skip to main content
10.1145/3523181.3523198acmotherconferencesArticle/Chapter ViewAbstractPublication PagesasseConference Proceedingsconference-collections
research-article

The Composable Finance Mosaic Bridge and Liquidity Forecasting

Published: 18 April 2022 Publication History

Abstract

We introduce the forecasting technology behind Composable Finance's (Composable's)1 cross-layer cross-chain bridge Mosaic2 . Mosaic contains a network of vaults, one on each layer and chain, and we present our research into improving the operation and robustness of said bridge via liquidity forecasting of its associated vaults. Users utilize the bridge to transfer assets from a source network and source layer (a source vault) to a destination network and layer (a destination vault). However, if a vault is near depletion when a transfer is initiated the transfer can fail due to inefficient liquidity. Thus, critical to maximizing the user experience of Mosaic is being able to forecast accurately when the liquidity will be depleted from vaults supported by Mosaic. After introducing Mosaic, and without loss of generality, we forecast liquidity movements between three vaults: one on the layer 1 (L1) Ethereum mainnet, one on the L2 Polygon (POL) network, and one on the L2 Arbitrum (ARB) network. We demonstrate state of the art results from our statistical and AI methods that can consistently achieve low average errors (e.g., <2% liquidity) and high coverage (e.g., >80%) of the true liquidity levels in their 1-week-ahead forecast confidence intervals. As a baseline, we first develop autoregressive integrated moving average (ARIMA) models for forecasting. The ARIMA models are optimized at each time step and provide conservative confidence intervals that allow us to safely assess the times when vaults will reach critical liquidity values and therefore, prevent depletion. To further enhance our forecasting framework, we also develop a machine learning approach with Gaussian Processes that can further capture time series trends, past replenishment events, where large discontinuities are present. Liquidity data for our forecasting methodologies will be provided in two levels of fidelity. We first introduce our in-house Liquidity Simulation Environment (LSE) - a state machine capable of running hypothetical vault liquidity scenarios under certain assumed cross-layer trades/swaps. Next, we forecast real-world data obtained from a recent PoC three-vault bridge setup run by Composable, and we apply the LSE to that. Finally, we provide our thoughts on forecasting beyond the PoC - our work will be of interest to layer 1, layer 2, cross-layer cross-chain, and AI researchers working on DeFi and blockchain technologies.

References

[1]
B. Øksendal, Stochastic Differential Equations, Springer, Berlin, Heidelberg, 2003.
[2]
R. J. Hyndman and G. Athanasopoulos, Forecasting: principles and practices, Otexts, 2018.
[3]
D. W. Scott, Multivariate density estimation: theory, practice and visualization, John Wiley & Sons, 2015.
[4]
C. C. Holt, "Forecasting seasonals and trends by exponentially weighted moving averages," International Journal of forecasting, vol. 20, no. 1, pp. 5-10, 2004.
[5]
C. K. Williams and C. E. Ramsussen, Gaussian Processes for Machine Learning, MIT Press Cambridge, MA, 2006.
[6]
C. K. Williams, "Computing with Infinite Networks," Advances in neural information processing systems, pp. 295-301, 1997.
[7]
R. M. Neal, "Priors for Infinite Networks," in Bayesian Learning for Neural Networks, 1996.
[8]
A. Damianou and N. D. Lawrence, "Deep Gaussian Processes," in Artificial Intelligence and Statistics, 2013.
[9]
W. Wilkinson, P. Chang, M. Andersen and A. Solin, "State space expectation propagation: Efficient inference schemes for temporal Gaussian Processes," in International Conference on Machine Learning, 2020.
[10]
O. Hamelijnck, W. J. Wilkinson, N. A. Loppi, A. Solin and T. Damoulas, "Spatio-temporal Variational Gaussian Processes," in Thirty-fifth Conference on Neural Information Processing Systems, 2021.
[11]
J. Hartikainen and S. Särkkä, "Kalman Filtering and smoothing solutions to temporal Gaussian Process regression models," in 2010 IEEE International workshop on machine learning for signal processing, 2010.
[12]
M. Titsias, "Variational learning of inducing variables in sparse Gaussian processes," in Artificial Intelligence and Statistics, 2009.
[13]
J. Hensman, A. Matthews and Z. Ghahramani, "Scalable variational Gaussian Process classification," in Artificial Intelligence and Statistics, 2015.
[14]
D. P. Kingma and J. Ba, "Adam: A method for stochastic optimization," arXiv preprint arXiv:1412.6980, 2014.
[15]
J. Durbin and S. J. Koopman, Time series analysis by state space methods, Oxford University Press, 2012.
[16]
Y. Yu, X. Si, C. Hu and J. Zhang, "A review of recurrent neural networks: LSTM cells and network architectures," Neural Computations, vol. 31, no. 7, pp. 1235-1270, 2019.
[17]
R. M. Neal, "Priors for infinite networks," in Bayesian Learning for Neural Networks, 1996.

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Other conferences
ASSE' 22: 2022 3rd Asia Service Sciences and Software Engineering Conference
February 2022
202 pages
ISBN:9781450387453
DOI:10.1145/3523181
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 18 April 2022

Permissions

Request permissions for this article.

Check for updates

Qualifiers

  • Research-article
  • Research
  • Refereed limited

Conference

ASSE' 22

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • 0
    Total Citations
  • 45
    Total Downloads
  • Downloads (Last 12 months)5
  • Downloads (Last 6 weeks)0
Reflects downloads up to 15 Feb 2025

Other Metrics

Citations

View Options

Login options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

HTML Format

View this article in HTML Format.

HTML Format

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media