Pith. sign in

REVIEW 4 major objections 6 minor 56 references

This paper claims that multi-currency constant-mean AMMs, with optimally chosen pool weights and correlation-based currency grouping, lower aggregate foreign-exchange trading costs relative to status quo USD vehicle-currency routing, quanti

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-01 16:38 UTC pith:KIPCN22U

load-bearing objection The theory is the real contribution and it holds up; the 13% savings figure is model-implied and needs a robustness check on the zero-mean return assumption before it is quoted. the 4 major comments →

arxiv 2607.26405 v1 pith:KIPCN22U submitted 2026-07-29 q-fin.TR

Multi-Currency AMMs for Decentralized FOREX Markets: Feasibility & Optimal Design

Reina Ke Xin Li , Andreas Park , Andreas Veneris , Srisht Fateh Singh This is my paper
classification q-fin.TR
keywords automated market makersconstant-mean invariantmulti-currency poolsimpermanent lossvehicle currencyforex trading costshierarchical agglomerative clusteringdecentralized finance
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Most currency pairs do not trade directly; they route through the US dollar, paying two spreads. The paper asks whether a multi-currency automated market maker—one pool holding many currencies at once—can do better. It derives a closed-form equilibrium trading cost for a constant-mean AMM, shows how the cost is minimized by choosing pool weights that balance liquidity depth against impermanent loss, and proves the optimized pool beats two-step dollar routing for a wide range of market parameters. It then partitions currencies into pools by correlating returns, and reports that on 43 non-pegged currencies over 2008–2023 the resulting architecture cuts realized aggregate costs by about 13%, roughly USD 2.5 billion, compared with routing everything through the dollar. The gains come from grouping currencies whose exchange rates move together, so shared liquidity does not generate outsized impermanent loss.

Core claim

The central claim is that the equilibrium aggregate trading cost of a constant-mean multi-currency AMM is c* = sqrt(Δ · H_w · Σ_i E[Q_i]/w_i), where H_w = Σ_{i<j} w_i w_j σ²_ij measures impermanent loss from the joint variance of currency returns, E[Q_i] is expected volume in currency i, and w_i are pool weights. The cost splits into price impact (which shrinks with pooled depth) and impermanent loss (which grows with volatility-weighted exposure); the break-even fee condition makes the two trade off. The paper shows the cost-minimizing weights tilt toward high-volume, low-relative-volatility currencies, with a closed-form optimal weight formula in symmetric environments and a square-root ap

What carries the argument

The load-bearing object is the constant-mean invariant, a weighted-geometric-mean generalization of the two-asset constant-product AMM. Equilibrium is defined by a break-even condition for competitive risk-neutral liquidity providers: fee revenue on noise volume equals expected impermanent loss. From that condition the paper derives the equilibrium cost formula c* = sqrt(Δ H_w Σ_i E[Q_i]/w_i), where H_w = Σ_{i<j} w_i w_j σ²_ij; this identity decomposes the cost into a volatility-risk term and a volume-weighted price-impact term, and it is what makes weight optimization meaningful. The partitioning stage uses hierarchical agglomerative clustering on distances d_ij = sqrt(0.5(1 − ρ_ij)) so tha

Load-bearing premise

The closed-form costs, including the 13% saving, assume exchange-rate excess returns are small and zero-mean (a second-order Taylor approximation) and trust that bilateral goods exports faithfully proxy real FX trading volumes; if the approximation error or the volume-proxy bias affects the pooled architecture differently from the bilateral-routing benchmark, the headline saving is unreliable.

