Tactical methodologies generally fall into one of three categories—trend following, counter trend following, or fundamentally based. In a Trend Aggregation approach no one methodology is better than the other, each has markets it works well in and markets it doesn’t, and a diversified tactical portfolio should have dynamic exposure to all three. However, counter trend strategies will generally provide the best risk vs. return and work well in more market environments than the other two. A counter trend model will use some sort of metric to determine whether a market is oversold (everyone who has sold is done selling and it is at a level that is likely to draw in bargain hunting buyers) or overbought (everyone who has bought is done buying and it is at a level that will entice people to take profits). The model would buy what is oversold and sell, or sell short, what is overbought. Basically acting counter to the trend.
The one “Achilles Heal” that counter trend models can have is around V shaped corrections. These are sharp selloffs that come with very little, if any warning, following by sharp reversals. Because counter trend models buy into weakness they can be hurt when that weakness continues for a while. In a “normal” market, a counter trend model would buy into a short term low and then sell on a bounce. In a V shaped correction, the short term low can keep going lower, and lower. Also, in a “normal” market a counter trend model would sell into a short term high and look to buy back in at a lower price. In the reversal from a V shaped correction a market can move up quite a bit, leaving the counter trend model behind.
Just like bond investors ladder bonds to protect from interest rate risk, tactical investors can ladder counter trend models to protect from V shaped correction risk. This is done by using models that have different metrics and time frames. That way, all models won’t go from bearish to bullish at the same time, they will tend to leg into markets on the downside. The same concept works on the upside where models will tend to leg out. A simple example of this could be combining different period RSI counter trend models. The RSI is a measure of the magnitude of prices changes in a security. Typically, high RSI readings can indicate a security is overbought and low readings could indicate that it is oversold. A 2 day RSI model could be combined with a 3 day model and a 4 day model. The 2 day model would need weakness over 2 days to get into a market and strength over 2 days to get out. The 3 day model would need weakness over three days, and the 4 day model would need 4 days. You could also change the levels that the model needed to hit, some models could need to be very oversold to trigger a buy, while others might just need to be moderately oversold. Finally, some models could use a moving average filter, while others do not. For example, an RSI model for the S&P 500 could shut off entirely when the S&P is below its 200 day moving average.
Instead of using one counter trend model that will be either all in or all out, investors will be better served using multiple models that don’t use the same time frames and metrics so they can scale in and out of volatile markets.