Rather than post the daily bond gauge and the weekly I now post only the average of the two. I do this because one of my trading algo’s prefers it. I have written about how the daily gauge is very volatile and how it often spends time at either extreme (0 or 100). The weekly is correspondingly nonvolatile since it looks at inputs using weekly bars. When we combine them we get this distribution:
Given the cluster around 50-60 we would look at any reading above 50 as acceptable to get a buy (though there may be other conditions attached to the signal). At no time in the past 10 years did the gauge get above 95.
Now let’s look at the returns for various levels of the gauge:
We should expect superior positive returns as the gauge level rises. Because the weekly gauge levels are included we should also expect that same relationship to hold for both two and three days forward. We do. There are clearly some outliers that break the relationship in the 60-65 region. That is mainly due to a 5 point decline on Nov.1 2016. The average absolute return for bonds over the period was .66 points.
Inputs: This combo-gauge uses inflation proxies, economic bullishness, as measured by stock market momentum, and bond market fear as its key components. Utility stocks performance contributes to the weekly bond component.