Believe those who are seeking the truth. Doubt those who find it. Andre Gide

Sunday, February 2, 2014

Monitoring Japan


I am as curious as anyone in ascertaining the effects of Japanese Prime Minster Shinzo Abe's QE experiment. Miles Kimball points us here to an early assessment by Marcos Nunes, who writes:
Shinzo Abe was elected in December 2012 on a promise to revive growth and put an end to deflation. How have his promises ‘performed’ one year after taking power? The ‘performance’ will be illustrated by a set of charts.
Nunes focuses on Japanese macroeconomic data beginning roughly with Abe's appointment as PM. But that's only about a year's worth of data. What I want to do here is compare these recent measurements with a longer sample, beginning in the year 2000 (the shaded region in my diagrams correspond to the Koizumi era, which I have written about before here).

First up is Japanese inflation (headline and core):




The evidence unfolding here really does seem to suggest that QE matters for inflation. My coauthor Li Li and I have recently remarked on this here.  Next, let's look at NGDP and RGDP growth:




Well, you know...this does not look so great, does it? While it is true that both NGDP and RGDP are growing, similar growth experiences are evident even in the earlier deflationary periods. Sure, it's nice to see RGDP growth rising recently, but it's still far too early to tell whether it will be sustained. And in any case, note the relatively robust period of growth during the "Koizumi boom" period--an era of deflation and fiscal austerity.

The exchange rate and the stock market:



 
So the stock market was booming late in the Koizumi era, the exchange rate stable, and core inflation negative. What about trade patterns? Take a look here:



 
I'll let you make up your own mind. Now for some comparisons with the Eurozone. First, a comparison of broad money growth:



Next, a comparison of inflation rates:



 And finally, a comparison of RGDP growth rates:



So sure, the Eurozone is underperforming as of late, and prospects in Japan are looking relatively good. How good in Japan relative to the Koizuma era, I'm not sure. And how much of the recent Japanese performance can be attributed to QE, one can only speculate. All that I conclude from this data is that QE may be influencing the inflation rate and the exchange rate. But whether it is having a quantitatively significant impact on the real economy is far less certain.

Addendum:

A comment by Noah Smith below suggests that Japanese CPI and GDP deflator are behaving quite differently. This indeed appears to be the case.


So, since about the time of the Asian financial crisis, the relative prices of non-consumer goods and services has declined steadily.

15 comments:

  1. There seems to be a slight discrepancy between the inflation graphs and the GDP growth graphs. The inflation graphs show inflation going positive, while the GDP graphs show nominal growth leading real growth. Is this a discrepancy between the GDP deflator and the CPI?

    ReplyDelete
    Replies
    1. Noah, I had my RA check it out. You are correct. I have posted the difference above.

      Delete
    2. Wow, interesting.

      Also: Do you happen to have up-to-date time series on Japese investment and consumption? I can't find them.

      Delete
  2. David, I agree with you, but my intention was to show the very recent developments.
    I longer run take on Japan is in the post below (before Abe). It seems to me to indicate that inflation targeting (even if implicit) may be "dangerous".
    The Koizumi boom is visible, and that was the result of a much weaker program.
    http://thefaintofheart.wordpress.com/2012/07/08/japan-poster-child-for-ngdp-targeting/

    ReplyDelete
  3. And this post is a complement to the previous one:
    http://thefaintofheart.wordpress.com/2013/07/18/liquidity-trap-is-just-a-failure-of-monetary-policy/

    ReplyDelete
  4. "And how much of the recent Japanese performance can be attributed to QE, one can only speculate. All that I conclude from this data is that QE may be influencing the inflation rate and the exchange rate."

    David, this statement does not do justice to your own discussion in the Fed note and the one we had earlier on this blog that the Abenomics QE is not like the previous ones because of the expected permanency of it. This is not just another test of QE. It is a test of a very different kind of QE. It is a new monetary regime change, one that so far seems to be credibly committed to larger nominal economy. The successes so far--and let's be clear these were the initial goals of Abenomics that many doubted would happen--should discussed in light of this new monetary regime.

