Paper: arXiv 2503.08693
Authors: Qi Deng, Zhong-guo Zhou
Abstract
We construct liquidity-adjusted return and volatility using purposely designed liquidity metrics (liquidity jump and liquidity diffusion) that incorporate additional liquidity information. Based on these measures, we introduce a liquidity-adjusted ARMA-GARCH framework to address the limitations of traditional ARMA-GARCH models, which are not effectively in modeling illiquid assets with high liquidity variability, such as cryptocurrencies. We demonstrate that the liquidity-adjusted model improves model fit for cryptocurrencies, with greater volatility sensitivity to past shocks and reduced volatility persistence of erratic past volatility. Our model is validated by the empirical evidence that the liquidity-adjusted mean-variance (LAMV) portfolios outperform the traditional mean-variance (TMV) portfolios.
Complexity vs Empirical Score
- Math Complexity: 7.5/10
- Empirical Rigor: 6.0/10 (capped from 8.0: transaction costs not modelled; see penalty flags)
- Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 8.0)
Why this score: The paper employs advanced econometric models (ARMA-GARCH with liquidity adjustments) and constructs new liquidity metrics (jump and diffusion), indicating high mathematical density. It validates the model using minute-level trading data, applies it to out-of-sample portfolio optimization, and compares performance across asset classes, demonstrating strong empirical rigor and implementation readiness.
Research Flowchart
flowchart TD A["Research Goal: Model volatility & optimize portfolios for illiquid assets (e.g., cryptocurrencies)"] --> B["Data: Cryptocurrency OHLCV data<br>with specific liquidity metrics"] B --> C["Method: Construct Liquidity-Adjusted<br>Return & Volatility (Jump & Diffusion)"] C --> D["Computation: Integrate metrics into<br>Liquidity-Adjusted ARMA-GARCH Framework"] D --> E["Outcome: Improved Model Fit<br>Higher sensitivity to shocks<br>Lower persistence of erratic volatility"] E --> F["Validation: Liquidity-Adjusted Mean-Variance<br>Portfolios outperform Traditional Portfolios"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a liquidity‑adjusted ARMA‑GARCH framework that incorporates novel liquidity jump and liquidity diffusion metrics.
Source sentence
Based on these measures, we introduce a liquidity‑adjusted ARMA‑GARCH framework to address the limitations of traditional ARMA‑GARCH models, which are not effectively in modeling illiquid assets with high liquidity variability, such as cryptocurrencies.
Assumes 3
- Traditional ARMA‑GARCH models assume markets are efficient and liquid. market
Source sentence
Traditional ARMA‑GARCH (and subsequent ARMA‑GARCH‑Jump) models are not effective for cryptocurrencies for several reasons. First, these models are developed under an implicit assumption that markets are efficient and liquid, which is often not the case for a majority of cryptocurrencies.
- The risk‑free asset has a 0% interest rate. model
Source sentence
𝑟,*+ is the annualized realized daily returns for the risk‑free asset, and assumed to be 0 for simplicity.
- Liquidity measures (liquidity jump and diffusion) capture intraday liquidity shocks. model
Source sentence
We introduce liquidity jump and liquidity diffusion as novel liquidity measures that capture both the magnitude and intraday volatility of liquidity shocks, refining existing liquidity risk modeling approaches.
Data
- Binance API minute‑level crypto trading data other public
Source sentence
with at least four years of complete historical data (October 15, 2020 to February 7, 2025
- Polygon.io US‑stock minute‑level trading data other public
Source sentence
Using the minute‑level trading data (crypto trading data from the largest cryptocurrency exchange Binance API, and stock trading data from the Polygon.io API), we model this perceived and unobservable volatility as a minute‑level liquidity‑adjusted variance.
- Yahoo Finance market‑cap data fundamentals public
Source sentence
We obtain the market caps of all 1,503 stocks from the Yahoo Finance API on October 12, 2024 at the market close.
Costs
Not modeled
Finds 5
- Liquidity‑adjusted ARMA‑GARCH improves model fit for 7 of the 10 selected cryptocurrencies at the 1% significance level. statistical
Source sentence
out of the 10 selected cryptocurrencies, seven have their model fit improved by the liquidity‑adjusted model at 1% significant level, and two experience no statistically significant improvement
- Liquidity adjustment increases volatility shock sensitivity (α) for 9 of the 10 cryptocurrencies. statistical
Source sentence
(𝒶< 𝒶ℓ) ... The results show that the liquidity adjustment enhances the sensitivity of volatility to past shocks for essentially all selected cryptocurrencies.
- Liquidity adjustment reduces volatility persistence (β) for 7 of the 10 cryptocurrencies. statistical
Source sentence
(𝒷> 𝒷ℓ) ... The results suggest that liquidity adjustment reduces the persistence of volatility, making current volatility less dependent on erratic past volatility typical for cryptocurrencies.
- Liquidity‑adjusted mean‑variance (LAMV) portfolios achieve higher Sharpe ratios than traditional mean‑variance (TMV) portfolios for cryptocurrencies. performance
Source sentence
the empirical evidence that the liquidity‑adjusted mean‑variance (LAMV) portfolios outperform the traditional mean‑variance (TMV) portfolios.
- For US stocks, liquidity adjustment yields mixed Sharpe‑ratio results, with improvements in 9 of 15 portfolios and declines in 5. performance
Source sentence
Panel A of Table 7 presents the portfolio performance (Sharpe Ratio) for each of the 15 TMV and 15 LAMV portfolios for stocks. Among these, nine exhibit an improvement in SR by liquidity adjustment
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 48 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.