What would settle it

Recompute the out-of-sample realized costs using the same 2008–2023 exchange-rate data but replacing the goods-export volume proxy with observed FX turnover (e.g., BIS Triennial Survey volumes) and compare the HAC-partitioned architecture against USD routing; if the aggregate saving no longer reaches roughly 13% (or becomes negative), the central dominance claim fails. Similarly, any currency pair whose pooled cost exceeds its status quo cost under the optimal weights would violate the paper's dominance result.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • If the equilibrium-cost formula is correct, AMM designers can set fees and weights in closed form—f* = sqrt(Δ H_w / (4 E[Q]²) Σ_i E[Q_i]/w_i)—and the optimized pool dominates two-leg USD routing whenever the cross-pair relative volatility stays below a derived threshold.
  • Pooling currencies with highly correlated returns is what makes multi-currency pools viable; the empirical clusters (European economies, East Asian trade partners, carry-trade targets) confirm the co-movement mechanism rather than assuming it.
  • The HAC partitioning runs in 1.6 seconds, making system-level pool design cheap enough for on-chain implementation or frequent recalibration.
  • The ~13% aggregate saving (USD ~2.5 billion) is out-of-sample over 2008–2023 and persists through the 2008 financial crisis and 2020 COVID recession, suggesting the cost advantage is not a calibration-window artifact.
  • The equal-weighted pool and dedicated bilateral pools are recovered as special cases of the general framework, so the cost comparison is internally consistent.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Because the empirical volume matrix is proxied by bilateral goods exports, the 13% figure is a first-pass estimate; replacing it with actual FX turnover (e.g., BIS Triennial Survey data) is a direct test that could move the saving up or down.
  • The weight rule w_i ∝ sqrt(E[Q_i]/H_i) resembles a square-root inverse-variance weighting; the same cost decomposition could extend to other multi-asset invariants or to dynamic pool rebalancing, which the paper lists as future work.
  • The paper keeps USD in every pool as the shared numeraire, so its 13% gain is a saving relative to dollar routing, not a de-dollarization result; removing the USD hub entirely is a different design the framework could be pushed toward.
  • If adopted at scale, such pools could reduce structural demand for USD as a vehicle currency, but the demonstrated gains are cost savings within the existing USD-centered system, not evidence of independence from it.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper studies whether multi-currency constant-mean AMMs can reduce foreign-exchange trading costs relative to the status quo of routing non-USD pairs through a USD vehicle. It derives closed-form equilibrium trading costs (Prop. 5) from a break-even condition for competitive liquidity provision, characterizes optimal pool weights for symmetric parametrizations (Props. 6–7), and proposes a square-root weight heuristic (Prop. 8). It then frames the system-level pool-membership problem and solves it with hierarchical agglomerative clustering on correlation distances. Using monthly exchange rates and bilateral export volumes for 43 non-pegged currencies over 2008–2023, the paper reports that the HAC-based pooling architecture reduces realized aggregate costs by about 13% (USD 2.5 billion) relative to USD vehicle routing.

Significance. If the cost model is accepted, the paper makes a meaningful contribution: it generalizes a standard AMM equilibrium framework to multi-asset constant-mean pools, gives closed-form expressions and a fast weight heuristic, and contributes a tractable clustering formulation for pool design. The derivations are detailed and, where checkable (e.g., Prop. 6 reducing to Prop. 7 at N=2, equal-weight limits in Prop. 2), internally consistent. The numerical validation of Prop. 8 and the fast runtime of HAC are strengths. The headline empirical result is policy-relevant but rests on the cost model's approximations and on a proxy for trading volume; as the paper itself notes, several assumptions remain to be relaxed.

