Monday, March 17, 2014

House of Debt

Atif Mian and Amir Sufi have started a blog related to their new book, "House of Debt." Amir and Atif are admirably data-oriented, which ought to make for good reading.

Today's post "Fed Meetings and Asset Prices" is a good example. They put together one-day returns on the June 19 "taper tantrum" when the Fed announced it might (heavens) start tapering bond purchases. There is, of course, a large literature studying announcement effects. Atif and Amir  put together an unusually wide spectrum of asset classes.



It's interesting that corporate bonds are most affected. Really, the credit spread widened. It's interesting that the intermediate government bond is hurt more than long and short -- maybe there is something to the idea that the Fed really affects the 10 year maturity where it's doing more buying. It's interesting that banks (financials) are not much affected. (Though be careful, these returns are scaled by volatility.)

My thought: This cross section is most interesting as an illumination of why QE might or might not work. Markets move, but never tell you why they move, and analysts jump too quickly to the conclusion that these reactions measure the direct economic effect of quantitative easing. Here are some theories to distinguish:

  1. QE works directly, by changing the M in MV=PY. The announcement of tapering means there will soon be less M so we should primarily see output and inflation effects. 
  2. QE works by lowering the long-term interest rate via some sort of segmented markets story. Thus, this is news that the yield curve will start to steepen, with no news about the low end
  3. QE is irrelevant per se, the Fed is buying green M&Ms in return for red M&Ms. But we all know that first comes tapering, then comes interest rate hikes, (then comes macroprudential meddling?) So, this is news that the short end of the yield curve will rise sooner than expected, with any long end effects coming from expectations hypothesis logic alone. We update to a higher but flatter yield curve.
  4. QE is completely irrelevant, but announcements reveal the Fed’s analysis of economic activity. The Fed is one of the most thoughtful economic forecasting shops around, with loads of private information, especially about what’s going on at the TBTF banks. If the newspapers had said “Mohamed El-Erian says economy stronger, rates to rise in 14”, markets might well have moved on the news too. 
  5.  QE is completely irrelevant, period. Markets think the Fed matters a lot, but it doesn’t. We’re out of rational expectations equilibria because nobody has seen a $ 3 trillion balance sheet, interest on reserves, and zero rates before. When in 44 AD the high priest of the temple of Jupiter came forth the day following the ides of iulius, to announce that he had spied the pattern in the murmuration of starlings, and that the harvest would taper down this year (I totally made this up), Roman grain futures markets fell. Are we really so much smarter?

I find the cross sectional results most interesting in how they might help us sort these stories out. Amir and Atif speculate a bit about what it means, and promise more in future posts. Worth watching.

13 comments:

  1. 1. "QE works directly, by changing the M in MV=PY. The announcement of tapering means there will soon be less M so we should primarily see output and inflation effects."

    Any increase in M caused by the central bank will be offset by a decrease in V - especially when the term structure of existing debt is short / or when most debt is adjustable rate.

    2. "QE works by lowering the long-term interest rate via some sort of segmented markets story. Thus, this is news that the yield curve will start to steepen, with no news about the low end."

    Depends in part on what happens with fiscal policy (supply of long term debt).

    3. "QE is irrelevant per se, the Fed is buying green M&Ms in return for red M&Ms."

    The Fed is buying less liquid assets for more liquid assets. Relevant only if you believe the economy suffers from a lack of liquid assets or you believe that the Fed can transform illiquid assets into liquid assets. (Would that make the Fed an Autobot or a Deceptacon?).

    5. "QE is completely irrelevant, period. Markets think the Fed matters a lot, but it doesn’t."

    It would be helpful here if you distinguish between the FOMC and the Federal Reserve. The open market committee sets the price they will pay for existing government bonds. The Federal Reserve sets the interest rate they are willing to lend money at.

    ReplyDelete
  2. Dr. Cochrane:

    I like your five your theories, especially the part about the Temple of Jupiter.

    But unless I miss it, in your five theories you make no mention of perceived credit risk, unless it is "taken for granted," so to speak, in No. 1

    It could be creditors think that without monetary expansion that debtors will have more difficulty in paying off IOUs, ergo there is higher credit risk.

    After such a long period of relatively dormant inflation, and after the Japan experience, bondholders may have concluded that interest-rate and inflation risks are smaller, but credit risks are to be reckoned with.

    After all, when the Fed tightened in 2008, the result for many bondholders was a catastrophe (even when they had insurance from AIG, and even if they bought AAA bonds).

    The chart somewhat supports the idea that bondholders fear credit risk more than inflation or higher interest rates, following tapering.

    PS

    Back on the Temple of Jupiter, true we have not seen a $3 trillion balance sheet at the Fed before, although the Fed did have a balance sheet of $800 billion in 2007 (which no one ever mentions) which they reduced to $400 billion before the bust in 2008.

