Pages

Wednesday, December 21, 2011

Much Ado About Nothing: Financial Repression Edition

There has been a lot of discussion on financial repression emerging in advanced economies as way for governments to handle the looming debt crisis.  According to some, financial repression is already in play in the United States as the Federal Reserve is keeping long-term interest rates artificially low to minimize financing costs to the Treasury.  Advocates of this view go on to note that the lowering of long-term interest rates is narrowing the net interest margins for banks and reducing the incentive for savers to fund the shadow banking system.  Financial repression, therefore, is causing financial intermediation to fall and is preventing a robust recovery. 

There is a big problem with this view: it wrongly assumes that the drop in long-term interest rates over the past few years is solely the result of the Fed's large scale asset purchases (LSAPs).  While it is true that there has been a spate of empirical studies showing the LSAPs have lowered the term premium portion of long-term interest rates, most of these studies only show modest effects.  It is unlikely, for example, that the LSAPs can account for much of the 300 basis points plus drop in the 10-year treasury interest rate since 2007.  The financial repression advocates, however, want to attribute all of this decline to the Fed's actions.

A far better explanation for the large drop in long-term interest rates is one, the growing global demand for safe assets and two, the ongoing slump in the economy.   The first of these factors is about the increasing scarcity of safe assets in the world economy even as the global demand for them grows.   U.S. treasuries remain the go-to safe asset for the world.  As I discussed previously, there are both structural and cyclical factors behind this shortage of safe assets with both implying a reduction in the term premium for U.S. public debt.  The second factor is that since the current and expected economic outlook continues to look bleak, the current and expected path of the short-term natural interest rate is low.  With the short-term natural interest rate expected to remain low, actual short-term interest rates will be expected to remain low too and thus pull down the long-term interest rate. Some observers seem to forget that the natural interest rate itself is determined by the state of the economy.

Another problem with the financial repression view is that the Fed's LSAPs, while very imperfect, were never explicitly intended to keep down government financing costs.  They were always about either saving the financial system (e.g QEI) or more recently the broader economy (eg. QEII and Operation Twist).  These programs had serious flaws--they should have explicitly targeted the level of nominal spending  without committing dollar sums upfront--but to attribute to them a motive of repressing the financial system to help save public finances seems unfair.   

For these reasons the financial repression view seems untenable to me.   It is much ado about nothing.

9 comments:

  1. David,
    Flight to safety may not explain a collapse in the term premium. Investors could choose to flee to T-bills and not sacrifice compensation for duration risk. At the zero bound, the term premium is a function of the expected volatility of future short term rates, which is in turn a function of the future path of policy. The Fed is signalling that that volatility will be low, both through LSAPs and ZIRP promises. Hence the flat yield curve.

    In the case of Japan, market-derived expectations of low nominal rate volatility co-existed with deflation expectations (at the zero bound, persistent deflation would produce rate vol of zero). In our case, positive inflation expectations and low rate volatility expectations can only coexist as a result of Fed policy. Any policy that conciously targets a lower risk-free term premium can be labeled as financial repression.

    ReplyDelete
  2. David Pearson,

    I think you are saying, that in addition to changing the supply of treasuries, the Fed can also affect the term premium by increasing the certainty of the future path of policy. Fine, but again the estimated effects from these Fed actions have been shown to be modest.

    The demand for safe assets (including the flight to quality) by my reading of the literature has a bigger effect on the term premium. Add in the weak economy's effect on the expected path of short-term interest rates and you should account for most of the drop in long-term interest rates. In short, I can't see that much of a role for financial repression in explaining recent variation in long-term interest rates.

    And this seem a bit strong to me: Any policy that conciously targets a lower risk-free term premium can be labeled as financial repression. Isn't motive important here? I thought the point of financial repression is to help public finance. The Fed's goal is stimulate the economy (which should ultimately raise real returns if it worked) not help the U.S. treasury.