major comments (4)
  1. [Lemma 4 / Prop. 5 / footnote 4] The zero-mean assumption E[ε_i]=0 is load-bearing and the omitted mean-return differential is first-order, not a higher-order effect. With μ_i=E[ε_i], the second-order expansion gives E[|IPL|] ≈ ½ Σ_{i<j} w_iw_j[σ²_ij+(μ_i−μ_j)²]. The paper's H_w drops the (μ_i−μ_j)² terms, so c* in Eq. (4) understates the true equilibrium cost whenever average returns differ across pool members. For 2008–2023 data containing TRY and other high-depreciation currencies, μ² is a material fraction of monthly variance. Because HAC clusters on correlations rather than mean returns, the bias need not cancel in the c(M) vs c_SQ comparison, so the 13%/USD 2.5B result in §8.4/Table 3 is not robust. This is a correction to the central cost formula, not the 'richer exchange-rate dynamics' future work listed in §10. Please re-estimate with H_w^μ = Σ w_iw_j[σ²_ij+(μ_i−μ_j)²] or demonstrate the term is negligible on t
  2. [§8.3 / Table 3] The empirical volume matrix is proxied by bilateral goods exports, not actual FX turnover. FX volumes are dominated by financial flows, and the proxy may misrepresent the Q_ij that drive both the HAC partition and the cost comparison. The paper offers no sensitivity analysis or comparison with BIS turnover data, despite citing the BIS triennial survey elsewhere. Since Table 3's dollar savings are directly proportional to these volumes, the headline figure needs at least a robustness check using an alternative volume matrix (e.g., BIS-reported turnover for major pairs) or a clear argument that the relative Q_ij across architectures is insensitive to the proxy.
  3. [§8.4 / Appendix C] The 13% savings are model-implied, not measured from actual market prices. The status-quo cost c_SQ and the HAC architecture cost c(M) are both evaluated with the same second-order formulas (Prop. 1 and Prop. 5). While this is internally consistent, the abstract and introduction present the savings as a realized reduction without noting that the comparison is entirely within the model. A calibration check against observed FX spreads for major pairs, or at least a clear caveat in the abstract, is needed before the USD 2.5B figure can be taken at face value.
  4. [§8.3 / §8.4] The text does not clearly state the sample on which the HAC threshold τ is selected. Section 8.3 says the HAC uses 2002–2007 data, but Section 8.1–8.4 describe a rolling-window 2008–2023 evaluation. If τ is chosen by minimizing c(M) over 2008–2023, the 'out-of-sample' claim in Table 3 is compromised. Please specify the threshold-selection sample explicitly and, if it is the evaluation period, redo the analysis with τ chosen only on 2002–2007 data.
minor comments (6)
  1. [Footnote 4] The phrase 'zero-mean excess returns ϵ_i = x_i − 1 ≈ 0' conflates two distinct assumptions: that returns are small and that they have zero mean. These have different implications for the Taylor expansion; please separate them.
  2. [§4] c_SQ, c_BP, and c_SYM are introduced as 'unit costs' but the formulas are aggregate costs over volume. Consistent terminology (e.g., 'aggregate unit cost' or 'total cost') would avoid confusion, especially since these are later used as aggregate costs in Props. 1 and 2.
  3. [Eq. (4)] In Proposition 5, c* is the aggregate equilibrium cost, but the notation could be confused with a unit cost. Please define explicitly as aggregate cost and note the dependence on the total volume E[Q].
  4. [Table 3] The table reports only aggregate costs. Since the paper discusses pool-level contributions (Pools 1–4), showing the cost per pool and for the unpooled/O set would make the decomposition transparent.
  5. [Figures 5 and 6] Figure 5 is described only in passing; please add an explicit reference in the text and explain what the quarterly ratio measures. Figure 6's heatmap order should be described in the caption so the diagonal blocks are interpretable without the main text.
  6. [Appendix D] The greedy algorithm is described as taking over 4 hours and giving 12% savings versus HAC's 13%, but the runtime comparison is informal. A brief note on implementation details (e.g., number of combinations scored) would strengthen the benchmark.

Circularity Check

1 steps flagged

Status-quo benchmark cost is imported from the authors' own unpublished [34]; empirical savings are model-against-model, though the multi-currency cost derivation itself is internally derived.

specific steps
  1. self citation load bearing [Section 4 'Benchmark Currency Exchange Architectures', footnote 4; also Section 2.3 Eq. (3)]
    "Following the work in [34], we consider three arrangements. ... By [34], the fee-optimized equilibrium unit costs for trades between i↔j of size ∆ in numeraire under the three arrangements are, respectively, ... [34] derives the costs in expectation using second-order approximations for IPL under the assumption of small, zero-mean excess returns ϵ_i = x_i − 1 ≈ 0. We use the same simplifying assumption throughout."

    The benchmark costs c_SQ and c_SYM—the status-quo arm of every optimality comparison and of the reported 13% savings—are asserted 'By [34]', an unpublished manuscript by co-author Andreas Park. Proposition 1's proof does not re-derive the per-pair cost formulas; it only sums the pairwise costs imported from [34]. The equilibrium break-even condition (3) used in Proposition 5 is likewise taken from [35], another Malinova-Park paper. Thus the comparison 'optimized multi-currency pool beats vehicle-currency routing' rests on a self-citation chain for the very formulas that define the status quo. The multi-currency Lemmas 3-4 and Proposition 5 are derived within the paper, so this is load-bearing self-citation rather than a purely definitional reduction.

full rationale

The main multi-currency cost machinery is internally derived: Lemma 3 derives price impact from the constant-mean invariant, Lemma 4 derives expected impermanent loss via a second-order Taylor expansion, and Proposition 5 follows algebraically from the break-even condition. The weight optimization and HAC partitioning are solved from this same model, and the empirical partition is selected on 2002-2007 data and evaluated out-of-sample on 2008-2023, so the 13% realized saving is not forced purely by in-sample minimization. The notable circularity-like feature is that the status-quo benchmark is imported from the authors' own unpublished [34], and both the HAC architecture and the status-quo benchmark are priced with the same model formulas, making the USD 2.5B saving a model-against-model comparison. This is best described as load-bearing self-citation plus model reliance, not a definitional equality or a fitted-parameter-renamed-as-prediction. The central derivation of c* is independent, so the score is moderate rather than high.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

