Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Forecasting Cryptocurrency Staking Rewards

Sauren Gupta, Apoorva Hathi Katharaki, Yifan Xu, Bhaskar Krishnamachari, Rajarshi Gupta

TL;DR

This paper investigates forecasting staking rewards for PoS cryptocurrencies using two simple time-series methods: a moving-window average and linear regression. It demonstrates ETH staking rewards can be forecast with RMSE relative to the mean as low as $0.7\%$ for 1-day ahead and $1.1\%$ for 7-day ahead horizons using a 7-day window, while performance varies across SOL, XTZ, ATOM, and MATIC. Short-term forecasting favors linear regression for XTZ and ATOM, whereas ETH and MATIC show strong performance with the moving window, though MATIC remains highly volatile. The results suggest staking rewards are relatively slow-moving and predictable for most assets, providing practical guidance for stake timing decisions, albeit with caution for highly volatile assets like MATIC.

Abstract

This research explores a relatively unexplored area of predicting cryptocurrency staking rewards, offering potential insights to researchers and investors. We investigate two predictive methodologies: a) a straightforward sliding-window average, and b) linear regression models predicated on historical data. The findings reveal that ETH staking rewards can be forecasted with an RMSE within 0.7% and 1.1% of the mean value for 1-day and 7-day look-aheads respectively, using a 7-day sliding-window average approach. Additionally, we discern diverse prediction accuracies across various cryptocurrencies, including SOL, XTZ, ATOM, and MATIC. Linear regression is identified as superior to the moving-window average for perdicting in the short term for XTZ and ATOM. The results underscore the generally stable and predictable nature of staking rewards for most assets, with MATIC presenting a noteworthy exception.

Forecasting Cryptocurrency Staking Rewards

TL;DR

This paper investigates forecasting staking rewards for PoS cryptocurrencies using two simple time-series methods: a moving-window average and linear regression. It demonstrates ETH staking rewards can be forecast with RMSE relative to the mean as low as for 1-day ahead and for 7-day ahead horizons using a 7-day window, while performance varies across SOL, XTZ, ATOM, and MATIC. Short-term forecasting favors linear regression for XTZ and ATOM, whereas ETH and MATIC show strong performance with the moving window, though MATIC remains highly volatile. The results suggest staking rewards are relatively slow-moving and predictable for most assets, providing practical guidance for stake timing decisions, albeit with caution for highly volatile assets like MATIC.

Abstract

This research explores a relatively unexplored area of predicting cryptocurrency staking rewards, offering potential insights to researchers and investors. We investigate two predictive methodologies: a) a straightforward sliding-window average, and b) linear regression models predicated on historical data. The findings reveal that ETH staking rewards can be forecasted with an RMSE within 0.7% and 1.1% of the mean value for 1-day and 7-day look-aheads respectively, using a 7-day sliding-window average approach. Additionally, we discern diverse prediction accuracies across various cryptocurrencies, including SOL, XTZ, ATOM, and MATIC. Linear regression is identified as superior to the moving-window average for perdicting in the short term for XTZ and ATOM. The results underscore the generally stable and predictable nature of staking rewards for most assets, with MATIC presenting a noteworthy exception.
Paper Structure (19 sections, 4 figures, 4 tables)

This paper contains 19 sections, 4 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Background: Stacking and rewards
  4. How Staking Works
  5. Determining rewards
  6. Staking Duration
  7. Feature Selection
  8. Considered Features
  9. Exclusion and Unexplored Features
  10. Experimental Setup
  11. Forecasting Algorithms
  12. Objective of the Model
  13. Data Collection and Utilization
  14. Training and Test Data Splitting Strategy
  15. Metrics
  16. ...and 4 more sections

Figures (4)

  • Figure 1: Ethereum (ETH) Data Visualization (2021-06-23 to 2022-08-06) - Showcasing Daily Price, Search Trends, and Reward Rate.
  • Figure 2: Staking Rewards Next Day Predictions of rewards for 5 cryptocurrencies ETH, SOL, MATIC, ATOM and XTZ, using MWA and SLR
  • Figure 3: Staking Rewards Next N-day Predictions of 5 cryptocurrencies ETH, SOL, MATIC, ATOM and XTZ using MWA
  • Figure 4: Staking Rewards Next N-day Predictions of 5 cryptocurrencies ETH, SOL, MATIC, ATOM and XTZ using SLR