In our final blog post, that finishes the trend-following series, we briefly review the results of the forward-tests of the profitability of various trend following rules in different financial markets: stocks, bonds, currencies, and commodities. The results of these tests allow us to better understand the properties of the trend following strategies, their advantages, and their disadvantages.
Prior to considering each specific financial market, let us elaborate on the factors that play role in the success of a trend following strategy.These factors are as follows:
These factors are as follows:
- The duration of a trend should be long enough to make the trend following strategy profitable. For the sake of illustration, consider the popular P-SMA(10) strategy which uses a 10-month simple moving average. We know (see Part 2 of this blog series) that SMA(10) identifies a turning point in the trend with the average delay of 4.5 months. Therefore, as a ballpark estimate, the duration of a trend should be no less than 10-12 months in order to make the P-SMA(10) strategy profitable. Since trends have various durations, longer trends should prevail over shorter trends.
- Whether the overall long-term market trend is upward, sideways, or downward. When the market trends strongly upward over the long run, bull markets are much longer than bear markets. Because short bear markets cannot be identified with a good enough accuracy, a trend following strategy is not profitable in these circumstances. A trend following strategy is usually profitable when the market goes sideways or downward over the long run. A possible indicator of the success of a trend following strategy can be the ratio of the average bull market length to the average bear market length. If this ratio is close to unity, then it indicates that the market has been going basically sideways and a trend following strategy may be highly profitable. The higher (smaller) the value of this ratio, the less (more) profitable a trend following strategy is.
- Bear markets should be not only long enough, but with a large enough price decrease. That is, generally the larger the drawdowns, the better the profitability of a trend following strategy. When the drawdowns are small and the price volatility is high, a trend following strategy is not profitable even if the ratio of the average bull market length to the average bear market length is close to unity.
When we talk about bull and bear markets, we talk about the so-called “primary markets” (a.k.a. “cyclical trends”); the length of these markets generally varies from one to five years. Besides the primary market trends, in each financial market one can easily observe the long-term trends known as “secular markets” or trends. A secular trend, which lasts from one to two decades, holds within its parameters many primary trends. For example, a secular bull market will have bear market periods within it, but it will not reverse the overlying trend of upward asset values. Similarly, a secular bear market will have bull market periods within it.
Why is it important to be aware of secular market trends? It is important because the dynamic of the bull-bear markets undergoes a regime shift when a secular market trend changes its direction. Specifically, a secular bull market is characterized by long bull markets and short bear markets. In contrast, a secular bear market is characterized by short bull markets and long bear markets. As a result, a trend following strategy outperforms its passive counterpart basically over secular bear markets. Over secular bull markets, on the other hand, a trend following strategy usually underperforms the corresponding buy-and-hold strategy.
We tested the profitability of trend-following rules using monthly data on five stock market indices in the US. They are the Dow Jones Industrial Average (DJIA) index, the large cap stock index, the small cap stock index, the growth stock index, and the value stock index.
The DJIA index is a price-weighted stock index. Specifically, the DJIA is an index of the prices of 30 large US corporations selected to represent a cross-section of US industry. All other indices are value-weighted indices. The large cap stock index resembles the S&P 500 index. By definition, growth stocks (a.k.a. the “glamor” stocks) are stocks of companies that generate substantial cash flow and whose earnings are expected to grow at a faster rate than that of an average company. Value stocks are stocks that tend to trade at a lower price relative to its fundamentals and thus considered undervalued by investors. Common characteristics of such stocks include a high dividend yield, low price-to-book ratio, and low price-to-earnings ratio. Small cap stocks are stocks of companies with a relatively small market capitalization.
Before reporting the results of our forward tests, let us consider secular bull and bear markets in stocks. There is an extensive literature on the secular stock market trends, see, among others, Alexander (2000), Easterling (2005), Rogers (2005), Katsenelson (2007), and Hirsch (2012). The figure below, upper panel, plots the Shiller’s Cyclically Adjusted Price-to-Earnings ratio (CAPE) computed using the 10-year average earnings divided by the price of the S&P 500 index. The lower panel in this figure plots the natural logarithm (log for short) of the S&P 500 index. Shaded areas in this panel highlight the primary bear markets. A secular bull market is characterized by a long term increase in the CAPE and the stock prices. A secular bear market is characterized by a long-term decrease in the CAPE. During a secular bear market the stock prices go sideways. Researchers disagree on the exact dating of the secular bull-bear markets, but the approximate dating of these trends is as follows. From 1900 to date, there were three secular bull markets and three secular bear markets. Secular bull markets covered the periods from 1920 to 1929, from 1949 to 1965, and from 1982 to 2000. It is quite likely that the new secular bull market started in 2009.
The results are hypothetical results and are NOT an indicator of future results and do NOT represent returns that any investor actually attained. Indexes are unmanaged, do not reflect management or trading fees, and one cannot invest directly in an index. Additional information regarding the construction of these results is available upon request.
We remind the reader that the trend-following strategy usually underperforms the passive strategy during secular bull markets that, in the stock market, last from one to two decades. For example, the trend-following strategy underperformed the buy-and-hold strategy over the period from 1982 to 2000. No wonder that trend-following went out of favor by the end of 1990s. Trend-following trading became again popular in the aftermath of the two recent severe stock market crashes (the Dot-com bubble crash and the Global Financial Crisis). Many investors started implementing