The central claim rests on assumptions imported from prior work and on model-internal choices: the break-even liquidity-provision condition (Eq. 3) from [35]; the small-return Taylor approximations in all closed forms; goods exports as a proxy for FX volume; and the constant-mean invariant as the design space. There are no invented entities (no new forces, particles, or conserved quantities). Two hand-set/fitted quantities matter: trade size Δ = 1M USD and the clustering threshold τ = 0.46 selected by cost minimization.

free parameters (2)
  • Trade size Δ = 1 (USD million)
    Fixed by hand (§8.3); all cost magnitudes and the savings estimate scale linearly with this choice.
  • HAC distance threshold τ = 0.46
    Selected by minimizing the model's expected aggregate cost c(M) over τ ∈ [0,1] (§8.2); a data-fitted hyperparameter that determines which currencies are pooled.
axioms (4)
  • domain assumption Risk-neutral, competitive liquidity providers break even: E[|IPL|]·V(0) = E[Q]·f (Eq. 3).
    Imported from [35] (a co-author's prior work); the entire cost model and the 13% savings estimate rest on this equilibrium characterization of pool depth.
  • domain assumption Small, zero-mean excess returns: x_i = 1 + ϵ_i, ϵ_i ≈ 0, E[ϵ_i] = 0 (footnote 4; Lemmas 3-4).
    Every closed form for price impact, impermanent loss, and equilibrium cost is a second-order Taylor approximation; applied to 2008-2023 data with extreme returns without a validity check.
  • domain assumption Bilateral goods-export flows proxy FX trading volumes (E[Q_ij]).
    Volume matrices in §8.3 are built from IMF goods-export data; the savings and especially the USD 2.5 billion figure inherit the proxy's bias and scale.
  • domain assumption Constant-mean (Balancer) invariant with a fixed trade size is the design space.
    Results are specific to this invariant and to fixed-size trades; the authors list stochastic trade sizes and other invariants as future work (§10).

pith-pipeline@v1.3.0-daily-deepseek · 21699 in / 21916 out tokens · 186002 ms · 2026-08-01T16:38:35.174895+00:00 · methodology

0 comments
read the original abstract

Most currency pairs lack a direct liquid market, so international foreign exchange relies on routing transactions through a dominant vehicle currency. Multi-currency automated market makers (AMMs) offer an alternative by sharing liquidity across many currency pairs, facilitating direct cross-currency trade while exploiting liquidity consolidation. This paper studies a multi-currency pool design that minimizes trading cost. Under a constant-mean AMM architecture, equilibrium trading costs reflect the trade-off between reduced price impact from consolidated liquidity and increased impermanent loss from joint return risk. This work derives closed-form costs, characterizes optimal pool weights, and shows that the optimized multi-currency pool dominates the status quo over a range of market parameters. It then formulates the system-level problem of partitioning currencies into multi-currency pools, which is solved using a hierarchical agglomerative clustering algorithm. Empirically, using exchange rate and trade data for 43 currencies over 2008-2023, the algorithm runs in 1.6 seconds and produces pools with geographic and economic structure. Notably, this reduces realized costs by ~13% relative to the status quo of vehicle-currency routing, with gains stable through episodes of global financial stress.

Figures

Figures reproduced from arXiv: 2607.26405 by Andreas Park, Andreas Veneris, Reina Ke Xin Li, Srisht Fateh Singh.

Figure 1
Figure 1. Figure 1: s vs. v regions for optimality over the status quo under the three-currency (v, s) parametrization [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 4
Figure 4. Figure 4: Coefficient of variation (standard devi [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Quarterly realized cost of Table 2 archi [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Threshold limits as v → ∞ for different values of N. C System Cost Under Currency Partition Consider a pool I = G ∪ {0}. Since i ∈ G trades all of its volume through the pool I (either directly to another pool member, or to an external currency routed through currency 0), the effective pool volumes are QI i = Qi for all i ∈ G and QI 0 = P i∈G(Q0i + P j̸∈I Qij ), the sum of the trades between 0 and pool mem… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

56 extracted references · 6 canonical work pages

  1. [1]

    Uniswap v2 Core

    Hayden Adams, Noah Zinsmeister, and Dan Robinson. Uniswap v2 Core. Technical report, Uniswap, Mar 2020. URL:https://uniswap.org/whitepaper.pdf

    2020

  2. [2]

    Uniswap v3 Core

    Hayden Adams, Noah Zinsmeister, Moody Salem, River Keefer, and Dan Robinson. Uniswap v3 Core. Technical report, Uniswap, Mar 2021. URL:https://uniswap.org/whitepaper-v3.pdf

