David R. Kotok co-founded Cumberland Advisors in 1973 and has been its Chief Investment Officer since inception. He holds a B.S. in Economics from The Wharton School of the University of Pennsylvania, an M.S. in Organizational Dynamics from The School of Arts and Sciences at the University of Pennsylvania, and a Masters in Philosophy from the University of Pennsylvania. Mr. Kotok’s articles and financial market commentary have appeared in The New York Times, The Wall Street Journal, Barron’s, and other publications. He is a frequent contributor to CNBC programs. Mr. Kotok is also a member of the National Business Economics Issues Council (NBEIC), the National Association for Business Economics (NABE), the Philadelphia Council for Business Economics (PCBE), and the Philadelphia Financial Economists Group (PFEG).
~~~~~
Some Observations about Spot Interest Rates and Forward Interest Rates: with help from Jason Benderly, Jim Bianco, Ned Davis & Howard Simons
October 18, 2009
“What going to happen to interest rates when the inflation comes?” This is a recurring question in our quarterly client review meetings.
In a normal cycle one can make some reasonable projections about the changes in interest rates when the economy bottoms. The usual sequence is that the Fed first allows the economic recovery to gain traction and then eventually starts to tighten policy by raising the short-term interest rate. Other rates also rise, first in anticipation of Fed action and then as the Fed persists. At some point the Fed reaches a level which slows the inflation tendency of the economy. The yield curve flattens and longer-term rates stop rising, even as short-term rates continue to do so. In extreme cases the short-term rate is pushed above the long-term rate.
We are not in a normal cycle.
The operation of interest rates and Fed policy is quite different this time, as the Fed is engaged in quantitative easing; there is no serious inflation, there is huge federal debt issuance and the policy-prescribed interest rate is effectively near zero. Traditional dynamics of monetary policy don’t work. There are many reasons why this is true, and we will discuss them in future Commentaries. Japan is an example of how this zero-rate status with no inflation and huge deficits can persist for a very long time.
The reason traditional policy may not fully apply to the United States, is because it is the world’s biggest provider of a reserve currency. But some of the reasons do apply, in part, and are at work today. That is why the long-term Treasury yield remains quite low and why those who keep sounding warnings about much higher interest rates have been frustrated by the markets.
A zero short-term policy rate is anchoring the entire Treasury yield curve. Benderly Economics recently reviewed the last 50 years of history and noted that “10 year yields have never moved more than 350 basis points above the Fed Funds rate.” Today the Fed Funds rate is targeted between zero and 25 basis points (one quarter of one percent). True to history, the 10-yr Treasury yield seems contained to an upper rate of 3.75%. That level has defined the top of a trading range; the bottom is defined as much lower.
Large deficits do not seem to matter much to the market’s setting of interest rates. This is and has been true in Japan and seems to be the case in the US. We all know that the government is borrowing trillions and will likely continue to do so for years and years. In spite of this information, the Treasury bond rate remains low.
Of course, a 3.75% 10-year rate and a zero short-term rate mean the yield curve is very steep. A steep curve allows the projection of forward rates, which quickly come into play in bond portfolio management. One of the great pieces of research work on forward rates and forward rate ratios is done by Howard Simons and Jim Bianco. We pay close attention to their commentaries every time they release them.
Forward rates allow a portfolio manager to make choices. Here is a simple example. Let’s say the very short-term rate is zero; the one-year rate is 1%, and the two-year rate is 2%. An investor can have absolute liquidity and earn nothing. Or she can tie up her money for one year and earn 1%. If she chooses the 2% rate for two years, she is implicitly betting that the one-year rate will be 3% or lower one year from now. Let’s ignore compounding and use just simple annual rates for this example; we will explain how we reach the conclusion.
The investor with the two-year time horizon can be guaranteed the 2% rate for the entire two years right now. If he chooses the one-year 1% rate instead, he is now betting that the second year will yield a higher rate, such that his total over the two years exceeds what he can get as a guaranteed result right now. In this case a 1% for the first year and a 3% for the second year would equal the 2% for the entire period of two years.
