Sunday, September 2, 2012

Jackson Hole Conference

Ben Bernanke's job is not easy, though that's partly his fault. By speaking extensively to the "maximum employment" part of the dual mandate, he has defined success in a way that dooms him to failure. By embarking on unprecedented and aggressive policy programs - quantitative easing (QE) and forward guidance of various sorts - he has taken on some big risks.

In his Jackson Hole speech, Bernanke's job was to do the best he could in defending past decisions, and in arguing that the Fed is prepared to face whatever the future holds.

Bernanke wants you to think of the people running the Fed as mechanics with a large box of tools. If the economy is to run the way we want it to, all we need to do is find the right tool, and fix it. The more tools, the better. In his speech, Bernanke wants to tell us about "balance sheet tools" and "communication" tools - QE and forward guidance, respectively.

QE and forward guidance are "unconventional" central banking tools. Why do we need these unconventional tools? According to Bernanke, the economy needs fixing, and we can't use conventional tools - open market purchases of short-term government debt - as the overnight nominal interest rate is as low as the Fed is willing to push it.

How do we know that the unconventional tools work? As economists, we feel more assured if we have sound theory and empirical evidence that we can argue are consistent with particular policy actions. What's the theory behind QE? According to Bernanke:
One mechanism through which such [long-maturity asset] purchases are believed to affect the economy is the so-called portfolio balance channel, which is based on the ideas of a number of well-known monetary economists, including James Tobin, Milton Friedman, Franco Modigliani, Karl Brunner, and Allan Meltzer. The key premise underlying this channel is that, for a variety of reasons, different classes of financial assets are not perfect substitutes in investors' portfolios.
For most modern macroeconomists, that won't be very assuring. "Porfolio balance theory" was the stock and trade of James Tobin more than forty years ago. In this class of models, one specifies a set of asset demand functions - one for each asset, where the demand for an asset depends on the rates of return on assets and on wealth. In general, open market operations in any asset will be non-neutral in such a model. If the central bank purchases a particular asset, then its price will go up, and the rates of return on other assets will change, depending on whether these assets are substitutes or complements for the asset the central bank purchased.

What's wrong with portfolio balance theory? Principally, it ignores how assets are used in retail and financial transactions, and how assets are intermediated by financial institutions. As well, complementarity and substitutability among assets are in practice functions of financial regulation and government policy rules, and should not be treated as "structure." Further, Tobin's models were static, and dynamic elements are critical for studying the effects of monetary policy.

But, Bernanke may be one step ahead of us, as in a footnote, he states:
For modern treatments of the portfolio balance channel in a macroeconomic context, see Andrés, López-Salido, and Nelson (2004). The portfolio balance channel would be inoperative under various strong assumptions that I view as empirically implausible, such as complete and frictionless financial markets and full internalization by private investors of the government's balance sheet (Ricardian equivalence).
Standard frictionless models have no role for quantitative easing. Indeed, the Arrow-Debreu framework has no role for a central bank or for central bank liabilities. To start thinking about monetary policy and what it does, we need a model with "frictions," i.e. we need a formal description of how different assets are used in exchange and an explanation for why conventional and unconventional central banking actions might matter.

The Nelson et al. paper builds a model based on conventional representative agent macroeconomic frameworks, and adds to this some frictions, of sorts. It's certainly not in the New Monetarist spirit. Central bank liabilities are in the utility function, there are costs to adjusting money balances, and the frictions that will give a role for QE are transactions costs associated with purchasing long-maturity bonds, and "self-imposed reserve requirements." This paper did not teach me much about why QE may or may not work.

Basically, Bernanke does not have much to go on in the way of theory to justify the use of QE. What about empirical work? Plenty of this has been done by now, and Bernanke cites some of it, but without a theory, where are we? We can find plenty of people - most of them in the Federal Reserve System - who are now willing to vouch for the fact that QE announcements move asset prices in the "right" direction. But what of it? What do the announcements mean? How do we separate the forward guidance aspect of a QE announcement from the basic QE effect?

On forward guidance, Bernanke makes clear what the "2014" language in the FOMC statement means, in case you were confused:
As the language indicates, this guidance is not an unconditional promise; rather, it is a statement about the FOMC's collective judgment regarding the path of policy that is likely to prove appropriate, given the Committee's objectives and its outlook for the economy.
Thus, there is no commitment implied by the language in the statement. Bernanke wants us to interpret the forward guidance in the FOMC statement as a forecast for when the Fed is likely to raise its policy rate. Maybe this is not such a bad idea. Under normal circumstances, every FOMC meeting gives us more information on the Fed's implicit policy rule, and theory tells us that the more we know about the policy rule, the better. But at the zero lower bound, we don't get much information about what the Fed is up to. Forward guidance might help in this respect, though I doubt that the FOMC will want to keep the interest rate on reserves at 0.25% for the next two years.