    So the Fed has had large balance sheets before, and if there was an adverse effect, it was when they unwound it.

    We have also some empirical results from Japan, 2002-7, when the BoJ conducted QE. John Taylor said the results were very good.

    It may be that too-rapid tapering could lead to unforeseeable, unintended but corrosive consequences for debtors, who will be less able to repay corporate IOUs. The sudden reduction of the Fed balance sheet from $800 billion to $400 billion just before the 2008 bust in worth pondering.



    ReplyDelete
  3. "QE works directly, by changing the M in MV=PY."

    I'm hoping this was just a momentary slip into macroeconomic tautologies on your part, and that you'll soon return to finance-based models of money that consider both the liabilities and the assets of the money-issuer.

    ReplyDelete
    Replies
    1. I listed that as a possibility, not an endorsement. Some people like it. Not me.

      Delete
  4. Just wondering... does not the methodology, scaled to daily vol, result in a potential and indirect emphasis on (or conflation of) liquidity issues specific to each ETF? Particularly on 'bad hair' days for the market? If ETF A (and the underlyings) has less liquidity (and more friction costs) than ETF B, then would it not be particularly amplified in that day's return? The round trip gets potentially more expensive for one but not for the other... as a function of liquidity rather than expectation.

    One suspects that would be the case with VCSH vs VGSH. Dunno.

    ReplyDelete
  5. QE seems likely to have two effects:
    1) Some flattening of the yield curve with most of the effect being felt in the financial markets rather than in production and consumption.
    2) The buying of mortgage backed securities is operating as a direct subsidy of the housing market. To the extent that the Fed is paying more for MBS than an undistorted market would the increase is a direct subsidy to housing and is more in the nature of a fiscal than monetary policy. If MBS were under priced than there is no harm. If the Feds actions have led to them overpaying for MBS then there is going to be a day of reckoning down the road.

    Personally, I think that QE has been a huge mistake and should be ended at the first opportunity. Instead, QE continues to pump tens of billions of dollars into the economy every month. To "taper" is just to slow down the ongoing flow.

    ReplyDelete
  6. John,

    What about that QE works by raising all types of asset prices, not just of bonds?

    ReplyDelete
    Replies
    1. How? The fed takes your red M&Ms, gives you green M&Ms in return, and this affects the risk premium on mortgage-backed securities, not just the level of the short term rate? Sure, we're all smart enough to come up with cocktail party frictions stories to generate anything we want, but we need something a bit more reliable before embarking on trillion dollar programs.

      Delete
    2. What about tobin's q and exchange rates?

      Delete
  7. University of Chicago Is Outlier With Growing Debt Load
    By Michael McDonald and Brian Chappatta Mar 17, 2014

    The University of Chicago has been trying to stand out from its elite rivals and is doing so in one category: amassing debt. That’s put its credit rating at risk.

    The university founded by oil magnate John D. Rockefeller in 1890 is in the midst of a $1.7 billion development plan. As municipal interest rates remain close to generational lows, it may borrow as much as $900 million in the next four years, according to Standard & Poor’s. The plan prompted S&P and Moody’s Investors Service to cut the school’s credit outlook to negative. Chicago already has $3.6 billion of debt -- the most relative to its endowment among the richest U.S. schools.

    “When you look at the University of Chicago, they are an outlier -- they are highly leveraged,” said Diane Viacava, an analyst at Moody’s in New York.

    --30--

    Well...maybe we can track University of Chicago bonds against the taper....

    ReplyDelete
    Replies
    1. It's not quite so simple. Universities borrow at municipal (tax free rates) for construction. The U of C is on a construction binge, for good reasons (my view). U of C also has $6.7 billion in endowment. In once sense, $3.6 billion in debt and $6.7 billion endowment really means $2.9 of assets and no debt. This is not like a corporation with only fixed assets and a lot of debt.

      Delete
    2. Professor, interested in commenting the following post?

      http://houseofdebt.org/2014/03/18/the-most-important-economic-chart.html

      Regards

      Delete
    3. John,

      "In one sense, $3.6 billion in debt and $6.7 billion endowment really means $2.9 of assets and no debt. This is not like a corporation with only fixed assets and a lot of debt."

      Huh?? Unless the endowment is buying up municipal bonds issued by the University, the value of assets and liabilities will not necessarily track together (future value of assets can be lower, future value of liabilities remains the same).

      Also, I would be willing to bet that the city of Chicago and / or the state of Illinois tax payers bear some risk if the University were to default. And so, in that sense, it is different than a corporate with a lot of outstanding debt (no taxpayer backstop).

      Delete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.