Boundary Conditions

A reader asked if I think the yield curve will flatten. The answer is yes, I do.

A steeply positive yield curve - as it is now - is the bond market's way of discounting a sharp rebound in the economy; I believe this optimism is unfounded and will soon be proven wrong, leading to a reversal. Sketching the picture in very broad lines, declining consumer spending (70% of GDP) will be the most serious drag on the US economy, due to persistently high and rising unemployment.

Let's look at the curve itself (see chart below).

US Treasury Yield Curve

The next chart shows the spread, or difference, between interest rates for 10-year Treasury bonds and 3-month bills since 1953. Currently the rates are 3.47% and 0.07% respectively, resulting in a spread of 3.40%, or 340 basis points (a basis point is 0.01%). That's very high by historical standards (see chart below, click to enlarge).

Data: FRB St. Louis

Next, I perform a frequency analysis for this spread and produce the following histogram (click to enlarge).

In the last 45 years the spread averaged 139 bp with a standard deviation of 119 bp. Quite obviously, the vast majority of the values are positive (above zero), since inverted yield curves are rare.

The charts above, however, do not tell the whole story because they track the difference in rates without taking into consideration their absolute levels. Because nominal interest rates cannot go below zero, it makes a whole lot of difference if a spread of 340 b.p. is the result of 6.40% minus 3.00% or 3.50% minus 0.10%. These two situations describe completely different economic and financial market conditions, even if the spread is exactly the same. So, we need to look at the spread from another perspective.

• Let's chart the spread as a ratio of the 10-year bond rate, i.e. take the spread and divide by the 10-year rate at any point in time. This provides a better sense of how wide the spread is in relation to the absolute level of interest rates.
  

The spread is now 96.5% of the 10-year yield (3.40/3.47= 0.965), i.e. a person by putting his/her money in 3-month bills is giving up almost the entire return possible from a longer-term bond. That's quite extreme and unprecedented in the last 45 years (see chart below, click to enlarge). The average over the same period is 25% and the standard deviation is 22%, i.e. the current ratio is in +3σ plus territory.

In a previous post (Bills And Swaps Indicate Great Optimism) I said that such low rates for Treasury bills (plus other indicators such as forward rate swaps) are the result of the Street's optimism. Speculators think the economy is bouncing back, bringing inflation, higher interest rates and lower bond prices with it. Therefore, they are piling into the shortest possible investments in order to shorten portfolio duration and avoid market value losses.

But, we have now demonstrably approached boundary conditions: T-bill rates are essentially zero, meaning that most speculators are 100% certain of the optimistic scenario. The only thing certain being that there are no certainties, I am more comfortable taking the opposite view, i.e. that the economy will weaken again and that the curve will flatten.
_______________________________________________________________
Addition by special request (Greenie): The 10y-3m spread for 1931-1952.

The spread between 1931 and 1952

___________________________________________________________
GDP Addendum
Third quarter GDP came in at +3.5% (real, seasonally adjusted, annualized), as usual "higher than expected". Crunching through the full set of numbers we can immediately distill the whole report to the following:

• Cash for clunkers accounted for 1.7%, i.e. half of the increase.
• Inventory adjustments (lower liquidation of goods in stock) accounted for another 1%.

In other words, 80% of this "growth": (a) came from a temporary government boost that is already gone (see chart below), or (b) was essentially technical in nature.