Woodford's paper from the Jackson Hole conference seems to be getting some attention. I wouldn't recommend reading all 96 pages of this tome unless the value of your time is close to zero. To get the general idea, read pages 82-87. Here's something I disagree with:
To the extent that central banks have been willing to make explicit commitments about the future conduct of policy, it has most often been to underline their commitment not to allow higher inflation than would correspond to their normal, long-run inflation target, even temporarily. Thus, in the case of the Federal Reserve, the introduction of more explicit forms of forward guidance and aggressive expansions of the Fed balance sheet have been accompanied by assurances that these policies should in no way suggest that there will be any relaxation of the FOMC’s vigilance when it comes to preventing any increase in inflation. Such assurances tend to contradict precisely the kind of signals that one would want such policies to send in order for them to be effective in providing people a reason to spend more.
A key problem for the Fed is that it needs to reassure people that it remains committed to controlling inflation, in spite of the unusual circumstances. The fact that inflation expectations - as reflected in TIPS yields, for example - remain subdued, reflects the Fed's efforts to maintain its hard-won credibility as being serious about inflation. I don't think Woodford, or anyone else, would like the results if the Fed were to abandon that commitment.

Woodford seems not to think much of QE, and goes off on an extensive discussion of forward guidance, most of which made me happy that Woodford is not in charge of forward guidance at the Fed. If he were, we would never understand what they are up to.


  1. very good ,,MAXIMUM EMPLOYMENT what mean? ''full''with unvulentary or invulentary?
    never have full employment statistical
    inflation ------how..?if there dont print money (ofcours with limit) and for specifice sector use not gv administration..

  2. Welcome back.

    SW: "The Nelson et al. paper builds a model based on conventional representative agent macroeconomic frameworks, and adds to this some frictions, of sorts. It's certainly not in the New Monetarist spirit."

    Are there any papers in the new monetarist literature that do find QE to be relevant? If so, what frictions do these papers invoke?

    If not, imagine you've died and gone to heaven. God says: "Steve, you know, QE actually wasn't irrelevant. I'll let you through the gates if you guess what frictions made it relevant."

    What would your three picks be?

    1. I think it's obvious there is something to "preferred habitat"-type models, where different households have different preferences for assets with different maturities. I think most of that would be coming from life-cycle considerations - I think it's reasonable to expect a 70-year-old with no children to have no desire to hold a 30-year Treasury bond, but it's obvious a 40-year old may be interested in such a bond. But this consideration doesn't mean QE is necessarily effective because financial intermediaries can still arbitrage this away - and it's implausible to argue that financial firms have some intrinsic preferred asset maturity.

      It's hard to make QE work in models, because for balance sheet policies to be effective the government more-or has to have some advantage relative to the private sector. Essentially what you need is just one thing - the Federal Reserve must be somehow better at arbitrage than the private sector. You could imagine that there is a situation where due to agency costs of some sort, like moral hazard or limited commitment, there exists balance sheet constraints and it inhibits financial intermediaries from obtaining funds. However, the Fed can overcome this agency friction because it can issue government liabilities like reserves and get funding elastically because it can credibly commit to honoring its debts. Ergo, it has the ability to affect asset prices and returns. While it would be nice to have explicit evidence for everything I just said, I think all of these things are reasonable assumptions. But everything I've just described is not quite enough to make QE effective. What I just described only makes sense if the Fed is equally as good at intermediating assets as the private sector. I think it's trivially obvious that the Fed is not going to be as efficient as the private sector at intermediating financial assets. If people really believed that, coupled with its superior ability to obtain funds, then the Fed should just replace all private financial intermediation. So, here is the question, which is not very obvious at all, is the Fed's reduced agency costs, coupled with its inferior intermediation skills, enough to give it an advantage relative to the private sector with its higher agency costs, but vastly superior intermediation abilities? Empirically trying to figure this out is very difficult. QE1 might have had positive effects, simply because the financial sector was an absolute mess and so perhaps agency costs were really high at that time. For example, credit spreads during the financial crisis suggest that there were excess returns to be had, which suggests there were private limits to arbitrage. This opens the door for QE1 to possibly be effective. But by the time QE2 rolled around and in the current state, are these limits to arbitrage arguments really even plausible? There isn't a lot of evidence in favor of significant excess returns at the moment, which suggest private intermediaries aren't balance sheet constrained at the moment and private arbitrage is probably working just fine.

