There is a virtual universal belief that the Fed is manipulating not only Treasury yields but also risk asset prices through its QE bond-buying program. However there is no empirical evidence that can prove Fed purchases reduce long-term interest rates or raise risk asset prices. There is only correlation. The Fed is buying bonds and bond yields are falling, therefore the Fed’s purchases of bonds is lowering bond yields. The Fed is buying bonds and stock prices are rising, therefore the Fed’s bond buying is raising stock prices. But correlation does not imply causation.
For the Fed to be responsible for lowering long-term interest rates, its purchases would need to represent demand in excess of what otherwise would be present relative to the supply of bonds at various levels of interest rates. In order to push rates lower, the Fed must be the buyer willing to absorb a larger percent of supply at ever lower prices as natural economic buyers exit the market.
To understand how this supply/demand balance works at various levels of interest rates, I look at the Fed’s own Flow of Funds
data. The Flow of Funds is essentially the balance sheet of the US capital market. The liabilities represent the issuers of credit while the assets represent the buyers of credit, and because one entity’s liability is another entity’s asset, these numbers net out to zero. The price is set where this supply and demand finds equilibrium. In theory, QE pushes this equilibrium below its natural price.
At the end of Q2 2013 there was a total of $57.562 trillion in outstanding liabilities of which $40.938 trillion was issued by non-financial entities, including the Federal government. Since 2008, total non-financial credit supply increased $7.27 trillion and the demand for this net new credit supply has been dominated by three main sources: the Federal Reserve, foreigners (the rest of the world), and mutual funds that, in aggregate, represent $6.96 trillion in demand for the increase in supply. The Fed bought $2.287 trillion, or 31.5%, of the increase. Foreigners bought $2.37 trillion, or 32.6%, of the increase. Mutual funds bought $2.3 trillion, or 31.7%, increase. Is one of these 30% buyers more responsible for setting the market price?
For purposes of this analysis, I want to focus on credit demand and the direction of interest rates over three periods: 2008-2010, 2010-2012, and Q2 2012-Q2 2013.
Between 2008 and 2010, due to the massive deleveraging of financial balance sheets, there was only a $293 billion in net credit growth. However, $3.02 trillion was in US Treasury supply, representing a whopping 47.5% growth from the 2008 levels. Over 98% of this supply was absorbed by the Fed, foreigners, and mutual funds. The Fed, which more than doubled the size of its balance sheet, accumulated $1.27 trillion in credit. Foreigners bought $968 billion and mutual funds bought $740 billion. The 10-year Treasury yield would begin this period at around 2.20% at the end of 2008 and end 2010 at 3.30%.
In the period between 2010 and 2012, as the credit markets thawed, issuance would pick up with a net supply of $3.1 trillion, of which $2.2 trillion was Treasury supply and $1.2 trillion was from non-financial corporations. Because QE2 ended in June 2011 and the balance-sheet-neutral Operation Twist dominated the Fed’s activity, it only added a net $410 billion in the period (QE2 ended June 2011). Mutual funds added $1.0 trillion; foreigners added $1.12 trillion. The 10-year Treasury yield would begin the period near 3.30% and end 2010 just above its all-time low at 1.75%.
In the period between Q2 2012 and Q2 2013, total credit supply increased by $1.97 trillion. This total included $862 billion in US Treasury issuance and $785 billion from non-financial corporations. Demand for this credit would come from the Fed’s QE3 program launched in the third quarter, which added $603 billion. Mutual funds added $551 billion followed by foreigners who added only $287 billion (banks added $288 billion). The 10-year Teasury yield would begin this period at its 1.65% lows and end the period at 2.50%.
The bond market is unique in that, unlike the stock market, there are uneconomic participants who buy regardless of price. Of the three main sources of credit demand since the financial crisis, mutual funds would be considered economic buyers; the Fed and foreigners are largely uneconomic.
Credit Market Demand
Click to enlarge
In 2007, mutual fund demand accounted for 6.73% of total net non-financial credit supply, and mutual funds were consistent owners of ~6.5% of non-financial credit supply prior to this period. In 2008, there was a clear change in trend in mutual fund demand for credit assets. As the crisis unfolded, mutual fund demand increased, and by the end of 2010, these investors were absorbing 8.0% of non-financial credit. In the latest quarter (Q2 2013), mutual funds were buying 10.4% of the net supply of non-financial credit.
The group that you would think to be the most price-sensitive was a consistent buyer regardless of price. In each period studied, mutual fund demand accounted for 31.5%, 31.2%, and 32.86% of the increase in non-financial credit supply for the 2008-2010, 2010-2012, and Q2 2012-Q2 2013 period respectively.