    2021

  3. [3]

    Improved price oracles: Constant function market makers

    Guillermo Angeris and Tarun Chitra. Improved price oracles: Constant function market makers. InProceedings of the 2nd ACM Conference on Advances in Financial Technologies, AFT ’20, page 80–91, New York, NY, USA, 2020. Association for Computing Machinery.doi:10.1145/3419614. 3423251

    doi:10.1145/3419614 2020

  4. [4]

    Replicating market makers.Digital Finance, 5(2):367–387, jun 2023.doi:10.1007/s42521-023-00082-0

    Guillermo Angeris, Alex Evans, and Tarun Chitra. Replicating market makers.Digital Finance, 5(2):367–387, jun 2023.doi:10.1007/s42521-023-00082-0

    work page doi:10.1007/s42521-023-00082-0 2023

  5. [5]

    Optimal routing for constant function market makers

    Guillermo Angeris, Alex Evans, Tarun Chitra, and Stephen Boyd. Optimal routing for constant function market makers. InProceedings of the 23rd ACM Conference on Economics and Compu- tation, EC ’22, page 115–128, New York, NY, USA, 2022. Association for Computing Machinery. doi:10.1145/3490486.3538336

    arXiv 2022

  6. [6]

    Liquidity provision by automated market makers, Oct 2020

    Jun Aoyagi. Liquidity provision by automated market makers, Oct 2020. URL:https://papers.s srn.com/sol3/papers.cfm?abstract_id=3674178

    2020

  7. [7]

    Triennial central bank survey: Report on global foreign exchange market activity in 2010

    Bank for International Settlements. Triennial central bank survey: Report on global foreign exchange market activity in 2010. Technical report, Bank for International Settlements, Basel, Switzerland, Dec 2010. URL:https://bis.org

    2010

  8. [8]

    Project mariana: Cross-border exchange of wholesale cbdcs using automated market-makers

    Bank for International Settlements. Project mariana: Cross-border exchange of wholesale cbdcs using automated market-makers. Technical report, BIS Innovation Hub, Sep 2023. URL:https: //www.bis.org/publ/othp75.pdf

    2023

  9. [9]

    Triennial central bank survey: Otc foreign exchange turnover in april 2025

    Bank for International Settlements. Triennial central bank survey: Otc foreign exchange turnover in april 2025. Technical report, Bank for International Settlements, Basel, Switzerland, Sep 2025. URL:https://bis.org

    2025

  10. [10]

    Dollar dominance: A source of dollar volatility?, May

    Cara Bordier, Lukas Frei, and Simon Stalder. Dollar dominance: A source of dollar volatility?, May

  11. [11]

    The adoption of blockchain-based decentralized exchanges

    Agostino Capponi and Ruizhe Jia. The adoption of blockchain-based decentralized exchanges. Papers 2103.08842, arXiv.org, Mar 2021. URL:https://ideas.repec.org/p/arx/papers/2 103.08842.html,doi:None

    Pith/arXiv arXiv 2021

  12. [12]

    China, Saudi Arabia sign currency swap agreement.Reuters, November 2023

    Joe Cash. China, Saudi Arabia sign currency swap agreement.Reuters, November 2023. URL: https://www.reuters.com/markets/currencies/china-saudi-arabia-central-banks-sig n-local-currency-swap-agreement-2023-11-20/

    2023

  13. [13]

    Arbitrage, factor structure, and mean-variance analysis on large asset markets.Econometrica, 51(5):1281–1304, 1983

    Gary Chamberlain and Michael Rothschild. Arbitrage, factor structure, and mean-variance analysis on large asset markets.Econometrica, 51(5):1281–1304, 1983. URL:http://www.jstor.org/stab le/1912275

    arXiv 1983

  14. [14]

    Commodity currencies.Journal of International Economics, 60(1):133–160, 2003

    Yu chin Chen and Kenneth Rogoff. Commodity currencies.Journal of International Economics, 60(1):133–160, 2003. Emperical Exchange Rate Models. URL:https://www.sciencedirect.com/ science/article/pii/S0022199602000727,doi:10.1016/S0022-1996(02)00072-7

    work page doi:10.1016/s0022-1996(02)00072-7 2003

  15. [15]

    UAE, Ethiopia sign currency swap agreement with value of up to$817 million

    Jana Choukeir. UAE, Ethiopia sign currency swap agreement with value of up to$817 million. Reuters, July 2024. URL:https://www.reuters.com/business/finance/uae-ethiopia-sign-c urrency-swap-agreement-with-value-up-817-million-2024-07-16/