    2. Ted, thanks for your reply. That's very helpful. I recall Miles Kimball also mentioning home bias as a friction that might render QE relevant.

      You point out that the Fed must be somehow better than the private sector at intermediating assets in order for QE to be relevant. My understanding was that the private sector simply had to be capable of reversing anything the Fed did in order for QE to be irrelevant. Perhaps these are the same points. Given the well-known short-termism problems that private fund managers face, and the Fed's somewhat sheltered existence - it really needn't worry about profits - it's easy to see why Fed purchases would not be reversed by fund managers. Why fight the Fed?

      This is just Richard's point from below:

      "Remember the famous paper, the Limits of Arbitrage – the money managers could lose their jobs easily for just a few years of poor performance from their trades."

    3. Ted clearly has a good understanding of the key issues here. We can argue that individuals have "preferred habitat" in the sense that they care more or less about payoffs at particular dates in the future. But that doesn't translate into preferences over assets, as assets are malleable. Financial intermediaries are in the business of transforming assets so that payoff characteristics are appealing to the people who hold intermediary liabilities. If we're trying to model the effects of QE, it's not good to start by assuming preferences over assets, or "home bias." Limited commitment on the part of private intermediaries might be a good place to start. Sometimes this can give you what might look like an arbitrage opportunity, but is really not one. For example, an intermediary might require an "excess return" to keep it honest (you won't behave well unless you have something to lose). That could be the basis for a theory of the term structure (which is what we need here).

    4. Ted, thanks for that comment. Very clear.

      I would say not "credit spreads" were wide during the GFC but specifically liquidity premia due to counterparty risk (in turn, due to a lemons problem). The Fed, by virtue of its irredeemable liabilities, carries no counterparty risk and therefore can arbitrage those premia. One would expect QE to be vastly less effective under normal liquidity conditions. It is surprising that Bernanke did not note this.

    5. "The Fed, by virtue of its irredeemable liabilities, carries no counterparty risk..."

      1. There is inflation risk.
      2. Not clear whether the intervention you are thinking about under "abnormal" liquidity conditions is discount window lending or QE.

    6. "But this consideration doesn't mean QE is necessarily effective because financial intermediaries can still arbitrage this away"


      What specifically is the arbitrage? If the Fed prints $1 trillion dollars and buys 10 year treasury bonds with it, then if that makes the 10 year T-bonds price go up, what exactly is the arbitrage that will appear? What exactly is bought and sold to create an arbitrage?

      And keep in mind, the classical definition of an arbitrage is absolutely zero risk, guaranteed payoff, with none of your money necessary (you borrow the funds to do the arbitrage).

      I see the arbitrage in Wallace's 1981 AER model, if the price of dollars goes up, you borrow dollars, sell them and buy state contracts that have the exact same payoff as dollars but cost less. But in the real world there are no state contracts that can exactly duplicate the payoff of any asset you want, and, people borrow at different rates than they save, and are borrowing constrained.

    7. The arbitrage is the following. Fed buys $x in Treasury bonds with $x in reserves. Given the resulting tendency for the price of the Treasury bonds to go up, some intermediaries that are holding Treasury bonds sell them, as this intermediation is becoming incipiently less profitable. Ultimately its all a wash. Private intermediaries hold $x less T bonds and have retired $x is liabilities, and the Fed holds $x more in T bonds and has issued $x more in liabilities.

    8. Stephen,
      I was referring to QE1 purchases of MBS, plus the various liquidity backstops (commercial paper, etc). These measures directly targeted abnormal liquidity premia caused by a counterparty lemons problem. They also made up for the fact that discount window lending was not getting to illiquid shadow banks.

    9. Thanks Stephen,

      I'll have to think about that, and find another paper to study in depth, hopefully a really recent one, the current state of the art as opposed to 1981.

