Explaining Momentum

The Rivkin Local Momentum Strategy is a purely systematic, price based trading system that produces excellent results based on our back-testing over the previous 15 years. The aim of this article is to give an overview of the concept of momentum in stock prices. My suspicion is that many investors have a general idea what momentum is but don't actually understand how it is measured or how we can profit from it.

In science, the concept of momentum is very well defined, with explicit formulas used to calculate it. When it comes to investing however, momentum can be defined in myriad different ways. For this reason, it is impossible to give an explicit definition of a stock's momentum and there are probably thousands of different momentum based strategies in existence around the world, all based on slight variations of the underlying concept.  

Nevertheless, just like in science, the phenomenon that an object once in motion tends to stay in motion (momentum) is why it appeals to stock investors alike. There have been numerous studies on the momentum effect with most finding that there is a genuine statistical basis for trend following strategies. One well known saying in investing "the trend is your friend" is a reference to the momentum effect.

Let's run through a basic momentum strategy to help explain the concept.

A simple momentum strategy could be based on the moving average of the stock price. A moving average is simply the average of the stock price over a given time range. For example, the 100 day moving average (100DMA) of a stock is the average of the close price for the current day and the prior 99 trading days. The term 'moving' means that tomorrow, the last price in the calculation will drop off to make way for the new price (tomorrow's close price) so that there are only ever 100 days used in the calculation. Similarly, a moving average for any specified number of days can be calculated, such as a 50 day moving average. As the number of days in the calculation is increased, the slower the average will respond to new prices, and visa versa. Therefore a sudden increase in the share price will, for example, cause the 50DMA to move up faster than the 100DMA.

Figure 1: ASX200 plotted over the past year. Blue line is the 100DMA and the orange is the 50DMA. The 50DMA crossed above the 100DMA in April 2016, indicating that the ASX200 is now in a positive trend. (Source: AmiBroker)

A basic momentum strategy can be formed by combining the 50DMA and the 100DMA. Such a strategy would state that a stock is considered to be in a positive trend if the 50DMA is above the 100DMA and in a negative trend if the 50DMA is below the 100DMA. A long momentum based strategy would therefore be looking for stocks that are in a positive trend. This criterion alone isn't generally enough to produce a successful trading strategy. There must also be criteria to trigger a trade entry signal. In our example strategy, such criteria might include a 50 day breakout in the direction of the trend. A 50 day breakout simply means that the stock closes at the highest level in the past 50 days. If these two criteria are satisfied (positive trend and 50 day breakout) a buy setup is generated and the stock is bought the next day.

A full trade strategy would also include a process for determining when to exit a position. The Rivkin momentum strategy uses a monthly rebalancing system to manage positions, but there are many other possible options for this. Sticking to a plan for exiting trades is arguably the most important part of any trading strategy. It is extremely important that, when using a systematic strategy, emotion is not allowed to dictate when to exit positions.

In this simple example, we have used only one selection criterion and one trade entry criterion; however, any number of criteria could be used to generate a trading strategy. The important thing is that any new strategy is back-tested in a rigorous manner to evaluate its profitability (or otherwise)

The sample strategy I have described isn't the exact strategy used for the Rivkin Local Momentum portfolio but the basic principles are the same. We have performed back-testing over 15 years on numerous combinations of criteria to find the combination that produces desirable characteristics (Including a strategy based on US stocks, click here for a comprehensive webinar recording of this strategy). Investors should realise that the overall profitability of a strategy isn't the only factor to be considered. Risk related measures, such as the maximum drawdown, should also be considered. For this reason, the Rivkin Investment Team has found a system that provides a good return for an acceptable level of risk. Furthermore, it must be stated that even the most profitable strategy will have periods of poor performance. Trading strategies like this are based on a statistical probability of being profitable in the long term. No trading strategy can have 100% success on each trade, and therefore there will be trades that lose money and there will be periods of poor performance for the portfolio as a whole.

Any investor interested in the Rivkin Momentum strategy can click here to learn more.