    ReplyDelete
  3. David,
    To the extent that the term premium reflects very weak NGDP growth expectations, corporate earnings estimates are bound to fall. The divergence in expectations between markets offers an arbitrage oppotunity (short stocks/long bonds). Two alternative explanations are more compatible with EMH: 1) flight to safety; and 2) Fed policy that directly targets a lower term premium. The former requires some theoretical basis for investors to accept less compensation for duration risk (aside from that created by weak domestic NGDP growth expectations).

    ReplyDelete
  4. David Pearson,

    To the extent that the term premium reflects very weak NGDP growth expectations

    Why would the term premium be reflecting weak GDP growth? Normally, weak growth is seen through the expected path of short-rates.

    The flight to risk is just a special case of the rising demand for safe assets. And that can be explained by the segmented markets hypothesis or portfolio theory. See Gagnon et al. (2011) (http://www.ny.frb.org/research/epr/11v17n1/1105gagn.pdf)

    ReplyDelete
  5. Some people just can't stand the idea that there is a glut of capital globally, and that is a good thing---if we could stimulate our economies enough to harness that great supply of capital out there.

    George Gilder, for all his flaws, used to understand this problem--and always advocated growth policies, not necessarily anti-inflation policies.

    At heart, the concerns about "financial repression" are expressions of gold nuts and others, who like to imagine capital should be scarce (giving to people with capital more leverage?), or that the supply of money is sacred.

    Those days are over. Capital will be abundant for generations, as global incomes rise and cultures with higher propensity to save become even richer.

    The challenge ahead is to not to fight inflation, but stimulate growth enough to suck down all that capital. (I am beginning to wonder if my lifelong desire for consumption taxes is off base now. If we have gluts of capital and not enough demand, should we tax investment and not consumption? Heresy!)

    We are like Japan in many ways---real estate assets are depreciating, and stocks limp. Just like Japan in that regard. Unit labor costs are declining in the USA.

    It could be the CPI overstates inflation--in that case we are at zero or minor deflation now.

    Market Monetarism appears to offer the only real course forward.

    ReplyDelete
  6. David,
    At the zero bound, any drop in long term rates caused by weak growth expectations translates into a drop in the term premium.

    BTW, "financial repression" is not a pejorative term. Rather, in my mind at least, it describes a situation in which the central bank produces abnormally low or even negative real interest rates for much of the yield curve, rendering deficit financing "costless". In the absence of such policy action, very weak economic growth in the presence of an output gap tends to produce deflation expectations and positive real rates.

    ReplyDelete
  7. David Pearson,

    1. The below is hard to reconcile with standard term structure theory [i.e. long-term rates are equal to the average of short-term rates over the same period plus a term (or risk)premium]:

    "At the zero bound, any drop in long term rates caused by weak growth expectations translates into a drop in the term premium".

    The only way this might happen is if the short-term interest rate is expected to be 0% over the entire horizon of the long-term rate leaving only the term premium to adjust. For the 10-year treasury interest rate this is unrealistic because somewhere over the next 10 years the economy should recover and raise short-term rates. If anything, the expected weak growth should pull down expected short rates, not the term premium.

    2. "... a situation in which the central bank produces abnormally low or even negative real interest rates for much of the yield curve."

    You are assigning too much power the Fed. The expected path of the federal funds rate is shaped largely by the expected state of the economy. Another way of saying this, is that over the long-run the Fed brings the federal funds to its natural interest rate level. the Fed may temporarily affect the expected path or modestly affect the term premium, but over the long run it is a slave to the fundamentals of the economy.

    ReplyDelete
  8. "A far better explanation for the large drop in long-term interest rates is one, the growing global demand for safe assets"

    driven at least partially by regulatory capital requirements, n'est-ce pas?

    ReplyDelete
  9. I don't think financial repression should solely refer to public finance. it could also be aimed at benefiting private borrowers. But I do think it should be limited to situations where there are binding interest rate price ceilings and various other regulations to limit ultimate savers to those price controlled debt instruments.

    Where the central bank is lending more one way or the other, and that results in lower interest rates, it isn't the same thing at all.

    ReplyDelete