The investor who thought that rates in year two would be much higher than 3% would choose the one-year option. The investor who thought the second-year rate would be lower than 3% would choose the two-year option. The investor who didn’t know what to do would choose the overnight rate of zero and defer the decision. Change the numbers slightly and you have the decision tree that every investor is wrestling with today.
The same logic applies when you compare the 10-year rate with the 5-year rate or many other combinations of forward rates. The process of forward rate adjustment is dynamic and ongoing which is why the movement of any rate impacts all rates. It is the forward rates that function to limit the steepness of the yield curve in times when the central bank’s policy is setting the shortest rate at zero.
There are many models used to forecast what the 10-year Treasury yield SHOULD BE. Each has strengths and weaknesses. Most practitioners have their favorites. But all of us know that no model works perfectly. If there were a perfect model and I knew it, I wouldn’t be writing this Commentary or sweating out these portfolio-management decisions.
A review of most of the models would suggest that the 10-year Treasury yield SHOULD BE about 3% on the low end, if the models are based on assumptions of some continuing economic weakness. The upper level of the SHOULD BE range is about 4.5%, based upon assumptions that the economy will be recovering on a sustained path and that some little bits of inflation will eventually emerge. If you average out the various models, the 10-year Treasury yield should be between 3.25% and 4%, or right about where it is now at 3.4%.
We like Ned Davis’ model, which uses four factors and creates a “fair value” of 4.27% today. This model allows the user to modify the assumptions about the direction of the factors. Ned admits that the present trends suggest his fair-value computation could soon be deriving a lower yield. That is our view, too. Ned’s model is not widely known, so it doesn’t suffer from a Goodhart’s Law effect. And it is tested by a nearly a half-century of data and has worked well in both high- and low-inflation-rate regimes and in both high- and low-interest-rate environments.
We get similar results when we apply our own internal and proprietary models. They are based on inflation-adjusted holding returns and use year-over-year computations to eliminate seasonality.
One cause of concern was recently noted by Howard Simons as he discussed China’s currency peg to the US dollar and how a floating yuan would cause US rates to face large upward pressure. Howard is on to something and demonstrates it well by comparing the euro vs. the dollar and how the China policy works with each one.
To this point, at Cumberland we do not expect the Chinese to engage in any policy that is not in their longer-term self-interest. Orderly markets and economic growth by selling stuff to the world are at the top of their agenda. China employs 50 million people making low-cost items to sell to Americans and Europeans. As long as that drives their policy they will continue to manage their currency such that they stay competitive as to price. That means their reserves will continue to grow, because they will have to keep managing their currency exchange rate against the dollar.
At Cumberland we have maintained long duration in our bond accounts for a considerable period, and clients have benefitted by this approach. We emphasized “spread product” over Treasury bonds and watched spreads narrow to the benefit of clients. We continue to do that today.
We expect to gradually reposition duration to a more neutral point over the coming months. We are not ready to go to very short duration. We do not see rates abruptly going higher. That is out in the future and not at hand today. We cannot say if significantly higher rates are a year away or three years away or five years away but we can say they are not destined to arrive real soon. For now, the Fed will remain at zero and forward rates will continue their role as anchorage of the entire yield curve. For the present time our duration shift in portfolios will be gradual.
One duration target change is working its way into the strategy and portfolios. We believe it is the beginning of the time period to start moving from long to neutral. It is way too soon to get fully defensive in fear of abruptly rising rates. It still pays to select certain spread product over Treasury bonds. And in some cases it makes sense to hedge bond portfolios to dampen the volatility.
We will make changes in this strategy in response to any event-driven input or when our preferred models suggest it is time to do so. For now we expect the short-term rate to remain near zero for a while longer and the 10-year Treasury yield to be limited by the upper bounds we discussed earlier in this Commentary.
David R. Kotok, Chairman and Chief Investment Officer, email: david.kotok@cumber.com
What's been said:
Discussions found on the web: