David R. Kotok
Chairman and Chief Investment Officer
The Fed and Treasury Volatility
September 8, 2010
Stewart Taylor runs a specialized fund for Eaton Vance. He has seriously studied volatility in US Treasury securities for decades. He also joins the annual August gathering at Leen’s Lodge and came back for a long Labor Day weekend. We talked.
We dissected market’s response to the Fed’s post-Lehman balance-sheet moves. First, we saw an explosion of new programs. That morphed into the $1.25 trillion in Fannie and Freddie holdings, which took the Fed’s balance sheet to about $2.4 trillion. Now the Fed is receiving the payments on those mortgages and is redeploying the cash flows into treasuries.
Stew and I exchanged many questions. A glass of wine on the veranda at Leen’s as the sunsets unfold allows for serious conversation in a relaxed setting. Some friendly smallmouth bass set the stage earlier in the day.
How can we estimate the impact of the Fed’s $1.25 trillion in purchases of mortgages? What are the security price effects of the transition from mortgages to treasuries as the Fed maintains the size of the balance sheet? How do we deal with the conflicting factors like flight to quality in treasuries or worries about the financing of the deficits? Is there any way to have a segregated and isolated test on treasury yields? So many competing factors influence them.
BCA Research has tried to quantify these issues. In a recent report they offered that “each $100 billion purchase of agency or MBS securities — or sale of Treasury securities — causes a narrowing in MBS spread of about 8 basis points.” The spread between mortgage rates and treasury yields was narrowest coincident with the ending of the Fed’s mortgage purchase program. The spread has been widening since. BCA notes that the widening spreads have “offset the rally in Treasury yields.” In other words, the Fed lowered Treasury yields but caused the mortgage yields to rise when it stopped supporting that targeted sector. Thus, recent Fed activity has not helped housing, since home mortgage rates did not fall.
How about the overall increase in the size of the Fed’s balance sheet? Can we estimate an impact from that? Does size matter?
BCA says yes. By examining the term premium, they estimate the impact of an increase in the Fed’s asset holdings. They argue it is about 7 basis points for every $100 billion in size increase. This approximates the Fed’s effort to measure the same effect. BCA found that the market adjusts rapidly to a change in the size. In fact, they have been able to measure the effect and find it occurs at the time of the Fed’s policy announcement. Clearly, markets believe the Fed when it says it is going to do something.
What about the expectations component we have written about in the past? Here, too, the BCA research paper sheds some light. After the release of the August 10 FOMC minutes, BCA estimates that the “term premium” dropped by 6 basis points. It continued to shrink another 15 basis points while narrowing. BCA suggests, “one interpretation is that the market has fully discounted a further $300 billion in asset purchases.” Another way to look at this, they say, is that there is “a 30% chance of another $1 trillion in balance sheet expansion.” Their conclusion is that a commitment by the Fed for a balance-sheet size of $3.3 trillion instead of $2.3 trillion would cause 10-year treasury yields to decline by another 50 basis points. One could also infer that a commitment to buy more mortgages would promptly lower mortgage rates, trigger more refis, and give an additional boost to housing’s tepid and questionable recovery.
Is there evidence of a yield reversal that supports these numbers? Stew and I speculated with many anecdotes. Consider how 10-year Treasury note yields jumped 14 basis points immediately after the positive ISM news hit the wires. On-the-run 10-year notes gapped from 2.62% to 2.76%. Why? Because the markets were repricing the probability of the Fed expanding the balance sheet. Markets then lowered yields based on expectations, and the amount seemed to be in line with BCA estimates.
Competing influences make this attempt to measure a Fed policy in terms of basis points very difficult. We applaud the work of BCA for attempting to do so and for using some of the Fed’s own research to independently support their conclusions.
What will the future show about this methodology? We speculate that it is valid now and will continue to be until Goodhart’s Law comes into effect. Goodhart’s Law is now part of the Fed’s dilemma. How long can the policy of large balance-sheet sustainability be efficacious? What happens when the markets fail to respond robustly to the next action? The latest employment statistics suggest that this test may be forthcoming.
For investors, that means the present forces will continue to operate for a while. Our stock accounts are fully invested using ETFs. We expect higher stock prices over the next year and as long as the global policy interest rates remain near zero.
Our bond accounts continue to emphasize longer duration. We are starting to increase the hedge exposure in selected accounts where a market-neutral position is desired. That remains at the early stage. Our goal is to construct a market-neutral position to take advantage of the spreads while dampening risk from parallel yield-curve shifts.
Labor Day already seems like ancient history. Stew and I will go back to email to compare notes. At Leen’s Lodge, the bass are feeding for the next few weeks and then will go into winter hibernation. Unlike the market’s lessons under Charles Goodhart’s tutelage, the bass follow a predictable annual pattern. It does not change as the Fed develops extraordinary policies for an extraordinary time.
David R. Kotok, Chairman & Chief Investment Officer, Cumberland Advisors, www.cumber.com