Wednesday, December 11, 2013

QE, Rising Yields, and the Right Way to Taper

Matthew O'Brien of The Atlantic gives us the 41 most important economic charts of 2013. The figures come from various contributors, including me. My figure is borrowed from chief economist Michael T. Darda of MKM Capital and shows that long-term treasury yields generally have risen under the QE programs. This pattern runs contrary to the stated objectives of the Fed and is also inconsistent with those who claim the Fed's large scale asset purchases (LSAPs) are draining the financial system of safe assets. Fortunately, there is a way to make sense of these developments.

The chart I borrowed from Darda came from a longer note that had other interesting charts and comments. I thought you might like seeing the rest of them, especially his thoughts on how to taper without tightening monetary policy. Here is Darda:
Most observers continue to (falsely) believe that QE works by lowering long rates and flattening the yield curve. However, a quick look at the data suggests this is not the case.


Others believe that QE is akin to “debasing the dollar,” but this too is false. The dollar has been stable at a low level for the last five years. Commodity prices haven’t responded to QE in a consistent manner because China’s growth trajectory, not QE, the dollar or developed country nominal GDP growth, has been the primary driver of commodity price trends in recent years.


We find many of the criticisms of the Fed/QE to be totally off-base. To wit: allowing the financial system to collapse in 2009 (instead of initiating QE1), allowing deflation to emerge in 2010 (instead of implementing QE2) or having the full force of the fiscal cliff/sequester hit in 2013 (instead adopting QE3) are not persuasive alternatives to QE/forward guidance, in our view.

The Fed has had a meaningful effect on financial conditions and has placed a floor under inflation expectations. Thus, the U.S. has avoided double dip recession and deflation despite the stiffest fiscal headwinds in six decades (unlike Europe in 2011 and the U.S.A in 1937 when premature monetary tightening capsized business cycles before full recovery was at hand and unlike Japan during the 1990s and 2000s when a too-tight BoJ blithely presided over monetary deflation).




Concerns about financial stability/over leverage are vastly overblown, in our view. Household leverage ratios have collapsed over the last five years even though the Q3 flow of funds data will likely show a new-all time high in inflation-adjusted net worth, thanks to rising nominal asset values and moderate nominal income growth (relative to flat or falling debt levels). Moreover, there is no evidence of a broad-based leverage or credit problem, unlike the early 2000s. The broadest measures of money (that include the shadow banking system) are growing at about the pace of NGDP. With unemployment/underemployment/long-term unemployment still a problem and inflation near record lows, the Fed is correct to focus on growth.



With inflation ~100 bps below the Fed’s target, and forward-looking inflation breakeven spreads below 200 bps at the five year horizon, the FOMC has to be careful not to send an inadvertently hawkish message if/as begins to taper asset purchases. One way to accomplish this would be to use forward guidance on the monetary base, promising that previous additions to the stock of high powered money will be permanent at least up until the point at which the Fed’s 2% inflation target its met (or is expected to be met). The same exercise could be done for a NGDP or price level path target. If this was done and viewed as credible, the demand for base money would ease and the Fed could end QE without endangering the recovery/failing on its inflation target. To some degree, this is already occurring. We note that even though long bond yields have risen in recent weeks, two year note yields are firmly anchored, meaning the market believes the Fed will not hike short rates over the next two years despite what appears to be a high likelihood that tapering will begin sometime between December and March. A steepening yield curve and easing financial conditions are consistent with a pickup in growth (which is what the Fed wants).






I find his proposal to do forward guidance on the monetary base rather than interest rate targets to be interesting. I look forward to more discussion on it.

9 comments:

  1. David,
    Does QE work through a "stock" or "flow" effect? The QE vs. yields chart implies a positive correlation with "flow".

    ReplyDelete
  2. This paper below says there are both stock and flow effects, with the latter being more modest.
    http://www.federalreserve.gov/pubs/feds/2010/201052/201052pap.pdf

    My sense is that most observers, including Bernanke, see the stock effect as being the most important one. So yes, these figures do create some interesting questions about the relative importance of the stock vs flow effect of LSAPs.

