A slowing in the fall.

Photographer: Christopher Furlong/Getty Images

Don't Count Out Seasonal Patterns to Predict Rates

David Ader is chief macro strategist at Informa Financial Intelligence.
Read More.
a | A

Bond strategists use many tools in attempting to forecast the path of interest rates, ranging from straightforward data analysis, to parsing the statements of Federal Reserve officials, to looking for trading patterns. The challenge lay in weighting these methods in terms of their market influence for time frames from the immediate to the very long haul.

There is, however, one simple technique that has proved very reliable when trying to provide a year-ahead outlook: monitoring seasonal patterns in interest rates and the yield curve. Simply put, there is a strong tendency for yields to rise into the May-June period and drop heavily into the end of year; 2017 may be no exception.

A glance at a chart of average yield moves since 2000 makes for a compelling story, even though each year has its own oddities and the seasonal approach will be subject to challenges (it didn’t quite work for 2016, for example). Nonetheless, this pattern held up well in 10 of the last 17 years, was neutral in two and didn’t work in just five. The pattern was even stronger from 1990 to 1999. Those are not bad odds.

Further, in two of the years when the seasonal pattern failed, the following year it either reverted to the norm or was neutral. In short there is strong argument to be made that yields this year will tend to rise into that May-June period, providing a buying opportunity.

Excluding the possibility of remarkable coincidences, what explains these seasonal patterns?

From the 1990s to just a few years ago, some of the onus was on the investment behavior of Japan, one of the largest owners of U.S. debt, Japan's fiscal year closes at the end of March, which was accompanied by repatriation of profits -- selling Treasuries -- and investors didn't begin reloading until the period around the U.S. government's quarterly refunding in May. This coincided with the period when investors needed the bonds to ensure liquidity. This pattern has broken down somewhat in recent years.

There also was a seasonal bias for a pickup in inflation early in the year (companies getting price increases in early to book them for the bulk of the year).  As a result, Treasury Inflation-Protected Securities also benefited in the first few months of a given year, with break-even rates -- the difference in yields between TIPS and conventional Treasuries, and an indicator of the rate of inflation expected by investors -- tending to reach their widest levels around May-June and then narrowing for the balance of the year.

Another noticeable factor was that companies tended to frontload bond issues in the first months of the year, funneling more supply into a relatively narrow window before slowing the pace somewhat into the summer and fall.

It's been a painful time to own bonds, with the market in recent months suffering one of its worse selloffs on the prospects for faster economic growth, inflation and pace of Fed rate increases. The Bloomberg Barclays U.S. Treasury Index has fallen 5.54 percent since peaking in July.

The recent pattern has been driven by a cycle of anticipation of Fed rate hikes that fizzled as some data point -- sluggish growth or slower inflation -- or an external event such as Brexit forced a change of course and restored the seasonal driver. 

As I’ve looked at these seasonals, I’ve been humbled every time I thought that the factors that create the seasonal pattern were about to break down or change.  Instead, they've held up incredibly well. 

Even though the seasonal pattern is just one of many things to look at and should be treated more as a road map than a trading model, it remains a predictor that I’ll follow closely as the year unfolds.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
David Ader at dader2@bloomberg.net

To contact the editor responsible for this story:
Max Berley at mberley@bloomberg.net