Working Papers

Predicting VIX with Adaptive Machine Learning

Yunfei Bai and Charlie X. Cai

This draft: 14 Jun 2021

Using 278 economic and financial variables we study the power of machine learning (ML) in predicting the daily CBOE implied volatility index (VIX). Designing and applying an automated three-step ML framework with a large number of algorithms we identify Adaptive Boosting as the best classification model chosen at the validation stage. It produces an average rate of 57% during the 11-year out-of-sample period. Potential significant economic gains are demonstrated in various applications with tradable instruments. Besides the modelling techniques, the weekly US jobless report is the most important contributor to the predictability along with some S&P 500 members’ technical indicators.

Keywords: Machine Learning, AutoML, Explainable AI, VIX, Predictability, Forecasting, Quantitative Trading, Big Data, S&P 500, Futures, US markets

Suggested citation:

Bai, Yunfei and Cai, Charlie Xiaowu, Predicting VIX with Adaptive Machine Learning (June 14, 2021). Available at SSRN: https://ssrn.com/abstract=3866415 or http://dx.doi.org/10.2139/ssrn.3866415

Economic Uncertainty: Mispricing and Risk Ambiguity Premium

Charlie X. Cai, Semih Kerestecioglu and Fu Xi

This draft: 7 Nov 2020

We study the effect of economic uncertainty exposure (EUE) on cross-sectional return differentiating the mispricing from ambiguity-premium effects. Conditional on a common mispricing index, we find that EUE induces disagreement which amplifies mispricing. The highest EUE quintile produces an annualized Fama-French six-factor mispricing alpha of 9%, more than double the unconditional mispricing effect. An ambiguity premium of 4.2% alpha is documented in the “non-mispricing” portfolio. The EUE induced mispricing effect is different from existing limits of arbitrage explanations, such as idiosyncratic risk. The ambiguity premium is a new source of the risk premium that is robust to the latest risk models, such as mispricing and q5 models.

Keywords: Economic uncertainty, Ambiguity aversion, Risk premium, Mispricing, Cross-section of stock returns, Return predictability

Suggested citation:

Cai, Charlie Xiaowu and Kerestecioglu, Semih and Fu, Xi, Economic Uncertainty Exposure and Cross-Sectional Return: Mispricing and Risk Premium (July 19, 2020). Available at SSRN: https://ssrn.com/abstract=3655670 or http://dx.doi.org/10.2139/ssrn.3655670

Informational Friction, Economic Uncertainty and CDS-Bond Basis

Charlie X. Cai, Xiaoxia Ye and Ran Zhao

This draft: Dec 10, 2020

We study how macroeconomic uncertainty (EU) manifests into the cross-sectional variations of the credit default swap (CDS)-bond bases. We develop a structural model in which common EU induces informational friction affecting the pricing in the bond and CDS markets. Higher EU will lead to a larger cross-sectional divergence in the bases. Furthermore, the difference between the two markets' exposure to EU measured by the EU betas can predict cross-sectional variations in the bases, which is confirmed in our empirical study. We also study the practical implication of EU as a new basis determinant in the context of the basis arbitrage.

Keywords: Uncertainty, Informational friction, CDS, CDS-bond basis, Uncertainty beta, Limits of Arbitrage

JEL Classification: G12, G13, G14

Suggested citation:

Cai, Charlie Xiaowu and Ye, Xiaoxia and Zhao, Ran, Informational Friction, Economic Uncertainty and CDS-Bond Basis (December 10, 2020). Available at SSRN: https://ssrn.com/abstract=3746637

Market Development, Information Diffusion and the Global Anomaly Puzzle

Charlie X. Cai, Kevin Keasey, Peng Li and Qi Zhang

This draft: Dec 15, 2020

Previous literature finds that anomalies are at least as prevalent in developed markets as in emerging markets; namely, the global anomaly puzzle. We show that while market development and information diffusion are linearly related, information diffusion has a nonlinear impact on anomalies. This is consistent with theoretical developments concerning the process of information diffusion. In extremely low efficiency regimes, without newswatchers sowing the seeds of price discovery and ensuring the long-run convergence of price to fundamental, initial mispricing and subsequent correction will not occur. The concentration of emerging countries in low efficiency regimes provides an explanation to the puzzle.

Keywords: Asset Pricing, Anomalies, Behavioral Finance, Multi-Factor Models, International Evidence

JEL Classification: G12, G14, G15

Suggested citation:

Cai, Charlie Xiaowu and Keasey, Kevin and Li, Peng and Zhang, Qi, Market Development, Information Diffusion and the Global Anomaly Puzzle (December 15, 2020). Available at SSRN: https://ssrn.com/abstract=2839799 or http://dx.doi.org/10.2139/ssrn.2839799

Investor Mix and Mutual Fund Performance: A Flow Based Measure of Relative Smartness

You Zhou, Peng Li, Charlie X. Cai, and Kevin Keasey

This draft: 30 Oct 2019

We study the information content of the mutual-fund investor mix at the fund level. Building on the fund-flow determinant literature, we develop a method to attribute the proportion of fund net-in-flow explained by a fund’s fundamental characteristics and past performance as smart and dumb money respectively. The fund-level Smart Dumb Ratio (SDR) positively predicts future cross-sectional fund return. A series of tests shows that SDR postively correlates with other skill measures and its effects are enhanced with investor sophistication. Our findings confirm that the investor composition can be a useful source of information to estimate the fund-level smart-money effect.

Keywords: mutual-fund flows, risk factors, non-risk factors, smart-money effect, CAPM

JEL Classification: G11, G12

Suggested Citation:

Zhou, You and Li, Peng and Cai, Charlie Xiaowu and Keasey, Kevin, Investor Mix and Mutual Fund Performance: A Flow Based Measure of Relative Smartness (February 11, 2020). Available at SSRN: https://ssrn.com/abstract=2839798 or http://dx.doi.org/10.2139/ssrn.2839798


Nonlinear Limits to Arbitrage

Charlie X. Cai, Jingzhi Chen, Robert Faff, and Yongcheol Shin

This draft: 08 Feb 2021

We capture the nonlinear nature of limits to arbitrage. Specially, we investigate a complex interaction between arbitrage costs and funding constraints in shaping the nonlinear relation between mispricing and arbitrage activity. When mispricing is small, arbitrage activity intensifies with mispricing because of the higher cost-adjusted return. However, at high mispricing levels, arbitrage activity is rather deterred by this larger mispricing as funding constraints become more binding. Working on the index spot-futures arbitrage with a Markov regime-switching (generalized) error correction model, we provide empirical evidence where mispricing and arbitrage activity are consistent with such an inverse U-shaped relation. The extreme regime is characterized with extremely large mispricing but least arbitrage activity and coincides with the market turmoils, suggesting that funding constraints become the main driver behind the limit to arbitrage.


Keywords: Index Arbitrage, Limited Arbitrage, Noise Momentum, Futures and Spot Prices, Markov Switching Model

JEL Classification: C12, C22, G13, G14

Suggested Citation:

Cai, Charlie Xiaowu and Faff, Robert W. and Shin, Yongcheol, Limited Arbitrage and Noise Momentum (July 19, 2012). Available at SSRN: https://ssrn.com/abstract=1571931 or http://dx.doi.org/10.2139/ssrn.1571931