      I do have an intuition for why QE wouldn't work, and couldn't raise the price of gold, if I might ask you about it:

      The government buys 100 million ounces of gold in a QE. The assumption is, of perfect foresight, perfect everything investors, that over the next several years, unemployment will go down and the Fed will reverse course, and then sell all of those 100 million ounces back again. Thus, the supply of gold in 10 years will be exactly the same as if the QE had never occurred. The gold just temporarily sits in government vaults (or with government ownership papers), rather than private ones, then goes back to the private vaults – No difference at all in 10 years. So, in 10 years the supply of gold is exactly the same, so the price of gold in 10 years will be exactly the same. If the price of gold in 10 years will be exactly the same, then its price today will be exactly the same, since with prefect foresight, perfect analysis, etc. investors, the today price is just the discounted 10 year price.

    10. That's a different, but not unrelated idea. The story does not have to involve gold. There's an idea that the "backing" for a money supply increase matters for its effects. For example, if the Fed issues money to finance lending at the discount window, with the promise that the money will be retired when the loans are paid off, that will have a different effect than if the Fed purchases some government debt, and then keeps rolling that over as the debt matures, so that the money circulates forever. Bruce Smith thought he had found evidence for the "backing" theory in colonial monetary regimes (again, it's an old paper - people don't think about these things much now):

    11. "The arbitrage is the following... Ultimately its all a wash."

      Still trying to grasp... Modern day QE is all about the central bank buying a fixed dollar amount of bonds with no specific price in mind, which you say will be canceled out by private sector selling. But say a central bank promises to conduct unlimited amounts of QE by buying bonds at price y (where price y is above from the market's price z). Would you say that arbitrage by private intermediaries will prevent price y from being realized, as in the first scenario?

    12. The "backing" idea is essential in Wallace 1981 AER. Condition (b) of the irrelevance proposition implies that in any "open market operation" the assets added to the Feds balanced sheet will be 100% sold back next period. This is pivotal to why the printing of more money and buying assets cannot increase inflation, or the price of any asset in the economy.

      The eventual selling back 100% of the assets added to the Feds balance sheet is pivotal and necessary in Wallace's 1981 AER, and it would seem in any paper that concludes QE can't work, not just a relatively obscure paper from Bruce Smith. This is how it looks to me at this point.

  3. Stephen,

    You have in the past cited Neil Wallace's 1981 AER article, “A Modigliani-Miller theorem for open-market operations” as an important piece of evidence for your contention that QE can't be effective (or do pretty much anything at all). I'm not the only one you gave this impression. For example, Miles Kimball coined the term Wallace neutrality in response, and wrote recently "...essentially all of my posts listed under Monetary Policy in the June+ 2012 Table of Contents are about Wallace neutrality" (at: ).

    So, being so intrigued and frustrated by your claims – for example:

    "No, in a liquidity trap, if the Fed purchases gold, it does not change the price of gold, just as it will not change the prices of Treasury bonds if it purchases them."


    I put in, what was for me, an enormous amount of my tiny discretionary time into really studying that paper, equation by equation, sentence by sentence. Here are my current conclusions:

    In the paper's model, the government's particular printing more dollars and buying an asset has no effect on the price of any asset, and no effect on inflation either. But let's look at the particulars:

    The government prints dollars and buys the single consumption good, which I like to call c's. It holds the consumption good for one period, storing it (investing it) at the return x, the same return anyone else gets for storing (investing in production) c's.

    Then, at the end of that period, it takes all of those stored (invested) c's, plus whatever return it got for them at x, and uses it all to buy back dollars at the then prevailing rate of dollars for c's.

    Now, note that Wallace does not say this explicitly, but if you study the equations and think about what they imply, you see, and can prove, that this is what must be happening. I spent a lot of time doing this (It would have been nice if he wasn't so amazingly terse and had explained/derived this – and a lot more).

    Now, in this economy, it's very simple. You either consume c's, or store them. You can buy (or sell) state-contingent contracts to get a c in a particular state next period (People, who are clones with perfect information, foresight, and rationality, only live two periods, a young period and an old period.), but those contracts are backed-up just by one thing, storage of c's (or next periods endowment, which is analogous to GDP not counting savings).

    So in this model it's simply a case that when government prints money, it's just storing c's for one period. People are going to want to store a certain amount of c's anyway, because that's utility maximizing to smooth consumption. What the government essentially does, in this model, is say, hey, store your c's with us instead of at the private storage facility. Give us a c, and we'll give you some dollars, which are like a receipt, or bond. We'll then store the c's – we won't consume them, we won't use them for anything (these are crucial assumptions of Wallace required to get his stunning results) – we will just hold them in storage (implied in the equations, not stated explicitly).