    ReplyDelete
    Replies
    1. David,
      I think you are misunderstanding Diego Espinosa's question. D'Amico and King look at the stock and flow effect of QE on the liquidity premium, and both are obviously negative. The QE vs. yields chart implies a positive correlation which obviously reflects an increase in the average of the short term interest rates that are expected to occur over the life of the bond. This is equivalent to the observation that 10-year T-Note yields are strongly correlated to expected NGDP:

      http://3.bp.blogspot.com/-fQOh2jjNS_8/T2kdMyGHxAI/AAAAAAAACXI/y5k6QONVFGM/s1600/treasyield_NGDP.jpg

      Delete
  3. Its not so much the relative importance, but the different signs on the correlations. Yields seem to have a strong negative correlation with stock, and strong positive correlation with flow.

    One other thing: the drop in yields at the expiration of QE1 and QE2 cannot be a function of the expiration itself, as in both instances the date was known a priori by market participants. Thus, the "flow" may be said to cause an increase in yields, but not the subsequent decrease. One could argue the market initially expected each QE to be extended, but I don't believe that was consensus at the time.

    In short, we have some evidence QE flow raises yields, but we have no explanation, other than QE stock, for why yields should fall both at QE flow expiration and during the entire period. Further, there were other major events impacting yields over the period (i.e. Europe). Overall, it seems difficult to ascribe any kind of causality to either flow and stock effects given their opposite correlations, the timing of yield movements and the noise from other variables.

    ReplyDelete
    Replies
    1. In my opinion the drop in yields at the expiration of QE1 and QE2 due to another longer run phenomenon relating to the decline in longer maturity real yields. I do not think this is attributable to QE.

      Long term interest rates will be the sum of a liquidity premium and the average of the short term interest rates that are expected to occur over the life of the bond. Theoretically, QE should reduce long term rates in the short run via a reduction in the liquidity premium. Over the longer run QE should increase yields via higher expected nominal GDP. But overlaid on top of that is this third even longer run phenomenon of declining real yields which I argue is not directly related to QE, and is something of a mystery. See this David Glasner post for example:

      http://uneasymoney.com/2013/01/09/why-are-real-interest-rates-so-low-and-will-they-ever-bounce-back/

      Delete
  4. In my Granger causality tests on the US monetary base since December 2008 I find that the monetary base Granger causes the 10-year T-Note yield (and the impulse response is *positive*) but that the monetary base does not Granger cause the price of oil, copper or gold. On the other hand I do find that the monetary base Granger causes the real broad dollar index, but I don't see how anyone can call that "dollar debasement".

    In any case there are still lots of myths about QE out there. It's important that those myths be debunked.

    ReplyDelete
  5. Excellent blogging.

    Yes, the Fed is at about 1 percent on the PCE index, or half of their "target," which I sense is not a target but a sacrosanct ceiling, and so the real target is 1.4 percent or so.

    This is too low, and thus we may enter the next recession with interest rates nearly flat....

    ReplyDelete
  6. "One way to accomplish this would be to use forward guidance on the monetary base, promising that previous additions to the stock of high powered money will be permanent at least up until the point at which the Fed’s 2% inflation target its met (or is expected to be met). The same exercise could be done for a NGDP or price level path target." --Beckworth

    Excellent, but I would add, that the 2 percent target would have to be exceeded for some period of time, as in two years. If inflation pokes its head about 2 percent for a month does the Fed stop QE? I hope not.

    Actually, I think the Fed can hold onto its bond-hoard for eternity with positive result...and that raises the forbidden question: What if monetizing debt works in the current and foreseeable context?

    ReplyDelete
    Replies
    1. 2% isn't a target, it is a ceiling. Overhead inflation can go down and that is a good thing some of the time.

      The only reason inflation appears to be falling is because yry energy prices have dropped globally. Hence, demand inflation is rising.

      Delete