    2024

  16. [16]

    Looking at the other side of carry trades: Are there any safe haven currencies? Working Papers 2014-03, CEPII research center, Feb

    Virginie Coudert, Cyriac Guillaumin, and H´ el` ene Raymond. Looking at the other side of carry trades: Are there any safe haven currencies? Working Papers 2014-03, CEPII research center, Feb

    2014

  17. [17]

    Crama and M

    Y. Crama and M. Schyns. Simulated annealing for complex portfolio selection problems.European Journal of Operational Research, 150(3):546–571, 2003. Financial Modelling. URL:https://www. sciencedirect.com/science/article/pii/S0377221702007841,doi:10.1016/S0377-2217(02 )00784-1

    work page doi:10.1016/s0377-2217(02 2003

  18. [18]

    Metaheuristics for the portfolio selection problem.International Journal of Operations Research, 5(1):13–35, 2008

    Giacomo Di Tollo and Andrea Roli. Metaheuristics for the portfolio selection problem.International Journal of Operations Research, 5(1):13–35, 2008. URL:https://www.orstw.org.tw/ijor/vol5 no1/2_IJOR_Vol5_No1_pp13_35.pdf

    2008

  19. [19]

    The axis of evasion: Behind China’s oil trade with Iran and Russia.Atlantic Council, March 2024

    Kimberly Donovan and Maia Nikoladze. The axis of evasion: Behind China’s oil trade with Iran and Russia.Atlantic Council, March 2024. URL:https://www.atlanticcouncil.org/blogs/ne w-atlanticist/the-axis-of-evasion-behind-chinas-oil-trade-with-iran-and-russia/

    2024

  20. [20]

    StableSwap—Efficient Mechanism for Stablecoin Liquidity

    Michael Egorov. StableSwap—Efficient Mechanism for Stablecoin Liquidity. Technical report, Curve Finance, 2019. URL:https://curve.fi/files/stableswap-paper.pdf

    2019

  21. [21]

    Optimal fees for geometric mean market makers,

    Alex Evans, Guillermo Angeris, and Tarun Chitra. Optimal fees for geometric mean market makers,

  22. [22]

    What and where are russia’s$300 billion in reserves frozen in the west?, Dec 2023

    Elena Fabrichnaya and Guy Faulconbridge. What and where are russia’s$300 billion in reserves frozen in the west?, Dec 2023. URL:https://www.reuters.com/world/europe/what-where-are -russias-300-billion-reserves-frozen-west-2023-12-28/

    2023

  23. [23]

    Brics pay as a challenge to swift network, Dec 2024

    Evan Freidin. Brics pay as a challenge to swift network, Dec 2024. URL:https://www.lowyinst itute.org/the-interpreter/brics-pay-challenge-swift-network

    2024

  24. [24]

    Cross-border exchange of cbdcs using layer-2 blockchain, 2024

    Krzysztof Gogol, Johnnatan Messias, Malte Schlosser, Benjamin Kraner, and Claudio Tessone. Cross-border exchange of cbdcs using layer-2 blockchain, 2024. URL:https://arxiv.org/abs/23 12.16193,arXiv:2312.16193

    Pith/arXiv arXiv 2024

  25. [25]

    Goldberg and C´ edric Tille

    Linda S. Goldberg and C´ edric Tille. Vehicle currency use in international trade.Journal of Inter- national Economics, 76(2):177–192, 2008. URL:https://www.sciencedirect.com/science/arti cle/pii/S0022199608000664,doi:10.1016/j.jinteco.2008.07.001

    work page doi:10.1016/j.jinteco.2008.07.001 2008

  26. [26]

    Behavior of liquidity providers in decentralized exchanges

    Lioba Heimbach, Ye Wang, and Roger Wattenhofer. Behavior of liquidity providers in decentralized exchanges. Papers 2105.13822, arXiv.org, May 2021. URL:https://ideas.repec.org/p/arx/pa pers/2105.13822.html,doi:None

    Pith/arXiv arXiv 2021

  27. [27]

    Annual report on exchange arrangements and exchange restrictions (areaer), 2024

    International Monetary Fund. Annual report on exchange arrangements and exchange restrictions (areaer), 2024. URL:https://www.elibrary-areaer.imf.org/

    2024

  28. [28]

    Exchange rates (er).https://data.imf.org/en/datasets/IMF.S TA:ER, 2026

    International Monetary Fund. Exchange rates (er).https://data.imf.org/en/datasets/IMF.S TA:ER, 2026. Accessed: 2026-05-18

