impact) or halt trading altogether.

Overall, virtual liquidity is a model that works ideally when the market has low volatility and the community of traders has little conviction; it can hardly scale up in size and is very costly for traders in volatile environments.

Cost implications

DEUS costs of sourcing liquidity on-chain, and of executing FX trades are defined by its partner brokers’ prices. The gap between a typical FX-vAMM fee structure and DEUS input costs shows the potential for both DEUS profit margins and improved trader experience.

This table contrasts the pricing of a partner broker with that of a typical FX-vAMM. An individual trade’s execution cost is expressed in basis points (100bps = 1%), while the costs of borrowing and funding are annualised. On the blockchain, borrow and funding costs are calculated hourly (which is 365*24 times lower than annually). In contrast, traditional CFD brokers charge borrow costs on any overnight positions, thus “borrow costs” are commonly referred to as ‘rollover costs’.

This table shows a partner broker’s pricing, compared to that of a typical FX-vAMM protocol.

One clearly sees that:

DEUS provides more affordable liquidity than FX-vAMMs, which translates into cheaper borrowing and funding rates.

  • DEUS liquidity costs depend on interbank rates, typically 2%-5%
  • FX-vAMM borrowing rate range from 10–200%, and can be quite volatile. Furthermore, the funding rate, can eat up profits from winning trades.

DEUS can guarantee FX is executed with true market impact, while FX-vAMM simulated market impact spirals off so as to discourage trading when the market is volatile.

Despite the fact that FX-vAMM can have low fees when the market is not volatile, any size of trade position can be very expensive for traders particularly when the market is volatile. Once the FX-vAMM books become unbalanced, traders must pay an additional cost to bid against the protocol (or its user-provided collateral). In fact:

  • At FX-vAMM, simulated market impact is deliberately set off-chart to prevent trading when there is potential price action.
  • Consequently, trading is usually disabled around time important news is announced.
A snapshot of deep, market-wide liquidity taken by Silas Baisch (unsplash)

A solution for traders and protocols

DEUS tech provides a solution that can be used universally, both for protocols and for traders.

Deus tech for traders

The DEUS technology is ideal for traders who are looking to trade in larger volumes, take advantage of news releases or market volatility, or enter strategic trades without fees eating up their profits.

  • When market conditions are ideal, with low volatility and limited long and short OI, FX-vAMM protocols such as the GNS can provide competitive trading conditions.
  • DEUS technology is the only one that can be used in times of high market volatility or when there is significant price action.
  • Furthermore, it is the only one that offers the ability to reliably maintain low rolling costs over a period of days, making it ideal for strategic trades.

Deus tech for protocols

FX-vAMM models can also use DEUS as a bridge to access deep liquidity from traditional finance markets when necessary.

For instance:

  • suppose the maximum open interest for an FX-vAMM protocol is $5m. Close to this threshold, a protocol could entirely hedge its positions with DEUS, and scale up.
  • FX-vAMM protocols tend to stop trading during news releases because of the larger risk that bears on them. They could redirect trades

Virtual liquidity protocols can benefit from the assistance of DEUS to expand their reach, make sure their traders can trade around news releases, and hedge their own risks.

Deus and partners shaking hands, photo taken by Austin Kehmeier (insplash)

The overall potential profit margin for DEUS

The online FX market has a tremendous amount of untapped potential, making it an ideal area for exploration. DEUS, the infrastructure that allows for connecting inexpensive tradfi liquidity on-chain, is set to finally allow the FX on-chain market to expand.

If the current on-chain daily trading volume is only tripled to $150 million, at least $100 million daily will be traded through DEUS. We’ll use this assumption below.

Stable borrow rates are beneficial for traders and an important source of revenue for DEUS

DEUS’s true cost of liquidity is 7–8% overnight, or 4% on a continuous basis if half of the traders do not maintain overnight positions

The most competitive on-chain borrow rates are around 12%, charged per hour, and FX-vAMM protocols charge varying borrow rates, typically in the 10–200% range.

Predictable borrow rates under 25% would be very competitive for FX traders, while a recurring source of revenue for the DEUS protocol since it can represent a 20% profit margin.

With an average total (long plus short) OI of $50m, a yearly $10m revenue would accrue to the DEUS ecosystem.

DEUS is the sole tech to offer true price impact, and ability to trade around news releases and in volatile markets.

DEUS could choose either a premium flat fee structure (eg fixed 5 basis points fee for all foreign exchange transactions) or a variable fee structure (e.g. 1bp during high-liquidity hours and 10bp in low volatility hours).

To give an order of magnitude, a flat fee 5bp trading fee would generate an annual income of $1.65 million in revenue, supposing a daily FX trading volume of $100 million.

Potential on-chain FX market

The potential online FX market is huge but largely untapped:

  • the FX market trades $7.5Trn per day even though the greatest fraction of the world population lacks access to traditional financial markets (and still needs to hedge their risks).
  • Onchain FX trading is still relatively small, with Gains Protocol accounting for the majority of the $50m traded daily.
A snapshot of fees and costs as described by many protocols.

Appendix III: formulas and examples of GNS traders’ costs

The dominant FX on-chain protocol, the Gains Network, gives an idea of competitive trade execution costs, as well as borrowing and funding costs for the unhedged virtual AMM trading tech.…/fees-and-spread

Trade execution cost can be decomposed for FX as:

  • open and close costs: 1.2bp for FX
  • fixed trading cost, called “fixed spread”, which is 1bp for FX
  • price impact, which depends on a parameter called “market depth”

FX trades on fixed spreads (no price impact on top of the 1bp fixed spread) apart from:

  • Major news: 1% depth is set to 250k on relevant pairs — for instance from 1h before to 10 min after a CPI release.
  • Market closing: 1% depth is set to 250k on all pairs 30 minutes before the market closes.
  • Low liquidity session: 1% depth is set to 10m on all pairs for 2 hours after New York close.

Price impact is called “Dynamic Spread”

Formula: (called “Dynamic Spread”) (%) = (Open interest {long/short} + New trade position size / 2) / 1% depth {above/below}.

Price impact can be decomposed as:

  • Price impact (%) from resulting pre-trade OI. Pre-trade OI / 1% depth
  • Price impact (%) of the trade itself: New trade position size / 2 / 1% depth
A snapshot of price impact on virtual, thin liquidity taken by Nicolas Ruiz (unsplash)

Example 1: suppose a $100k notional trade on a book in the two hours after NY market close:

  • with an OI of $1m in the direction of the trade, price impact is ($1m+$100k/2)/$10m > 0.1% = 10 bp
  • with an OI of $4m in the direction of the trade, price impact is ($4m+$100k/2)/$10m > 0.4% = 40 bp

Example 2: suppose a $100k notional trade on a book during news releases or market close (1% depth assumed to be $240k)

  • with an OI of $1m in the direction of the trade, the price impact is ($1m+$100k/2)/$240k > 4% = 400 bp
  • with an OI of $4m in the direction of the trade, the price impact is ($4m+$100k/2)/$250k > 16% = 1600 bp



Architecture for sustainable and scalable derivatives trading. Connecting users and brokers directly, solving counterparty risk by utilizing bilateral agreements and offering instant, on-chain settlement.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store

Advisor to DEUS Finance. Focus on incentives, tokenomics and stability