  4. Next period, you give us back those dollars, and we give you back your c's, plus some return (from the dollar per c price changing over that period). In equilibrium, the return from storing c's via the dollar route, must be equal to the return from storing c's via the private storage facility route. Or at least the return must be worth the same amount at the equilibrium state prices. So either way you go, you can arrange at the same cost in today c's, the same exact next period payoff in any state that can occur.

    Note that dollars in this model are just zero-coupon bonds. Wallace assumes no value in lowering transactions cost, in convenience, in liquidity. Transactions, liquidity, and convenience costs are zero in his model. People will hold no dollars (buy no dollars with their c's) unless their value appreciation will be equal (at the equilibrium state-prices) to if they stored their c's privately.

    So essentially, in the Wallace model the "open market operation", the QE, the printing of dollars, is just the government offering storage of c's that's exactly equivalent to what the private sector is offering, at no better a price (or maybe an epsilon better to get people to switch). So what happens? Private saving goes down, and government saving goes up by an equivalent amount. No prices change, and people's consumption in youth and old age doesn't change.

    It is analagous to Miller-Modigliani, in that if the corporation increases its debt holding, then shareholders will just decrease their private debt holding by an equivilent amount, so that their total debt stays exactly where it was, which was the amount they previously calculated to be utiltiy maximizing (and there's a lot of very unrealistic and material assumptions that go with this that have been long acknowleged as such in academic and praticioner finance)

    Now, let's get to the problems when thinking if this occurs with QE in the real world:

  5. 1) In Wallace's model, the government doesn't buy financial assets with the newly printed dollars; it buys c's, the consumption good. What if the government bought up state-price contracts for payment in state i, next period? Well, then when next period comes, if it's state i, the government will have a ton of c's that it will use to buy dollars, making dollars very valuable in that state. So there will be less state-contracts in supply for period i, but less demand for them because people will be holding dollars, which will be worth lots in state i. The government would, basically be a (costless) middleman for some of the state i contracts. So interestingly, it looks like Wallace's contentions still actually work, but I haven't proven it mathematically. But:

    2) The whole state-price paradigm has some serious unrealism problems. At first glance it looks cool. You've added randomness. People do know with certainty what each asset will pay in any state that can occur, but they only have probabilities for whether each state will happen. The problem is this is still a lot simpler and more certain than reality. If you extended this to reality, there would be a bazillion significant states, with all of the possibilities of 7 billion interacting humans in a very complicated world. And you would not know perfectly what every asset in the world's price would be in everyone of those states. So there would not be, and is not, a state price contract for each state, so that you could funge any asset into any other, and synthetically construct perfectly any asset from state price contracts.

    So referring to your quote above, Stephen, if the government bought gold and stated pushing the price up, everyone wouldn't just sell their gold and then go buy a portfolio of state price contracts that's cheaper but has the same exact payoffs in every possible state. Nothing like that exists. There's no way to synthetically construct gold.

    You could say, in Wallace's model the government will just sell that gold back again next period, and everyone knows that, but in the real world everyone doesn't know that. There are tons of investors – especially in gold – that have very little knowledge or understanding of finance. And even for those who have a lot, the government may not sell back that gold for five or ten years or more. Remember the famous paper, the Limits of Arbitrage – the money managers could lose their jobs easily for just a few years of poor performance from their trades.

    There's more, when I have time.

  6. professor Williamson, do you believe that the fed's policy of interest on reserves have affected the TIPS low?
    Do you also believe, lowering a .25% interest on reserve return would have a more than negligible impact on our economy?

  7. The payment of interest on reserves is important for how monetary policy works. This makes it possible to have a large stock of excess reserves outstanding, with the interest rate on reserves determining the overnight interest rate. However, with the interest rate on reserves at 0.25%, that interest rate does not matter much for TIPS yields, relative to an interest rate on reserves of zero. Which answers your second question I think. Lowering the interest rate on reserves to zero would have little effect. The reason the Fed does not do this is that, rightly or wrongly, they are worried about the profit margins of money market mutual funds.

  8. "We can find plenty of people - most of them in the Federal Reserve System - who are now willing to vouch for the fact that QE announcements move asset prices in the "right" direction"
    You sound skeptical. Are you skeptical about whether QE announcements move asset prices, or whether those moves are in the right direction?

  9. I'm skeptical that it's the asset swaps that are having the effect. I think it's whatever the announcement signals about the future policy rate.