    2026

  29. [29]

    International trade in goods (by partner country) (imts).https: //data.imf.org/en/datasets/IMF.STA:IMTS, 2026

    International Monetary Fund. International trade in goods (by partner country) (imts).https: //data.imf.org/en/datasets/IMF.STA:IMTS, 2026. Accessed: 2026-05-18

    2026

  30. [30]

    Kraft.A Software Package for Sequential Quadratic Programming

    D. Kraft.A Software Package for Sequential Quadratic Programming. Deutsche Forschungs- und Versuchsanstalt f¨ ur Luft- und Raumfahrt K¨ oln: Forschungsbericht. Wiss. Berichtswesen d. DFVLR,

  31. [31]

    The international role of the dollar: Theory and policy implications

    Paul Krugman. The international role of the dollar: Theory and policy implications. In John F. O. Bilson and Richard C. Marston, editors,Exchange Rate Theory and Practice, pages 261–278. University of Chicago Press, 1984

    1984

  32. [32]

    Decentralized exchange: The uniswap automated market maker

    Alfred Lehar and Christine Parlour. Decentralized exchange: The uniswap automated market maker. Journal of Finance, 80(1):321–374, February 2025. URL:https://ideas.repec.org/a/bla/jfin an/v80y2025i1p321-374.html,doi:10.1111/jofi.13405

    work page doi:10.1111/jofi.13405 2025

  33. [33]

    Automated market-making for fiat currencies, 2021

    Alex Lipton and Artur Sepp. Automated market-making for fiat currencies, 2021. URL:https: //arxiv.org/abs/2109.12196,arXiv:2109.12196

    Pith/arXiv arXiv 2021

  34. [34]

    Dethroning the dollar? multi-currency automated market making for fx

    Katya Malinova and Andreas Park. Dethroning the dollar? multi-currency automated market making for fx. Unpublished. 14

  35. [35]

    Learning from defi: Would automated market makers improve equity trading?, Aug 2023

    Katya Malinova and Andreas Park. Learning from defi: Would automated market makers improve equity trading?, Aug 2023. URL:https://papers.ssrn.com/sol3/papers.cfm?abstract_id=45 31670

    2023

  36. [36]

    Manning, Prabhakar Raghavan, and Hinrich Sch¨ utze.Introduction to Information Retrieval

    Christopher D. Manning, Prabhakar Raghavan, and Hinrich Sch¨ utze.Introduction to Information Retrieval. Cambridge University Press, USA, 2008

    2008

  37. [37]

    Resurgent Political Risk Derails Rallies Across Emerging Markets.Bloomberg Markets, May 2026

    Maria Elena Vizcaino, Zijia Song, and Nicolle Yapur. Resurgent Political Risk Derails Rallies Across Emerging Markets.Bloomberg Markets, May 2026. URL:https://www.bloomberg.com/news/art icles/2026-05-24/resurgent-political-risk-derails-rallies-across-emerging-marke ts

    2026

  38. [38]

    Portfolio selection.The Journal of Finance, 7(1):77–91, 1952

    Harry Markowitz. Portfolio selection.The Journal of Finance, 7(1):77–91, 1952. URL:http: //www.jstor.org/stable/2975974

    arXiv 1952

  39. [39]

    Russia, China find payments workaround as US sanctions net widens, sources say.Reuters, June 2024

    Alexander Marrow. Russia, China find payments workaround as US sanctions net widens, sources say.Reuters, June 2024. URL:https://www.reuters.com/world/russia-china-find-payment s-workaround-us-sanctions-net-widens-sources-say-2024-06-20/

    2024

  40. [40]

    Balancer: A Non-Custodial Portfolio Manager, Liq- uidity Provider, and Price Sensor

    Fernando Martinelli and Nikolai Mushegian. Balancer: A Non-Custodial Portfolio Manager, Liq- uidity Provider, and Price Sensor. Technical report, Balancer Labs, Sep 2019. URL:https: //balancer.fi/whitepaper.pdf

    2019

  41. [41]

    Moallemi, and Tim Roughgarden

    Jason Milionis, Ciamac C. Moallemi, and Tim Roughgarden. Automated market making and ar- bitrage profits in the presence of fees. Papers 2305.14604, arXiv.org, May 2023. URL:https: //ideas.repec.org/p/arx/papers/2305.14604.html,doi:None

    Pith/arXiv arXiv 2023

  42. [42]