For the Fed to be manipulating bond yields below their natural prices, you would have to assume it was buying bonds that were uneconomic for natural buyers. However, despite these apparently uneconomic prices, credit assets continued to find economic buyers willing to increase their exposure.
One way to rationalize this increased appetite for credit assets by economic buyers is that credit assets are closer to their natural prices than what many assume. Long term interest rates tend to discount long term growth and inflation, or the average growth rate of nominal GDP. The growth rate for nominal GDP is 3.0%. The growth rate for personal consumption, the largest component of NGDP, is 2.8%. The growth rate for corporate earnings, which grow at NGDP, is 2.9%. Maybe the natural rate of interest is 3.0%?
NGDP Vs. PCE and 10-Year
While economic mutual funds were consistently adding exposure despite price, seemingly uneconomic foreigners seemed to be more price-sensitive. Presumably, both the Fed and foreign buyers are uneconomical in that these buyers are swapping dollar reserves back into dollar credit assets. However while foreigners may not be asset price-sensitive, they do seem to be exchange rate price-sensitive.
JPY Vs. 5-Year
Juxtapose the period between the four quarters ended Q2 2012 when the dollar was falling with the four quarters ended Q2 2013 when the dollar was rising. In the four quarters ended Q2 2012, Treasury supply increased $1.3 trillion. Foreigners bought $456 billion and the Fed saw a net reduction in assets by $89.4 billion (mutual funds bought $498 billion). In the four quarters ended in Q2 2013, Treasury supply increased $862 billion. Of this new supply, Fed demand accounted for $603 billion with foreign demand accounting for $287 billion. However in the last two quarters, the imbalance is even more pronounced. Through the June quarter, the Fed has purchased a total of $544 billion in total credit while foreigners were net sellers of $89 billion.
In periods when the Fed’s increase in demand represented the largest percent of the increase in net new supply, interest rates were rising. In the periods when the increase in foreign demand represented the largest percent of the increase in supply, interest rates were falling. This poses the question: Is US credit demand dominated by the Fed’s QE purchases, or is it dominated by the foreign purchasing power of the US dollar?
Back on June 17 in The QE Carry Trade Is Imploding Right Under Everyone’s Noses,
I provided my assessment of the mechanics of how QE impacts market prices:
QE does not influence markets via pumping money into the bond market, which in turns inflates asset prices. QE expands the monetary base, which in turn increases the inflation premium in the yield curve. This inflation premium manifests itself in lower, and in recent cases, negative real interest rates, which in turn raises the present value of cash flows from risk assets.
The spread between nominal Treasury yields and TIPS yields, a.k.a. the breakeven spread, represents the bond market’s inflation premium for the given term structure. This spread is extremely important because of the discount it represents.
In effect nominal yields are being pushed around by real yields that are a function of the QE inflation premium. This suggests that the QE trade is not a function flow but rather a function discount. This is a major distinction that is key to understanding what is going on today as real interest rates rise.
The Fed’s influence is more psychological than physical. The Fed is targeting real interest rates, which are a function of inflation discounts, which are a function of participants' expectations. The value of the US dollar is also a function of inflation discounts, which is a function of participants' expectations. If there are more dollars in circulation than demand for those dollars, then the excess liquidity increases inflation expectations, which reduce the value of the dollar.
As the Fed exits QE, the bond market is going to be looking to find natural equilibrium between price-sensitive credit buyers and price-sensitive exchange rate buyers. As the Fed reduces balance sheet expansion, the growth in excess liquidity will wane and the value of the exchange rate should rise, reducing demand from foreign investors. This should also see inflation premium fall, raising real interest rates. Rising real rates should slow nominal growth rates, which will lower the natural rate of nominal interest, reducing the risk of economic investors and increasing the demand. At what rate this finds equilibrium will depend on expectations for growth and inflation, which today could conceivably be at a 2% real rate, a 1% inflation rate, and a 3% natural nominal rate.
The Fed doesn’t set the price of money, the market does. Interest rates are low because the demand for money is weak and the demand for credit is strong. As the demand for money adjusts upward or downward, so, too, will the price of money and the participants who are willing to pay that price.
As we approach year-end, the story of 2013 is about the prospect of exiting QE and rising long term interest rates. The 10Y-year is up over 100bps from the lows reached this time last year with most of the move occurring while the Fed was increasing the rate of purchases while foreign investors substantially reduced exposure. The fundamental conditions behind this move were not economic in nature but rather a reduction in inflation expectations driving a rise in real interest rates and a rally in the value of the US dollar. This is a critical distinction to understand how rates will behave in a post-QE market.
What is having a bigger impact, the flow of the Fed’s purchases or the inflation discount expected by market participants? I suspect when the postmortem is written, investors will come to realize it is the inflation discount and value of the currency that had the bigger impact.