Monday, November 18, 2013

Quantitative Easing myths

The Fed continues to assert that its Quantitative Easing bond purchases will boost economic growth by lowering borrowing costs for businesses and consumers. But the evidence shows that QE bond purchases have actually coincided with increases in long-term interest rates. The Fed is thus guilty of misleading the public; it should be arguing that its QE efforts have been successful because they have boosted long-term interest rates—interest rates are up because the outlook for the economy has improved, and QE has likely contributed to that improvement by satisfying the world's huge demand for safe assets.


The above chart shows the yield on 10-yr Treasuries (blue), overlaid by shaded areas representing periods of Quantitative Easing (darker green) and the one period of Operation Twist (light green). Note that yields rose on net during each period of Quantitative Easing, and only fell when the Fed was NOT purchasing bonds. Operation Twist also failed to deliver on its promise, since 10-yr yields were roughly unchanged despite purchases of $400 billion longer term bonds and corresponding sales of shorter term bonds.


The above chart compares the yield on 10-yr Treasuries (red line) with the percent of marketable Treasuries held by the Fed as a result of its QE purchases (blue line). Note that the Fed today holds about 18% of marketable Treasuries, which is almost identical to what it held in the years leading up to QE, yet yields are substantially lower—there's no discernible correlation between the amount of bonds the Fed owns and the level of yields. Note also that over the past four years the Fed has significantly boosted not only its purchases of Treasuries but also the portion of marketable Treasuries it holds, yet interest rates are largely unchanged on net. Note further that 10-yr Treasury yields have jumped over 100 bps since early May, even as the Fed has stepped up its purchases and its percentage ownership of marketable Treasuries. There is no evidence in either of the two charts above to support the notion that QE bond purchases have inflated bond prices or depressed bond yields.


The same arguments apply to the Fed's purchases of mortgage-backed securities. As the chart above shows, the Fed has dramatically increased its ownership of MBS over the past several years, yet there is no indication that increased purchases have had a depressing effect on mortgage yields. Over the past year, during which the Fed has purchased $40 billion of MBS per month, mortgage yields have actually increased by about 100 bps.

One explanation for why the Fed's QE efforts have not produced their promised results is relatively simple. The Fed has purchased only a relatively small portion of the outstanding supply of Treasuries and MBS, and that is not enough to materially change the yield on all outstanding bonds. The outstanding supply of marketable Treasuries and MBS is almost $21 trillion, while the Fed owns only about $3.6 trillion. More important is the fact that there are tens of trillions of corporate and non-U.S. bonds that are effectively priced off of Treasuries. To artificially depress the yield on Treasuries the Fed would have to not only buy a huge portion of outstanding Treasuries but also a significant portion of corporate and non-U.S. bonds.

It's also a myth that the Fed has been "printing money" with abandon as a result of its QE bond purchases. This myth persists, despite strong evidence to the contrary (i.e., the fact that inflation remains very low despite four years of massive QE purchases), because the majority of observers fail to understand the mechanics of Quantitative Easing. QE bond purchases do not create money; they create bank reserves, which are very different from money. Bank reserves can't be spent anywhere, because they exist only on the Fed's balance sheet. Banks don't lend bank reserves, they use them to collateralize their deposits, which in turn are a function of total lending activity. The huge expansion of bank reserves could potentially result in a huge expansion of the money supply (banks need extra reserves to support increased lending), but it has not, because a) banks have been reluctant to expand their lending activities, and b) the public has been reluctant to borrow more. Not only banks, but the entire world remains relatively risk-averse, preferring to hold significantly more cash and cash equivalents, of which bank reserves are an important part.

12 comments:

  1. Ed Yardini: Bond Bubble

    I see more bubble-like conditions than Janet Yellen does. Her lack of concern may be justified currently, but by expressing it she increases the odds of triggering melt-ups in asset markets, especially if she turns out to be even more dovish than Bernanke, as I expect. Consider the following:

    (1) Emerging market debt. The result of the widespread “reach for yield” is that debt sales by emerging markets rose to $439 billion this year through October, within reach of last year’s record of $488 billion. According to the 11/6 FT article on this subject, “record debt sales from the developing world have rebounded after a summer of turmoil and the US budget crisis stunted demand, leading analysts and bankers to predict the second record year of bond issuance in a row.” Tapering chatter during the summer nearly burst the bubble in emerging market debt. Since the 9/18 decision not to taper, more air has been pumped into it.

