Tuesday, May 30, 2017

Treasury and Fed Funds Rate

The Federal Reserve seems determined to raise the overnight federal funds rate it controls. Many believe that two or three more 25 basis-point increases this year are all but certain, unless the economy suddenly weakens appreciably or a major financial or political crisis unfolds here or abroad.

The prospect of steady rate increases is a worry to many market participants with the taper tantrum of May and June 2013 still a vivid memory. Then-Fed Chairman Ben Bernanke only hinted at a slowing of Fed security purchases, then running at $85 billion per month, but yields on Treasury notes and bonds rose sharply higher nonetheless. In fact, the Fed didn’t taper until December of that year and finally discontinued adding to its huge portfolio in October 2014.

Interestingly, the taper tantrum may have prepared markets for that eventuality since no cataclysmic actions ensued in the market for Treasuries. Also, stocks, which were propelled by Fed largess since their March 2009 bottom, merely leveled and didn’t sell off, even as the central bank stopped adding more fuel to the equity bonfire. The Fed did, however, pledge to maintain the size of its asset horde by using the proceeds of maturing securities to buy more bonds.

The question now is: How much effect do changes in the fed funds rate have on longer-term Treasury yields? This question assumes that the Fed is the prime mover and Treasury markets follow. But you can argue that causality runs the other way. If investors fear an overheating economy and resulting inflation, they probably would sell Treasuries, pushing bond yields higher. That could worry the central bank, which then delays raising short-term rates. Furthermore, the Fed, along with other major central banks, reduced overnight rates to essentially zero in reaction to the financial crisis and held them there until recent years. So, the Fed was essentially on the sidelines in influencing long-term rates.

As I’ve often said, causality can’t be proven with statistics. After beating a drum, a solar eclipse goes away; that’s 100 percent correlation, but not causality. Still, it’s probable that the causality runs from changes in the fed funds rate to Treasury yields. My firm’s research shows that the highest correlation between the fed funds rate and 10-year Treasury note yields is with the central bank rate leading the note yield by 10 months.

The fed funds rate and yields on 10-year and 30-year Treasuries all increased with inflation from the end of World War II to the early 1980s and then fell with disinflation. So, correlating the raw data would imply stronger relationships than existed. To measure the true effects of fed funds rate changes on long-term Treasury yields, we statistically removed the up and down trends and correlated the deviations that were left.

The statistical fits appear meaningful, though they are much lower than if the trends had not been removed. But again, good statistical fits over time may have no causal relationships. Global steel and meat consumption correlate highly over time, but only because both grow with economic development.

A 100 basis-point increase in the fed fund rate, using our analysis, results in a 46 basis-point increase in 10-year Treasury yields, on average. In other words, about half the change in the fed fund rate spills over to Treasury yields. So the influence is far less than one-to-one. As you’d expect, the spillover from fed funds changes to 30-year Treasury yields is even smaller. A 100 basis-point rise in the fed funds rate results in only a 35 basis-point rise in yields.

The statistical results reflect the entire post-World War II era, but the influence of changes in the fed funds rate on Treasury yields may be even less today. Thirty-year yields are lower, not higher, than when the Fed first raised its reference rate by 25 basis points in December 2015 and then made further quarter-point increases in December 2016 and in March of this year.

One reason for this weaker response may be the Fed’s use of forward guidance to advertise its intentions regarding the future path of rates. The spillover effect on Treasuries may already be in place. Another may be the lack of effectiveness of the fed funds rate in recent years. It’s the rate that banks with excess reserves at the Fed lend to those with deficient reserves. But with the huge amount of excess reserves -- about $2.7 trillion at present -- that have been created by the Fed’s bailout of Wall Street and then quantitative easing, most banks have ample reserves and few need to borrow from other banks in order to meet their reserve requirements.

So the fed funds rate is simply a reference rate with little practical meaning. Market participants believe it measures Fed policy, but the emperor may have no clothes. That’s why the central bank has considered alternatives such as paying enough on bank reserves, currently 1 percent, to make them attractive for banks to hold. That also, in effect, would sterilize those excess reserves because lending to the Fed would, risk-adjusted, be more rewarding than lending them to private borrowers. 

Also, sterilizing excess reserves would remove the likelihood that when the economy resumes rapid growth, which I expect in two or three years, those reserves would be lent and re-lent to eager borrowers. The result could be an explosion of the money supply and velocity, which might propel excess economic demand and inflation. I already look for robust economic growth of 3 percent to 3.5 percent, to be driven by massive fiscal stimuli.

Alternatively, the central bank is thinking about letting its $4.5 trillion portfolio run down by not reinvesting maturing securities. To do so would require great skill and uncommon good luck to avoid squeezing or frightening lenders and borrowers to the point that a recession follows. Indeed, key Fed officials have suggested that the Fed may not return to its $1 trillion pre-crisis portfolio but perhaps keep it at $2 trillion. So it still would have massive excess reserves to deal with.

via bloombergquint