    Also, what about the M3 series you show? It too seems to be breaking from its old trend. Your scaling makes it harder to see, but it is there in your comparison graph. Broader measures of money like the "L" money series the Bank of Japan tracks shows a similar pattern. In some ways this is an even more important development, because it shows how endogenous money creation can come via a change in expectations wrought by the new monetary regime. Many believed and claimed that QE leads to less or at best no change in financial intermediation. The growth in broad the money supply belies this claim.

    ReplyDelete
    Replies
    1. Hi David. All I am saying is that I remain agnostic. To me, it looks like QE has influenced inflation. Do you call this a success? I suppose, if inflation is your goal. But I like to look at real GDP. It's pattern doesn't look much different than what I've seen in the past, even under "contractionary" regimes. This is all just eye balling stuff to be sure. I just think we need more data. Hopefully, time will tell.

      Delete
    2. I agree we should not count our chickens before they hatch, but the count sure looks promising! My point, though, is different. The results we see so far are special not because they appear to show QE works, but because they provide a test of the importance of permanent injections. It is a natural experiment of sorts for this theory.

      Delete
  5. Abenomics at Crossroads

    February 3, 2014

    Needless to say, it is the Abenomics that drastically changed the economic policy regime in Japan for the better, and brought about the surge in Tokyo stock market in 2013, being accompanied by the Japanese currency’s rapid depreciation against the US dollar and the other major currencies. With close to 60% annual gain in price appreciation in the past year, the Tokyo stock market was clearly one of the best performers in the global equity markets. The yen/dollar rate now stands more or less in line with the PPP, the long run fundamental value. (See my book entitled by Japan’s Great Comeback with Abenomics (February 2013))


    Unfortunately, the Abenomics is now at great risk, however. The government’s fourth arrow of large fiscal consolidation program starting from April 2014, contradicts the expansionary fiscal policy, the second arrow of the Abenomics.

    In fact, 10 trillion yen (2.1% of the GDP) fiscal expansion mainly through public works under the supplemental budget in FY2013 has been playing the major role in boosting the economic recovery under the Abenomics. By contrast, it is difficult to assess how much the BOJ’s doubling the size of the monetary base itself has been contributing to the strong recovery of employment and the robust economic growth in general, although it has been successful for addressing the currency misalignments since 2008, and for boosting corporate earnings mostly for the major exporters.

    In particular, the sales tax hike to 8% as of April 2014 is highly likely to bring Japan back to the secular stagnation and deflation, as forcefully advocated by Mr. Summers in the IMF seminar last November. The depressed private consumption due to the tax increase is highly likely to contract the Japan’s aggregate demand by about 8 trillion yen (1.8% of the GDP).

    In order to offset such economic contraction, the government has proposed 5 trillion yen (1.2% of the GDP) supplementary budget including public works. Nevertheless, it would fall short of the size of the public expenditure in the previous supplemental budget by about 5 trillion yen. In other words, in FY2014, the government expenditures will decline by the same magnitude.

    Clearly, the Japanese economy is highly unlikely to be able to ward off such large negative fiscal shocks due to both large tax hike and significant reduction of the government expenditure. Given such large negative fiscal disturbances, the BOJ’s quantitative monetary easing will become ineffective.

    Moreover, unlike the US Fed now in the process of tapering the QE3, the BOJ would continue to commit itself to increase its monetary base. It seems to be self-evident under the Japanese government’s policy mix of large fiscal contraction and aggressive monetary expansion that the yen is going to further depreciate sharply, which could trigger the repeat of vicious cycle between Japan’s sales tax hike and the Asian currency crisis in 1997.

    Sincerely,

    Tomo Nakamaru
    An Ex-World Banker

    ReplyDelete
    Replies
    1. "In fact, 10 trillion yen (2.1% of the GDP) fiscal expansion mainly through public works under the supplemental budget in FY2013 has been playing the major role in boosting the economic recovery under the Abenomics."