    Moallemi, Tim Roughgarden, and Anthony Lee Zhang

    Jason Milionis, Ciamac C. Moallemi, Tim Roughgarden, and Anthony Lee Zhang. Automated market making and loss-versus-rebalancing. Papers 2208.06046, arXiv.org, Aug 2022. URL:https: //ideas.repec.org/p/arx/papers/2208.06046.html,doi:None

    Pith/arXiv arXiv 2022

  43. [43]

    Chair’s Statement and Outcome Document at BRICS Foreign Ministers’ Meeting, May 2026

    Government of India Ministry of External Affairs. Chair’s Statement and Outcome Document at BRICS Foreign Ministers’ Meeting, May 2026. URL:https://mea.gov.in/bilateral-documen ts.htm?dtl/41144

    2026

  44. [44]

    Uniswap: A good deal for liquidity providers?, Jan 2019

    Pintail. Uniswap: A good deal for liquidity providers?, Jan 2019. URL:https://pintail.medium .com/uniswap-a-good-deal-for-liquidity-providers-104c0b6816f2

    2019

  45. [45]

    Time-variant safe haven currencies.International Review of Economics & Finance, 93:316–328, 2024

    Ayano Sato, Hayato Nakata, and Jay Percy. Time-variant safe haven currencies.International Review of Economics & Finance, 93:316–328, 2024. URL:https://www.sciencedirect.com/sc ience/article/pii/S1059056024002685,doi:10.1016/j.iref.2024.04.015

    work page doi:10.1016/j.iref.2024.04.015 2024

  46. [46]

    Loaded With Massive Reserves Of Rupees In Indian Banks, Russia Buys ”Sensitive” Goods From New Delhi Despite Embargo - Reports, September 2024

    Ritu Sharma. Loaded With Massive Reserves Of Rupees In Indian Banks, Russia Buys ”Sensitive” Goods From New Delhi Despite Embargo - Reports, September 2024. URL:https://www.eurasi antimes.com/ne-loaded-with-massive-reserves-of-rupees/

    2024

  47. [47]

    Dollar dominance in fx trading.Management Science, Apr 2026.doi:10.128 7/mnsc.2024.07656

    Fabricius Somogyi. Dollar dominance in fx trading.Management Science, Apr 2026.doi:10.128 7/mnsc.2024.07656

    arXiv 2026

  48. [48]

    Central bank currency swaps tracker, Apr 2026

    Benn Steil, Yuma Schuster, and Samuel Zucker. Central bank currency swaps tracker, Apr 2026. URL:https://www.cfr.org/articles/central-bank-currency-swaps-tracker

    2026

  49. [49]

    UN Comtrade Database, 2025

    United Nations Statistics Division. UN Comtrade Database, 2025. Accessed: 2026-05-19. URL: https://comtradeplus.un.org/

    2025

  50. [50]

    Saudi Arabia ‘open’ to petroyuan, closer China ties, minister says.South China Morning Post, sep 2024

    Kandy Wong. Saudi Arabia ‘open’ to petroyuan, closer China ties, minister says.South China Morning Post, sep 2024. URL:https://www.scmp.com/economy/global-economy/article/327 7788/saudi-arabia-open-petroyuan-closer-china-ties-minister-says

    2024

  51. [51]

    Decentralised finance and automated market making: Predictable loss and optimal liquidity provision, 2024

    ´Alvaro Cartea, Fay¸ cal Drissi, and Marcello Monga. Decentralised finance and automated market making: Predictable loss and optimal liquidity provision, 2024. URL:https://arxiv.org/abs/23 09.08431,arXiv:2309.08431

    Pith/arXiv arXiv 2024

  52. [52]

    Portfolio optimization with dis- crete simulated annealing, 2022

    ´Alvaro Rubio-Garc ´ ıa, Juan Jos´ e Garc ´ ıa-Ripoll, and Diego Porras. Portfolio optimization with dis- crete simulated annealing, 2022. URL:https://arxiv.org/abs/2210.00807,arXiv:2210.00807. 15 A Proofs A.1 Proofs of benchmark arrangement trading costs Proposition 1(Status quo trading cost).The expected aggregate trading cost in the status quo forN+1 c...

    Pith/arXiv arXiv 2022

  53. [1988]

    URL:https://books.google.ca/books?id=4rKaGwAACAAJ

  54. [2014]

    URL:https://ideas.repec.org/p/cii/cepidt/2014-03.html,doi:None. 13

    2014

  55. [2021]

    URL:https://arxiv.org/abs/2104.00446,arXiv:2104.00446

    Pith/arXiv arXiv

  56. [2026]

    URL:https://ssrn.com/abstract=6757018