Long term bond yields rose sharply in the first quarter, reflecting either expectations for faster economic growth, higher inflation, or both. Around half of the 0.8% increase in US 10-year yields appears to be related to inflation if judged by the increase in yields on Treasury Inflation Protected Securities (TIPS).
This move higher in yields occurred in spite of the Fed’s continuing program of $120 billion per month in treasury and agency securities purchases, which act to depress bond yields.
You could be forgiven for thinking what would have happened to yields if the Fed was doing nothing?
And while it’s certainly not a given, if bond yields rise further, is there a point at which they threaten to choke off the economic recovery or threaten to increase the federal government’s debt service costs to an uncomfortable degree? The latter is unlikely in the next few years given debt service costs remain near all time lows, but with federal debt levels rising sharply, yields sustainably above 2.5% across an average of bond maturities would start to raise eyebrows at 1500 Pennsylvania Avenue.
Looking even further ahead (too far perhaps?), some analysts have asked whether the Federal Reserve and the Treasury might need to work together to put a lid on long-term bond yields in a process known as “yield curve control”. It’s a question worth asking even if officials have denied that collaboration between the Fed and the Treasury is on the cards.
Our question is more simple: if the Fed and Treasury did choose to cap bond yields, could they even do so without inadvertently stoking higher inflation?
To stop bond yields rising the central bank would need to buy sufficient quantities of bonds to maintain prices consistent with a particular level of yields. But in the process it would create new bank reserves and potentially new money in circulation. Isn’t that inflationary?
To get some answers we looked at the last time yield curve control was tried in the U.S. during and immediately after World War II.
Turns out, it actually worked, at least for a while.
When the US entered WWII in 1942 the Fed and the Treasury agreed to fix interest rates across different maturities from 3/8ths of 1% for 13-week bills, to 2% for 10-year bonds and 2.5% for longer dated bonds. This yield curve control program allowed the government to finance the war at affordable rates while still attracting sufficient private investment. The program worked as long as inflation did not push investors to seek higher yields than were available under the program, which would have forced the Fed to either allow yields to rise or to purchase the majority of securities issued by the Treasury, which in wartime tends to be a lot. In the process this would create new bank reserves and accelerate inflation.
As it turned out, the Fed had to do very little actual purchases of longer-dated bonds to implement the 2.5% yield cap. With short term rates anchored at 3/8ths of 1% and the market sniffing out the Fed’s commitment to cap longer-dated yields at 2.5%, investors were able to move further out the yield curve and yet face virtually no price risk. Why own lower-yielding short-term paper when long-term debt offered higher yields without the potential for a decline in price?
Rather than having to buy large quantities of long-term bonds to keep their prices from falling, the Fed found itself with exactly the opposite problem – by late 1944 the Fed held substantial amounts of short-term bills (in fact holding nearly three quarters of bills on issue) and its balance sheet had increased substantially, but it had nearly run out of long-term bonds in its open market accounts as private investors snapped these up. The Fed even had to ask the Treasury to issue more 2.5% long-term war loans to satisfy private investor demand!
Problems with yield curve control only started to emerge after the war when the relaxation of price controls allowed inflation to surge. Prices began to rise rapidly in 1946 as demand for American goods grew and price controls were lifted. As measured by the annual rate of change in the consumer price index, inflation was 8.5% in 1946, 14.5% in 1947 and 7.75% in 1948. Initially long-term yields stayed low, either because the rise in prices was perceived as transitory (sound familiar?), or because of a belief that the Federal Reserve would act at some point to restrain inflation.
When the Fed and Treasury allowed the rate on short-term bills to rise, long-term bonds started to weaken as well, as investors sold long-term bonds and bought shorter-term bills. Shorter term rates became more attractive and participants feared that perhaps the 2.5% cap on long-term yields would also be removed. The Fed started to worry what declining long term bond prices would do to the banking system, which was sitting on an enormous pile of longer-duration bonds.
So in 1947 the Fed finally had to start buying quantities of long-term bonds to support prices and enforce the ceiling on yields. But the Fed was able to offset this by simultaneously selling short-term securities such that its overall balance sheet size was unchanged. There was no increase in total bank reserves and no new money created.
For the period of WWII war and its immediate aftermath yield curve control worked.
It wasn’t until the Korean war that things broke down. In 1950 the Korean War broke out and inflation surged in anticipation of possible renewed wartime rationing. Inflation jumped to 17.2% in the first quarter of 1951 with inflation expectations rising rapidly. Once again the Fed was forced to step in to buy increasing quantities of bonds to maintain the yield cap of 2.5%, but this time the Fed was running low on short-term securities to sell. Bank reserves started to expand.
In contrast to WWII, by late 1950 the Fed was caught in a trap where the act of purchasing more long-term bonds threatened to catalyze the process of inflation. In March 1951 the Fed and Treasury signed an accord agreeing to end the direct setting of long-term interest rates, thus recognizing: “the dilemma presented by the conflicting problems of debt management and credit restraint in the inflationary situation which developed” (Federal Reserve Annual Report, 1951, p. 98).
What are the lessons for today’s investors from this 1940’s episode of yield curve control?
Firstly, it actually worked, and for longer than one might have expected. The program was in place for nearly a decade between 1942 and 1951.
Perhaps it worked due to widespread price controls during WWII that prevented the true level of inflation from being reported. Or perhaps it was because investors in the 1940’s were willing to buy lots of U.S. government bonds at 2.5% as part of a national effort to win the war. For whatever reason, the market never really tested the Fed’s 2.5% cap until nearly five years into the program. Would that still be the case today?
The durability of yield curve control also appears to have depended on the Fed’s ability to sell short-term securities to offset its buying of longer-dated bonds. The Fed currently holds around $1 trillion of short-term securities that it could sell and could engage in ‘repo’ transactions (selling, with a promise to buy back) or other forms of innovative ways to ‘borrow’ and withdraw short term money from the market, but there also needs to be willing buyers or lenders on the other side. Recent changes to regulatory capital ratios for banks could prevent them from stepping in to fill the void.
Ultimately there is a fundamental contradiction between the goals of capping yields and controlling inflation. How long that dilemma can be avoided appears to depend on other factors. It worked during WWII but by 1950 the conditions for ongoing success were no longer in place. Much may depend on inflation expectations and on the willingness of American citizens (or others) to finance their government at low rates.