Tuesday, June 29, 2010

Yield curves in Japan and the US: similar but not the same

Andy Harless presents the case for a double dip (second recession) - I would re-order #1 and #2 on that list - and that for a sustained recovery. #6 of Andy's case for a sustained recovery (he calls it Case Against a Second Dip) caught my attention, pointing me to an earlier Paul Krugman article about positively-sloped yield curves in a zero-bound policy environment.

In a related article, Krugman argues that a current policy of near-zero short-term rates precludes the lowering further of future short-term rates. Therefore, the steep yield curve reiterates that rates have nowhere to go but up rather than that the economy is expected to improve.

Reasonable; but it was Krugman's comparison to policy during Japan's lost decade that got the mental wheels rolling:
Indeed, if we look at Japan we find that the yield curve was positively sloped all the way through the lost decade. In 1999-2000, with the zero interest rate policy in effect, long rates averaged about 1.75 percent, not too far below current rates in the United States.
In my view, current Fed policy is generally more credible than policy undertaken by the Bank of Japan in the early 2000's. The fed funds target has been near-zero since December 2008; and the new reserve base (liquidity) peaked quickly since the onset of QE and has since remained in the banking system.

Therefore, it would stand to reason that as long as policy remains consistent and big (the latter on the fiscal side is the problem right now), the US yield curve can, in my view, be interpreted as an auspicious sign - all else equal, as they say - as compared to the positively-sloped one in Japan.

Monetary policy in Japan: 1998 - 2006

The Bank of Japan has a solid history of rescinding their own policy efforts. They did it earlier this year; but more importantly their policy announcements spanning the years 1999 to 2006 have on occasion been rather deceiving. Notice that the 2-10 yield curve never became inverted.

The shortened version of the timeline (illustrated in the chart above):
  • From Bernanke, Reinhart, and Sack (2004): "In April 1999, describing the stance of monetary policy as “super super expansionary,” then-Governor Hayami announced that the BOJ would keep the policy rate at zero “until deflationary concerns are dispelled,” with the latter phrase clearly indicating that the policy commitment was conditional."
  • In August 2000, The BoJ raises the overnight call rate to 0.25%, up from near-zero.
  • In February 2001 the BoJ lowers the overnight call rate to 0.15%.
  • In March 2001, the BoJ announces its quantitative easing strategy, initially targeting current account balances (essentially reserves) at 5 trillion yen and lowered the overnight call rate target to near-zero.
  • Until 2004, the BoJ raises the current account reserve target several times until it peaks at 30-35 trillion yen.
  • In March 2006, the BoJ exits QE.
I concur with Paul Krugman, that the deflation threat is very very real. I do not think that it is completely fair to compare the current US yield curve to that to early 2000's Japan.

To be sure, the likelihood of rates rising is the only possibility built into the US yield curve right now (no possibility of lower rates); but since the Fed is relatively more credible and consistent, the probability of rates rising is much higher compared to that in early 2000's Japan.

And the current US curve is steep! The chart below compares the dynamics of the 2-10 yield curve in Japan from its low in 1998 through 2006 to that in the US from its low in 2007 through June 24, 2010.

Rebecca Wilder

Reference for paper in final chart: Luc Laeven and Fabian Valencia (2008), Systemic Banking Crises: A New Database, IMF Working Paper WP/08/224.


  1. Mainstream economic thinking has been wrong for the last 55 years.  Ever since 1965 the operations at the "trading desk" have been dictated by the assumption that the money supply can be controlled via interest rates. 
    <span>ZIRP? No, by definition the interest rate "floor" is the remuneration rate. There is no “liquidity trap”.<span>  </span>The FED is "pushing on a string" with its new policy tool - IORs.</span>
    <p><span> </span><span>What the FED has fostered is a contractionary policy with the use of a risk rate penalty. </span>
    </p><p><span>The floor on the FFR (or the interest rate on excess & required reserves), now @ .25%, has created dis-intermediation among the non-banks (an outflow of funds), and has reduced money velocity, in the thrifts, as well as the CB system.</span>
    </p><p><span> </span><span><span><span>IORs have caused massive portfolio shifts in the earning assets among the commercial banks ($1,047,858T in new excess reserves).<span>  </span></span></span></span>
    </p><p><span> </span><span><span><span>These portfolio shifts have induced system-wide bank credit contraction (the remuneration rate on IORs will have exceeded all 4-week, 3-month & 6-month Treasury bills for 2 years as of this Nov 5th).<span>  </span>Back then, the target FFR was @ 1% (on 11/05/08).<span>  </span></span></span></span>
    </p><p><span>The liquidity preference curve is a false doctrine. (see Alfred Marshall's "money paradox").The money supply can never be managed by any attempt to control the cost of credit (whether the FFR & IORs), & the Taylor Rule is a fictitious "sign post".</span>

    </p><p><span> </span><span>Nominal GDP will cascade in the 4th qtr (down in every month - Oct, Nov, & Dec), without extra (upwards of the linear path), stimulus.</span></p>

  2. I'll throw in one other issue here:  the US is a net debtor and consistently runs a large trade deficit. Japan is (IIRC) a net creditor and consistently runs a trade surplus. There is thus, in the case of the US, a particular reason to worry that there will be an eventual rebalancing of trade that will involve a significant decline in the foreign exchange value of the currency.  Presumably, the further you project into the future, the more likely it is that the dollar will have fallen significantly, and therefore the more reason there is to demand an interest rate premium to compensate for the anticipated currency translation loss (or opportunity loss, in the case of a US investor).  The difference in trade balances could thus possibly explain the difference in yield curve steepness between the US today and Japan at an analogous point in time.

    Also, there may be more reason to attach an inflation risk or default risk premium to longer-term government securities in the US than there was in Japan.  For one thing, technically, the US Treasury has actually defaulted a couple of times in the past 30 years.  While those defaults may not have had much real substance, they do suggest that the political situation in the US presents risks for investors.  Perhaps more important, the US is facing health care costs that are projected to rise at a rate that would eventually bankrupt the government, and recent experience suggests that it may be very difficult politically to address those costs in a satisfactory manner.  (On the other hand, Japan did suffer an actual rating downgrade at one point, which doesn't appear to have had any noticeable lasting effect on the slope of their yield curve.)

    You make a valid point, though:  Krugman has clearly overstated his case.  On the face of it, the Japanese evidence is really rather more against his point than for it.

  3. When, in your view, will it be reasonable to stop paying banks to keep their customers' deposits instead of lending them out? (interest on excess reserves)