Mikhail Zhitlukhin

Stock Market Crashes: Predictable and Unpredictable and What to do About Them

This book presents studies of stock market crashes big and small that occur from bubbles bursting or other reasons. By a bubble we mean that prices are rising just because they are rising and that prices exceed fundamental values. A bubble can be a large rise in prices followed by a steep fall. The focus is on determining if a bubble actually exists, on models to predict stock market declines in bubble-like markets and exit strategies from these bubble-like markets. We list historical great bubbles of various markets over hundreds of years.

When to Sell Apple and the NASDAQ? Trading Bubbles with a Stochastic Disorder Model

In this paper, the authors apply a continuous time stochastic process model developed by Shiryaev and Zhutlukhin for optimal stopping of random price processes that appear to be bubbles. By a bubble we mean the rising price is largely based on the expectation of higher and higher future prices. Futures traders such as George Soros attempt to trade such markets. The idea is to exit near the peak from a starting long position. The model applies equally well on the short side, that is when to enter and exit a short position. In this paper we test the model in two technology markets. These include the price of Apple computer stock AAPL from various times in 2009–2012 after the local low of March 6, 2009; plus a market where it is known that the generally very successful bubble trader George Soros lost money by shorting the NASDAQ-100 stock index too soon in 2000. The Shiryaev-Zhitlukhin model provides good exit points in both situations that would have been profitable to speculators following the model. who employed the model.

The High Price–Earnings Stock Market Danger Approach of Campbell and Shiller versus the BSEYD Model

We show how to use Campbell and Shiller’s work on price-to-earnings (P/E) ratio and the predictability of long-term returns to create a crash prediction measure: the high P/E measure. Next, we present a statistical procedure to test the accuracy of crash prediction models. We use this procedure to test the accuracy of the bond–stock earnings yield differential (BSEYD) and high P/E models on a 51-year period on the US market, starting on January 1, 1962, and ending on December 31, 2014 (12,846 daily data points). At the end of the Chapter, we expand the analysis beyond the US market to look at the two main Chinese stock markets: Shanghai and Shenzhen. Material in this chapter is based on Lleo and Ziemba (2017) and Lleo and Ziemba (2016c).

Discovery of the Bond–Stock Earnings Yield Differential Model

We discuss the bond–stock earnings yield differential (BSEYD) model starting from when Ziemba first used it in Japan in 1988–89 in various countries. The model has called many but not all crashes. Those have high interest rates in the most liquid long-term bonds relative to the trailing earnings-to-price ratio (EP ratio). In general, when the model is in the danger zone, there will almost always be a crash. The model called the 2000 and 2002 US crashes. A long horizon study for the US, Canada, Japan, Germany, and UK shows that being in the stock market when the bond–stock signal is not in the danger zone and in cash when it is in the danger zone provides a final wealth about double buy and hold in these five countries during 1975–2000 or 1980–2000.

Effect of Fed Meetings and Small-Cap Dominance

In this chapter we discuss the generally positive effects on the US stock market of FED meetings and small cap stocks.Sixty-fourty pension fund fixed mix and presidential party effects are studied along with the effects on the stock market when congress is in session.Five strategies are presented and two simple presidential party strategies have over long periods produced higher returns than small or large cap stocks and about twenty times more final wealth than the highly recommended 60-40 stock-bond mix.

A Monotone Performance Measure Based on the Sharpe Ratio

In this paper we present a modification of the Sharpe ratio, which is monotone with respect to second-order stochastic dominance. We study its properties and obtain a representation which allows to compute it in an efficient way.