Detecting Depegs: Safer Passive Liquidity Provision on Curve Finance

Table of Contents

1 Gabatarwa

Binciken ya ba da shaida kan yadda za a iya ƙirƙira ma'auni na ƙididdiga da algorithms na ganowa don sanar da masu ba da ruɗar ruɗar (LPs) cikin sauri game da yuwuwar karkatar da tsayayyen kuɗi da abubuwan da ake samu na lalata. Binciken ya mayar da hankali kan tafkunan StableSwap na Curve Finance kuma ya haɓaka ingantaccen tsarin ganowa.

1.1 Bayan Fage

Stablecoins da kuma abubuwan da ake amfani da su don samun ruwa a cikin hanyoyin saka hannun jari, suna da alaƙa da kuɗaɗen da ke yawo a kasuwa. Stablecoins galibi suna da alaƙa da Dala na Amurka, yayin da abubuwan da ake amfani da su don samun ruwa a cikin hanyoyin saka hannun jari, sukan kasance masu alaƙa da ETH ko wasu kuɗaɗen cibiyar sadarwa. Ƙungiyoyi suna kiyaye hanyoyin fansar kuɗi, amma idan ’yan kasuwa suka rasa imani, farashin na iya raguwa a kasuwannin sakandare - wannan aikin ana kiransa depegging.

2 Methodology

Muna ƙirƙiri tsarin ma'auni da aka tsara don gano yuwuwar raguwar kadarorin bisa farashi da bayanan ciniki daga tafkunan Kuɗi na Curve.

2.1 Depeg Detection Metrics

Tsarin ganowa ya haɗa da ma'auni da yawa ciki har da karkata farashi, ɓatan adadin ciniki, rashin daidaiton tafkin ruwa, da al'amuran saurin canji na tarihi. Ana haɗa waɗannan ma'auni don ƙirƙirar cikakken tsarin tantance haɗari.

2.2 Bayesian Online Changepoint Detection

Muna daidaita algorithm na Bayesian Online Changepoint Detection (BOCD) don faɗakar da LPs game da yuwuwar raguwa. Tsarin BOCD yana sarrafa bayanan rafi kuma yana gano katsewar tsari a cikin bayanan lokaci-lokaci a ainihin lokacin.

3 Experimental Results

An horar da kuma gwada algorithm na gano sauyi a kan farashin LP token na Curve don rafukan StableSwap 13 a cikin 2022 da 2023.

3.1 USDC Depeg Detection

Tsarin mu, wanda aka horar da shi akan bayanan UST na 2022, ya gano raguwar darajar USDC a cikin Maris 2023 da ƙarfe 9 na dare UTC a ranar 10 ga Maris, kimanin sa'o'i 5 kafin USDC ta faɗi ƙasa da centi 99. Wannan ganin farko ya ba da babban gargaɗi ga masu samar da ruwa.

3.2 Performance Evaluation

Tsarin ya nuna ƴan ƙararrakin ƙarya a cikin lokacin gwaji na wata 17, yana nuna ƙwararrun aiki a cikin abubuwan da suka faru na depeg da yawa ciki har da UST, USDC, da stETH depegs.

4 Aiwarta Fasaha

4.1 Tsarin Lissafi

Algorithm na Gano Canjin Canjin Kan Yanar Gizo na Bayesian ya dogara ne akan tsarin lissafi mai zuwa:

Tsawon gudu $r_t$ a lokacin $t$ yana wakiltar lokacin da aka samu canjin canji na ƙarshe. Ana sabunta yuwuwar tsawon gudu a jere:

$P(r_t | x_{1:t}) = \sum_{r_{t-1}} P(r_t | r_{t-1}) P(x_t | r_{t-1}, x_t^{(r)}) P(r_{t-1} | x_{1:t-1})$

Ina $x_t^{(r)}$ yana wakiltar bayanan tun daga canjin canji na ƙarshe, kuma aikin haɗari $H(r_t)$ yana ƙayyade yuwuwar canjin canji:

$P(r_t | r_{t-1}) = \begin{cases} H(r_{t-1} + 1) & \text{if } r_t = 0 \\ 1 - H(r_{t-1} + 1) & \text{if } r_t = r_{t-1} + 1 \\ 0 & \text{otherwise} \end{cases}$