    (2) Corporate bonds. The Fed’s Flow of Funds data show that the outstanding amount of corporate bonds issued by non-financial corporations rose to a record $6.1 trillion at the end of Q2-2013. That’s up by a record $625 billion y/y. That may not seem like a bubble since corporations may be using some of the proceeds to reduce their short-term debts, and have lots of liquid assets. However, corporations also are likely to be using some of their bond proceeds to buy back shares, which artificially inflates earnings per share. That’s a bubble-like development if stock prices are getting a significant lift from debt-financed share buybacks rather than actual earnings growth.

    http://blog.yardeni.com/

    ReplyDelete
  2. Can you post that old chart of the 10 year yield and what it says about the economy? 4% normal growth sub 3% reccession sub 2% depression...

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete
  4. To me the more plausible rationale - the market is front-running the Fed and selling into their purchases. Therefore, you would expect yields to rise as they embark on each round of QE.

    ReplyDelete
  5. scott, totally agree. which begs the question;why does the fed persist in behavior that doesn't help? seems like the more we get QE, the more the market assumes a weak economy and the lower the 10 yr trades in yield. but this also puts the brakes on economic investment as business assumes same and holds back on hiring. if the fed tapered/stopped QE, the markets would assume opposite which would spur investment. this ain't rocket science. I contend that the fed continues QE for POWER. central bankers want to be players and by taking a backseat to the economy, they lose power.

    ReplyDelete
  6. In this post:

    http://scottgrannis.blogspot.com/2013/10/bank-lending-to-business-picks-up.html

    You spoke about the pick-up in bank lending to business. So is the adverse effect from the individual not borrowing? Or banks just not being able to keep up with QE and lend at the same pace?

    Seems to me that the Fed will have to print (a lot) if losses are consumed in their portfolio. But, we don't get to see the actual numbers thanks to their public/private status and being able to keep assets reported at cost rather than FMV. The problem being their liabilities (all the reserves ballooning) have no duration, while their assets (all the MBS and treasuries bought at $85B/month) have about a 7 year duration.

    ReplyDelete
  7. I was trying to tell this to Mr Grannis, but he refused to take any stock in because Hans is a radical nutcase..

    If you place 100% reliance in governmental units facts and figures
    then you are subject to undermining your competency..

    http://blogs.marketwatch.com/capitolreport/2013/11/19/u-s-unemployment-rate-faked-n-y-post-critic-charges/

    ReplyDelete
  8. Outstanding analysis, not available from any of conventional media. Thank you

    ReplyDelete
  9. RE: borrowing. Bank lending to small and medium sized businesses has indeed picked up, but it has not yet surpassed the level of 2008. Many households and many businesses continue to deleverage, but deleveraging is fading away. In aggregate, I observe that there has not been any unusual degree of credit creation or money supply expansion in aggregate. I think this shows that the economy in aggregate is still somewhat risk averse.

    The Fed is essentially forced to take on risk to balance the risk aversion of the private sector. This will not go on forever, however, and I think we are in the early stages of the unwinding of QE and the return of confidence.

    ReplyDelete
  10. Hans: I have always cautioned that government statistics are inferior to market-based indicators. Government stats are subject to revision, faulty seasonal adjustment, and delays. It is very disturbing to think that the unemployment numbers may have been manipulated for a few months in 2012 for political purposes. I hope this is not the case, but if there had been manipulation it would have been reversed by now.

    In any event, I would never want to have total confidence in any of the employment related numbers, if only because they are almost always revised significantly after the fact and can be very volatile from month to month.

    ReplyDelete
  11. Radical Nutcase:

    It seems Jack Welch was wondering the same thing way back in October, 2012 ....

    Was Jack Welch Right: Jobs Numbers Under Fire

    ReplyDelete
  12. The Fed owned 20 percent of the Treasuries market in 2003? Really? Why?

    ReplyDelete