      In my opinion the most objective way of measuring fiscal policy stance is the change in the general government cyclically adjusted balance, particularly the cyclically adjusted primary balance (CAPB). The cyclically adjusted balance takes into account any changes in the general government budget balance due to the business cycle. Thus changes in the cyclically adjusted balance are mostly due to discretionary fiscal policy, and consequently may be taken as a proxy for the degree of fiscal stimulus. The CAPB goes a step further, factoring out changes in net interest on government debt and thus ensuring that practically all of the changes in fiscal balance are discretionary in nature. The best place to find CAPB data is the IMF Fiscal Monitor. You can find the CAPB on the bottom half of Table 2 on page 70:

      http://www.imf.org/external/pubs/ft/fm/2013/02/pdf/fm1302.pdf

      Japan's CAPB is estimated to be (-7.7%), (-8.4%), (-8.5%) and (-6.0%) of potential GDP in calendar years 2012-15 respectively. Thus the change in CAPB is (-0.7%), (-0.1%) and (+2.5%) of potential GDP in calendar years 2013-15 respectively. In other words fiscal policy is estimated to have been mildly expansionary in 2013, to be largely balanced in 2014, and to be significantly contractionary in 2015.

      How has the 10.3 trillion yen, or about 2.1% of GDP, FY 2013 fiscal stimulus affected this result?

      For the sake of tractability let's assume the additional spending under the fiscal stimulus is evenly distributed throughout FY 2013. Given Japanese fiscal years start on April 1, this implies that in the absence of the fiscal stimulus Japan's CAPB would have been roughly (-6.9%) and (-8.0%) in calendar years 2013 and 2014 respectively. Thus the CAPB would have changed by (+0.8%), (-1.1%) and (+2.0%) of potential GDP in calendar years 2013-15 respectively without the fiscal stimulus.

      In other words, in the absence of fiscal stimulus, fiscal policy would have been mildly contractionary in 2013, mildly expansionary in 2014, and significantly contractionary in 2015. The effect of the fiscal stimulus therefore seems to have been to shift fiscal policy stance from mildly contractionary to mildly expansionary in 2013, mostly at the expense of shifting fiscal policy stance from mildly expansionary to balanced in 2014.

      In short, as I hope this simple analysis makes clear, I'm not sure it's at all fair to describe fiscal policy as "playing the major role in boosting the economic recovery under the Abenomics".

      Delete
  6. "So, since about the time of the Asian financial crisis, the relative prices of non-consumer goods and services has declined steadily."

    Actually this pattern in Japanese inflation goes back a lot further than that. Interestingly, from at least 1956Q1 until 1962Q2 yoy GDP implicit price deflator inflation always exceeded yoy headline CPI inflation. There was a period of transition with CPI inflation being roughly equal to GDP implicit price inflation around 1970. The gap seemed to stabilize around the mid-1970s. Since then yoy headline CPI inflation has averaged about 0.9% more than yoy GDP implicit price deflator inflation.

    Furthermore, this pattern is hardly unique to Japan. The US CPI also has a long history of rising faster than the US GDP implicit price deflator.

    ReplyDelete
  7. I don't think you're making a valid comparison here. Surely the external environment was much better in the Koizumi era and surely this matters quite a bit as evidenced by the fact that Japanese ouptut crashed along with everyone elses in 2009.

    To get growth back to the same rate with a poorer external environment seems to imply QE is having a real effect, no?

    ReplyDelete
  8. What does a "quantitatively significant impact on the real economy" actually mean...in light of the inflation and import/export charts presented here? Inflation looks up, so is that not evidence of its impact?

    ReplyDelete
  9. That chart of growth in imports and exports looks like a percentage change chart starting at an arbitrary baseline period, correct? I assume its measuring change in volume expressed as yen...and its interesting to me becasue I think it logical exports rise when more money is pumped in, but I would have thought imports would fall as the yen falls against the currency of trading partners.

    ReplyDelete
  10. Anybody else watching the big banks as they start to cut americas gdp forecasts...coincidentally after the fed taper has begun?

    So if it works in japan, why does it not work here?

    ReplyDelete