4.2 Aiwarar Code

class BayesianChangepointDetector:

5 Aikace-aikacen Gaba

This research can be extended to dynamically de-risk Curve pools by modifying parameters in anticipation of potential depegs. Future applications include:

• APIs na sarrafa hadarin lokaci-lokaci don tsarin DeFi
• Gyara sigogi na tafki mai sauyi bisa alamun hadari
• Tsarin gano depeg na tsarin-tsare
• Kayayyakin inshora masu samar da ruwa
• Kayan aikin sa ido na ka'idoji masu fitar da stablecoin

6 Bincike na Asali

The research by Cintra and Holloway represents a significant advancement in real-time risk management for decentralized finance. Their application of Bayesian Online Changepoint Detection to stablecoin depeg scenarios demonstrates how sophisticated statistical methods can be adapted for blockchain financial markets. The methodology bears similarity to change point detection techniques used in traditional finance, such as those described in the seminal work by Adams and MacKay (2007) on Bayesian online changepoint detection, but adapted for the unique characteristics of automated market makers.

What makes this approach particularly innovative is its real-time capability and minimal false positive rate. Unlike traditional financial surveillance systems that might rely on simpler threshold-based alerts, the Bayesian framework incorporates uncertainty quantification and sequential updating. This aligns with modern machine learning approaches in anomaly detection, similar to techniques used in cybersecurity and network monitoring. The system's ability to detect the USDC depeg 5 hours in advance is remarkable, considering that most market participants were caught by surprise during the Silicon Valley Bank collapse.

Binciken ya ginu akan ka'idoji na kafaffe daga duka kididdiga na gargajiya da koyon inji na zamani. Tushen ilmin lissafi ya samo asali daga hanyoyin shigar bayanai na Bayesian makamancin waɗanda ake amfani da su a cikin karkatwar tsarin Gaussian da Monte Carlo na bi-de-bi, kamar yadda aka ambata a cikin ayyuka irin su "Pattern Recognition and Machine Learning" na Bishop (2006). Duk da haka, aikace-aikacen ga samar da ruwa na DeFi yana wakiltar gudunmawa mai ban sha'awa. Aikin tsarin a cikin tafkuna daban-daban guda 13 sama da watanni 17 tare da ƙaramin faɗakarwar ƙarya yana nuna ƙwararrun iyawar gama gari.

Idan akwai sauran hanyoyin sarrafa hadarin DeFi, kamar tsarin bayanan farashin da aka yi amfani da su a cikin ka'idojin lamuni ko kuma hanyoyin da ake katsewa a wasu musayar tsakiya, wannan hanyar tana ba da hanya mai zurfi da kuma tsari. Ba wai kawai tana mayar da martani ga motsin farashi ba amma tana gano sauye-sauyen tsari a cikin tsarin halayen kasuwa. Wannan na iya haɗuwa da tsarin daidaita sigogin AMM mai ƙarfi, kamar aikin da Angeris et al. (2021) suka yi akan ingantattun ka'idojin farashi, suna ƙirƙirar cikakkiyar tsarin sarrafa haɗari don musayar da ba ta da tushe.

Aiwatar da aikin yin amfani da API don masu samar da ruwa ya nuna amfanin binciken nan take. Wannan yana rage tazara tsakanin hanyar ilimi da amfanin duniya, yana magance wata muhimmiyar bukata a cikin haɓaka saurin DeFi. Yayin da kasuwannin stablecoin ke ci gaba da girma da kuma fuskantar binciken ka'idoji, irin wadannan tsare-tsaren ganowa za su zama masu muhimmanci ga mahalarta da masu kula da su.

7 References

1. Adams, R.P., & MacKay, D.J.C. (2007). Bayesian Online Changepoint Detection. University of Cambridge.
2. Bishop, C.M. (2006). Pattern Recognition and Machine Learning. Springer.
3. Angeris, G., et al. (2021). Improved Price Oracles: Constant Function Market Makers. Proceedings of the ACM on Measurement and Analysis of Computing Systems.
4. Bolger, M., & Hon, H. (2022). When the Currency Breaks. Llama Risk Research.
5. Egorov, M. (2019). StableSwap - ingantaccen tsari don kudin ruwa na Stablecoin. Curve Finance Whitepaper.
6. Goodfellow, I., et al. (2016). Deep Learning. MIT Press.