Interest Rates: The Bigger Picture
Board: Value Hounds
Interest rates are rising -- it's a no-brainer.
It seems to me that many investors have focused on the decline of interest rates since the 1980's and therefore with a cyclical upturn in the economy (car sales, housing prices, employment numbers) then interest rates likely have bottomed and away we go. Its been a great run for the bond bulls but all great things must come to an end -- right?
But couldn't you argue that the bubble was in positive real interest rates caused by Paul Volcker -- ok, that's a bit over the top given the inflation scenario the Chairman faces, but consider from 1980 till now -- real interest rates have been positive w/ brief dips into negative territory in 2008 and 2012. Looking back at Professor Shiller's data set starting in 1872; this time period is a pretty big outlier for consistently positive real interest rates with the decade before the Great Depression coming in second place (eyeballing it here).
Taking the tin-foil hat off -- what gets me though is the lack of discussion of our debt overhang when it comes to interest rates. I guess with the R&R 90% debt-to-gdp takedown -- debt overhangs don't matter. The debt-to-gdp numbers for the Federal Govt (excluding entitlements) though are pretty high. Last time we saw this level of debt-to-gdp was in the mid 1940s. The CBO is projecting that its going higher based on 2% or less net interest costs through 2015 and 2.3% in 2016. (The 10-year yield tracks the net interest cost of the Federal Govt pretty close.)
So if you go back to pre-Volcker and more specifically to the last time we had such a huge debt overhang gets us to a period of 1940-1980. While its somewhat interesting to look at nominal rates -- its the real rate that we should be looking at. Inflation actually is beneficial to reducing debt overhangs but introduces other challenges -- but I digress.
You can break that 40 year period into two rough groups: 1940-1952 and 1952-1980. The major deleveraging from that period ended in the early 1950s. US Household debt as % of net worth bottomed about 6-7% in early 50s. It topped out in 2007 around 22%. The Federal Govts debt to the public didn't bottom from its peak in the 40s until the mid 1970s. Spending on the Vietnam War clearly slowed the process of deleveraging at the Federal level as well as in Korea in the early 50s.
Saying that -- from 1940-1952; real interest rates were volatile but also predominantly negative with the highest peak at 5% in 1949. The median monthly rate from 1940-1952 was a negative .92% and the average rate was a negative -2.84%.
From 1952-1980 the median monthly real interest rate on the 10-year was 1.8% and the average was 1.3%. There didn't look to be one month of real interest rates above 4.0%.
So if we start with a situation where deficit spending is expected to continue for sometime driving the debt-to-gdp ratio even higher based on low net interest costs; the rise in interest rates are only going to compound the problem. Which gets me to the crux:
As everyone expects interest rates to rise -- history suggests something else is likely to happen. And we can look explicitly at the impact of higher rates on the Federal Government's financial situation to suggest action at the Federal level to keep debt down at minimum at a relative level. (The recent taper talk was a major screw up in my opinion based on rapid shift in rate expectations, which were likely not intended). While Federal cutbacks (austerity) are in play -- this line of action has its own drawbacks. And we can add the structural challenges including wonky topics such as labor vs capital, education, inequality, demographics etc which are other headwinds to economic growth and thus interest rates. Base on this line of thinking, sans short-term volatility, I expect its more reasonable to see a prolonged period of low interest rates as can be thought of as a stealth tax required to liquidate the debt overhang. Inflation is clearly an x-factor (which has failed to materialize yet.)
Again -- I don't think it is so obvious that rates have to rise when you broaden your historical perspective past the 1980s and review the previous experience when we had a significant debt overhang challenge as a country (as did many developed countries back then). As for my own thoughts -- negative rates on the 10-year in 2012 was not an anomaly but more likely the norm for sometime. I have mentally penned 15 years (+/-) for the deleveraging period that began in 2007 and that puts us at 2022 (+/-). Since Dec' 2007 the average and median monthly real interest rate on the 10-year was .83% and .23% respectively which carries us half way through the 6th year of a guesstimated 15-year deleveraging cycle.
Time will tell
My 2c worth
PS. My thoughts on fair value for the 10-year Treasury
1# Based on a forward projection of the FOMC projections including inflation and unemployment expectations then I calculate a 2.2% FV (inline w/ the much embattled Bill Gross) and 2.0% if you remove the 0% boundary (which has been worked around via QEx) and 2.5% in 2014 and no expectations of any QE as the short-term rate expectations should be slightly positive.
This is based on the market expectations hypothesis where the 10-year is the average of the expected short-term rates experienced over the next 10-years. The short-term rates are estimated based on FOMC projects and a regression derived formula between unemployment and inflation.
2# If you regress the ISM New Order's index, 3-month T-Bills, and Core CPI -- and looks at June data -- I get a FV ~ 2.5%; This data can be obtained through the STL FRED database.
3# Average of past 5 years of nominal GDP. This generates the highest estimate as the 2008 recession data rolls off and I get 3.4% versus 2.3% for 2012 and 2.4% in 2011. This trend has had reasonable correlation over the past but since 1980, the interest rates have fallen slower than the slowdown in nominal GDP showing slow shifts in inflation expectations.
I am leaning towards the 2.4% area for now and have thought of the range of 2.0% to 2.4% as FV -- but with the ISM regression, I might want to raise it to 2.2% to 2.5%. The 2.0% is driven by the obtainment of "implied" negative short-term rates through QEx on the long-end of the curve. (IE 2012 and 2013 implied short-term rates were negative 1.9% and 1.8% based on unemployment rates above 7% and low inflation rates.)