5-Year Cycles: About Time To Chuck Treasuries?

by: John Overstreet


The growing evidence suggests that the bond market is running out of steam and the risks are rapidly outgrowing the avowed safety of Treasuries.

It is prudent to stay away from Treasuries for the next three years.

Investors will have to start becoming more aggressive about how they protect their money.

Slowly accumulating rate-sensitive assets like financial stocks might allow one to begin preparing for a market that could turn surprisingly quickly within the next year.

In my previous article, I made a case for why it was best to avoid Treasuries for the rest of the year. Now, I want to make a separate but overlapping case for why it is prudent to stay away from Treasuries for the next, say, three years. Last time, I mentioned August as a good place for a top in the bond market this year, but there is no reason to be dogmatic about this or that date. The growing evidence simply suggests that the bond market is running out of steam and that the risks are rapidly outgrowing the avowed safety of Treasuries.

A rather grizzly metaphor comes to mind, that of oil trucks in third world countries overturning and then exploding while residents try to scoop up fuel gushing out onto the street. But, I will try to show you what I mean.

In a slightly different context, I showed this rather prosaic chart of ten-year Treasury rates.

(click to enlarge)10-year Treasury yieldsClick to enlarge

If you look at the rates over the last thirty years to 1983, there was only one occasion in which rates did not drop to distinct lows with years ending in '3' or '8'. That one exception was 1988, when that low seemed to arrive about a year early. Most of those lows were followed by fairly strong snap-backs.

The years 1968 and 1978 appear to largely conform to the general pattern, as well, although the dip occurs a little early in the latter instance. It is interesting that the two biggest outliers--1973 and 1988--coincided with Leeb oil shocks (an 80% yoy increase in WTI prices), although rates in 2008 appeared relatively unperturbed by the oil shock of that year. It is possible that the "premature" fall in interest rates prior to 1973 and 1988 are somehow tied to the flattening of the yield curve that preceded the 1973 oil crisis and the 1990 spike in oil after the Iraqi invasion of Kuwait. That is speculation for a different occasion, however.

For now, let's stick with Treasuries and see how much we can model or quantify. There is always the danger of seeing things in charts that aren't really there, so I decided to look at the data from a few different angles.

First, I broke up each of these supposed five-year cycles and compared them back to 1958. (Note: all the charts in this piece are based on data from the St. Louis Fed as of the end of June 2012).

(click to enlarge)historical 5yr cycles treasuriesClick to enlarge

"Year 1" would be equivalent to a year ending in a '3' or '8', such as 1978 or 2003. Setting "Year 1" here was fairly arbitrary; I only set it there for purposes of visualization.

In any case, this chart gives one pause. This looks more like how we tend to think of markets, a cacophony with just enough coincidental rises and falls to suck in someone desperate to see a pattern.

So, I decided to take an average of each step along the possible cycles from 1958-2007 (since the 2008-2012 cycle is not yet complete) and I found the following, which seems suggestive.

(click to enlarge)Average cycles vs current cycleClick to enlarge

There are two years of rapid acceleration followed by three years of a more gentle but less consistent fall.

But, if you overlay an individual cycle like we are going through now, you can see that although there appears to be something of a family resemblance, it is not a close bond. It is interesting that at the end of 2008, that drop managed to squeeze itself in just in the nick of time, however.

Here is the 2003-2007 cycle next to the averages.

(click to enlarge)5yr cyclical interest rates (ave) vs 2003-2007 cycleClick to enlarge

In the 2003-2007 cycle, there is an erratic rise in the first two years, but the peak that "should" have come in 2005 seems blunted or perhaps displaced to 2006. This appears to be Greenspan's conundrum in action.

Comparing each of the 11 cycles of the last 55 years (if we include the current one) to the 1958-2007 average shows some inconsistency in the correlations, but to my mind, they are still surprisingly strong.

(click to enlarge)correlation between individual cycles and overall averageClick to enlarge

I also looked at three-cycle moving averages.

(click to enlarge)Moving averagesClick to enlarge

This seemed to be mostly consistent, although there were two periods that did not seem to conform especially well:

(click to enlarge)outliersClick to enlarge

Much like the 2003-2007 cycle, the "Year 3" appears to be blunted or displaced, but the cycles still seem to be largely intact. I take it that this is due to the "premature" dips prior to 1973 and 1988 that I mentioned above.

As one final aside, I wanted to mention what I thought was an interesting correlation with data in Carmen Reinhart and Ken Rogoff's famous This Time is Different. In Table 13.1, they list the "Big Five" systemic banking crises. Those episodes began in 1977, 1987, 1991 (in two countries), and 1992. These each occurred within two years of Leeb oil shocks and three of them happened in "Year 5" while the other two were in "Year 4". Turning to Table 15.1, the post-Depression "global banking crises" occurred in 1987-1988, 1991-1992, 1994-1995, 1997-1999, and 2007-"present" (i.e., 2009). I think we could include the Euro crisis of 2011-2012, as well. The only outlier is the 1994-1995 "Tequila crisis".

It is somewhat gratuitous for me to ask if there actually are five-year cycles at this point, because I believe there are. But, it is a blunt instrument nevertheless, and I haven't a clue as to why such a cycle should exist and therefore have no basis upon which to assume that it will continue to operate in the future, although I think that there are avenues of investigation that are opening up.

In the meantime, based on a blind application of it to our current situation, bonds could be expected to stabilize or strengthen into 2013 before sharply reversing into 2015.

But, because there has historically been so much variation within each particular cycle, it is very difficult to be sure whether the cyclical bottom will manifest itself this year or next, at least without referring to other tools at our disposal. In other words, although there seems to be little reason to panic at present, it also seems likely that pressure will grow for interest rates to rise and that investors will be forced to look for alternatives to the Treasury market.

Whether we should look to commodities or stocks for solace is a question for another time and probably largely depends on what the catalyst for a rise in interest rates would be. At present, a sorting out of the Eurozone crisis might create conditions under which bullishness could return while simultaneously bringing the United States' fiscal position under the spotlight.

Shorting bonds (NYSEARCA:IEF) at this moment seems unnecessarily risky, but if we are on the verge of an unwinding of this 30-year bull market, investors will have to start becoming more aggressive about how they protect their money. Slowly accumulating rate-sensitive assets like financial stocks and ETFs (NYSEARCA:XLF), as I mentioned in my previous effort, might allow one to begin preparing for a market that could turn surprisingly quickly within the next year.

Disclosure: The author is long BAC. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it. The author has no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I am long September WTI.