Johnny Hong    Home    CV    Research/Projects    Coursework    Teaching    Miscellaneous


Abstract: In the setting of high-frequency trading, stock price movements are fairly predictable. We use support vector machines (SVMs) with various kernels and random forests to predict the midprice movements of SPY (SPDR S&P 500 Trust ETF). We find that the machine learning approach provides good prediction accuracy. This work is based on the paper “Modeling high-frequency limit order book dynamics with support vector machines” by Kercheval and Zhang.

Abstract: We construct two HMMs to model the stock returns for every 10-day period. Our first model uses the Baum-Welch algorithm for inference about volatility, which regards volatility as hidden states and uses a mean zero Gaussian distribution as the emission probability for the stock returns. Our second model uses a spectral algorithm to perform stock returns forecasting. We analyze the tradeoffs of these two implementations as well. Code