Pages

Monday, March 30, 2015

Ben Bernanke and the Secular Stagnation Debate

Ben Bernanke is back and mulling over low interest rates. From the Washington Post:
Blogging isn't dead. At least, Ben Bernanke, former chairman of the Federal Reserve, doesn't think so: He's now blogging at the Brookings Institution. In his first post, he says he wants to write about this fascinating chart, which shows the steady decline in interest rates over the past 30 years or so.
The question of why interest rates keep falling is an important one these days. Former Treasury Secretary Larry Summers has argued that declining real rates are a symptom of secular stagnation, meaning the global economy just isn't what it used to be, for some reason, and might never recover its old strength. Here's what Summers told Wonkblog about his theory, and here's a response from Western Kentucky University's David Beckworth, who disagrees with him.
It would be great to see Bernanke engage this secular stagnation debate. Although Larry Summers never directly responded to me, he did reply to Marc Andreessen's tweetstorm where my critique was raised. The crux of my argument was that the long-decline in real interest rates given as evidence of secular stagnation ignores the sustained decline in risk premiums. Once this phenomenon is recognized, there is no long decline in real interest rates. 
 
Larry Summers replied to my critique with the following:
Markets – in the form of 30-year indexed bonds – are now predicting that real rates well below 2 percent will prevail for more than a generation... I think it is quite plausible and consistent with Marc’s picture that equilibrium real rates were roughly constant at around 2 percent until the mid-1990s and have trended downward since that time.
Looking to inflation indexed bonds or TIPs as a guide to the market's prediction of real rates is also misleading. It too fails to account for a liquidity premium priced into TIPs. On his second point, below is an updated version of the picture to which Larry Summers says he sees a downward trend since mid-1990s. Note that the real interest rate is now turning sharply up whereas before it was still flat. The real interest appears poised to return to its previous trend.


That the real interest rate appears to be returning to its previous trend--rather than finding a new lower one as suggested by Summers--makes sense if we plot this interest rate against the CBO's output gap. This is done in the figure below and reveals a striking fit. It also suggests the real culprit behind the sustained low rates is a prolonged business cycle. Now that the economy finally appears to be on a path to full recovery, the output gap is closing and the real interest rate is following it. So much for the smoking gun of secular stagnation.


I really hope Ben Bernanke joins this conversation. He had a great speech back in March 2013 that hinted at some of these topics. Welcome to the blogosphere Ben!

P.S. Although my original secular stagnation critique was made at the Washington Post, I did a more thorough follow-up piece with Ramesh Ponnuru at the National Review. I also did an interview with a Brazilian newspaper on secular stagnation.

8 comments:

  1. On interest rates, I think there is also a question of supply and demand. There are now many global sovereign wealth funds accumulating assets regardless of market rates. There are also insurance companies globally collecting assets in a manner somewhat removed from market dictates. Likewise many pension funds accumulate assets continuously. With rising incomes there is also an upper-class able to save. All in all there is a global of capital and it would be natural to assume lower interest rates. I would say large institutional imperfections play a role in global interest rates.

    ReplyDelete
  2. David: When one argues that the real rate is rising because the price environment has become deflationary, as a way of denying the existence of secular stagnation, it is hardly an argument from strength. Also, please consider that the risk premium you are drawing from market data becomes massively skewed at the zero lower bound. In fact, the real rate itself is skewed by ZIRP. Consider, if you will, that the real rate in Japan was rising (as was real GDP) during Japan's deepest deflationary periods of the past 15 years when nominal GDP was flat or falling. Please, I know that cyclicalists have a deep need to find reversion in data, but sometimes - really - long term trends are really worth considering. Especially when you have developed economies working through a problem as vexing as the absorption of emerging labor forces that, in the aggregate, vastly outnumber the labor of developed economies, are over producing, and are under consuming. This, together with obvious demographic issues, comprises a real secular challenge against with business cyclical theory is no match.

    ReplyDelete
    Replies
    1. Daniel,
      If you take a look at the original pieces you will see that I am not arguing the real rate is rising because of an deflationary environment. Also, the risk premium measure is not a market asset price directly. Rather, it is an estimated series. I would encourage you to look at this piece for more on how I constructed the real risk-free 10-year treasury rate: http://macromarketmusings.blogspot.com/2015/01/dont-worry-be-happy-falling-treasury.html (The only difference here is that I am using the quarterly SPF consensus expected 10-year inflation rate whereas on that post I am using breakevens for the daily 10 year expected inflation rate)

      I agree that that a part of the decline in observed nominal yields that is a long-term structural development--the rising demand for safe assets exceeding the capacity to produce them--but that is being captured in the term premium not the real risk-free rate shown above. See the link above for more.

      Delete
  3. You got your wish:

    http://www.brookings.edu/blogs/ben-bernanke/posts/2015/03/31-why-interest-rates-low-secular-stagnation

    Ben Bernanke must be reading you.

    ReplyDelete
    Replies
    1. Bernanke gave a speech on the subject yesterday before this blog post. But possibly he read the older post on risk premiums.

      http://www.c-span.org/video/?325073-1/ben-bernanke-remarks-economic-growth-employment

      If I understand him correctly, he does say in part that it was cyclical: the economy is recovering and rates are going back up as David mentions.

      But he also talks about the "saving glut" dynamics which temporarily push down interest rates. In the speech he mentioned China's exchange rate policy which he says is now moderating. He also mentioned Europe, where Germany has record trade surpluses and even the periphery is now in surplus because their depressed economies are not importing. As the periphery and Europe recovers, interest rates will go back up. He said he will have another blog post filling in the data details.

      Delete
    2. JP and Peter, I watched the video and read the post. He makes great points, but doesn't mention my falling risk premium critique so I doubt he's reading my blog. But hey, he shares my optimistic outlook for the US economy and I'll take that.

      Delete
  4. David,

    I had a methodological question on whether it makes sense to take the nominal yield, subtract expected inflation and subtract risk premium? The ACM risk premium calculation uses nominal rates to begin with, so one might argue that inflation expectations are embedded in their meausure of risk premium. If that's so your real interest rate calculation would be double counting inflation.

    ReplyDelete
  5. Carlos, the risk premium--which here is the term premium--is distinct from expected inflation. See http://macromarketmusings.blogspot.com/2015/01/dont-worry-be-happy-falling-treasury.html

    ReplyDelete