Are Econometric Models Useful in Trading?

We have previously presented time series analysis for identifying autocorrelation properties of stock indices and econometric techniques such as ARIMA and GARCH for estimating volatilities. We also highlighted an article [1] that demonstrated the usefulness of advanced volatility estimators in trading by reducing trading strategies’ turnover.

On the same topic, Reference [2] attempted to directly incorporate econometric models into existing trading systems. Specifically, it utilized three econometric models to forecast SPY prices: (i) Auto Regressive Integrated Moving Average (ARIMA), (ii) Generalized Auto Regressive Conditional Heteroskedasticity (GARCH), and (iii) Vector Autoregression (VAR). It then integrated these models into existing trading strategies that use two technical indicators, Bollinger Bands and Moving Average Convergence Divergence (MACD).

However, unlike Reference [1], the authors concluded that incorporating these econometric models did not improve the trading strategies’ performance.

In this paper, we consider linear process models using the VIX Index as a proxy for market sentiment, to predict the movement of the SPY ETF prices. The coefficients on the SPY lags of the ARIMA(2,1,1) model turn out to be statistically insignificant with the addition of the VIX Index, even though this particular model performs better out-of-sample than the ARIMA(2,1,1) model without the VIX Index. Furthermore, the one-step forward forecasts generated by a VAR(1) model performs better than an ARIMA(2,1,1) model. However, in the context of algorithmic trading strategy, it counters the many signals generated by Bollinger Bands and MACD. The one-step forward GARCH(1,1) volatilities do not help in the context of algorithmic trading strategies either, perhaps because of the limited predictive capability of the VAR(1) model.

In summary, the 2 articles’ results are contradictory.

So what do you think? Are econometric models useful in trading or not?

References

[1]  Baltas, Nick and Kosowski, Robert, Demystifying Time-Series Momentum Strategies: Volatility Estimators, Trading Rules and Pairwise Correlations. “Market Momentum: Theory and Practice”, Wiley, 2020

[2] YM Kobara, C Pehlivanoglu, OJ Okigbo, A Linear Process Approach to Short-term Trading Using the VIX Index as a Sentiment Indicator, Preprints 2021, 2021070673

Leave a Reply

Your email address will not be published. Required fields are marked *