Tuesday, December 3, 2013

The Intuition is in the Financial Markets

I thought I would take another stab at giving Paul Krugman some intuition about what I have written about here, here, and here.

Krugman quotes something Andolfatto wrote, and then writes:
OK, so “agents require” a fall in the inflation rate to induce them to hold more currency. How does this requirement translate into an incentive for producers of goods and services — remember, we’re talking about stuff going on in the real economy — to raise prices less or cut them? Don’t retreat behind a screen of math — tell me a story.

So, let's take that seriously. We'll tell a story, but telling a story about pricing in goods markets is not going to get at the essence of the problem. For Krugman, and for many macroeconomists, pricing is indeed the essence of the problem - their problem, not mine. Five years ago we went through a financial crisis, and we call it that for good reasons. Most of us are in agreement that the problem was financial. Thus, if we're looking to understand the crisis and its aftermath, we should start by thinking about financial markets. Financial market problems can indeed create problems for how goods markets and labor markets function, and that's part of what we want to understand.

Here's the story. No symbols. No equations. No numbers even.

Let's start with basic things we know about financial markets, neglecting risk for now. Savers who hold assets care about the rates of return on those assets. Indeed, what would seem to have to be true in a world without risk is that the rates of return on all assets would equalize - otherwise who would want to hold the low-return assets? But, we observe that, in the real world, different assets have different rates of return. Risk can explain some of that, but there are some assets - take currency and U.S. Treasury bills, for example - which appear to have exactly the same risk characteristics, but typically have different rates of return.

How to explain rate-of-return differentials? We might suppose that people just like some assets better than others. For example, maybe they enjoy holding General Motors stock more than they enjoy Microsoft stock. This might help us explain the rate of return on housing, relative to other assets, for example. If I own a house and live in it, I get an implicit flow utility return, and this will make me willing to hold a low-return house as an asset rather than high-return General Motors stock. Of course, I'll need to be concerned about the fact that it's possible to invest in housing as an asset and rent it out. In any case, trying to explain some rate-of-return differentials by positing a flow utility return from the asset won't be very satisfying in most contexts. For example, that won't produce any revelations if we're trying to explain why the rate of return on T-bills is typically higher than for currency.

But economists think about another factor - often expressed in vague terms - which helps to explain the rates of return on assets. That other factor is liquidity. So what is it? Liquidity is some extra value we get from an asset - a payoff not captured in it's measured rate of return - that comes from its acceptability in exchange for other assets or consumption goods, from its use as collateral in credit contracts, or for other reasons. We can all understand that currency is more liquid than T-bills, in the sense that no one accepts T-bills in retail transactions, but currency is widely accepted. However, there is a sense in which T-bills could be more liquid than currency in particular uses. For example T-bills are much more convenient for use as collateral in overnight repurchase agreements. Ultimately, it seems liquidity is a matter of degree. It seems hard to separate assets into the liquid and not-liquid. In one sense, a house is illiquid, in that it is hard to sell - not much use in making retail transactions, for example. In another sense, it has some liquidity value because homeowners can take out a home-equity loans using their houses as collateral.

We can conclude, though, that if the rate of return on asset A is higher than the rate of return on asset B, then part or all of the explanation for this is that asset A is less liquid than asset B. In other words, the two assets can carry liquidity premia - implicit flow liquidity returns - and the differences in rates of return in asset A and asset B could be all or partly due to differences in liquidity premia on the two assets, with the liquidity premium on asset B being higher.

So, those are some things we know about the determinants of rate-of-return differentials. What explains the average rate of return or, put another way, the aggregate rate of return on assets in an economy? That's just supply and demand logic. The demand for assets comes from the willingness of savers to hold those assets, and supply is determined by the actions of firms, financial intermediaries, and the government. Given demand, an increase in the supply of available assets will cause the rates of return on all assets to go up, so as to induce savers to hold those assets.

It's generally recognized that the effects of central bank actions in financial markets have something to do with liquidity. A conventional open market purchase is basically a swap of money for short-term government debt. Under conventional conditions, the rate of return on short-term government debt is higher than the rate of return on money (the nominal interest rate is positive), and that has to be solely due to the liquidity premium being higher on money than on short-term government debt, as money and short-term government debt have basically the same payoffs - they're equivalent in terms of risk. So under conventional conditions (a positive nominal interest rate) an open market purchase increases the average liquidity of assets in financial markets. How that matters is of course the subject of an enormous amount of theorizing and empirical work.

To get to the point, let's think about what happens in a liquidity trap. What's that? For some reason, possibly engineered by the central bank, the economy is in a state where the rates of return on money and short-term government debt are the same. The nominal interest rate is zero. Further, the rate of return on both of these assets is equal to minus the inflation rate. What is happening with liquidity premia in this liquidity trap? In the trap we are currently in, one can make the case that the aggregate stock of liquid assets is relatively small. Treasury debt and the liabilities of the Fed are liquid assets, but there are other liquid assets that are created by the private sector. Such private liquid assets include the liabilities of financial intermediaries and asset-backed securities. The financial crisis acted to reduce the supply of these private liquid assets, and we can also argue that there was an increase in the demand for U.S. liquid assets, coming from the demand abroad for our Treasury debt. Thus, in the liquidity trap we are in, the liquidity premia on money and short-term government debt are the same, and positive.

Next, conduct a thought experiment. What happens if there is an increase in the aggregate stock of liquid assets, say because the Treasury issues more debt? This will in general reduce liquidity premia on all assets, including money and short term debt. But we're in a liquidity trap, and the rates of return on money and short-term government debt are both minus the rate of inflation. Since the liquidity payoffs on money and short-term government debt have gone down, in order to induce asset-holders to hold the money and the short-term government debt, the rates of return on money and short-term government debt must go up. That is, the inflation rate must go down. Going in the other direction, a reduction in the aggregate stock of liquid assets makes the inflation rate go up.

But in a liquidity trap, since money and short-term government debt are equally liquid, if the central bank swaps one asset for another then this has no effect. That's why we call it a liquidity trap. Similarly, suppose that short-term government debt is more liquid than long-term debt. To be more precise, short-term debt may be a more preferable asset to use as collateral, perhaps because its market price is less volatile. Then, even though we are in a liquidity trap, quantitative easing (QE) - swaps by the central bank of money or short-term government debt for long-term government debt - will increase the effective stock of liquid assets in the aggregate. Just as I outlined in previous paragraph, that has to reduce the inflation rate.

What's the qualification? There are various short-run effects of monetary policy that could come into play. However, I think it's fair to argue that any of those short run effects have played themselves out in the financial crisis and its aftermath, and now we're looking at the effects I've described.

Done.

88 comments:

  1. ^ Hey bro, what the eff are you talking about? Why wouldn't the rate of return equalize demand and supply of financial assets?

    ReplyDelete
    Replies
    1. Hey bro, what the eff you thinkin? Where ya get da idee I ain't sayin dat.

      Delete
    2. It is a lot more complex than that.

      It is hard to conceive the thought experiment of increasing the supply of financial assets without changing the budget constraints of the agents in the economy. Therefore the supply and demand analogy is inappropriate.

      Delete
    3. You're never happy. If I give you the technical story, you complain. If I give you words only, in simple economics, you complain. I'm sure if I said the sky was blue, you would say it's green.

      Delete
    4. Mein gott, you give it up too easily!

      And by the way, deleting substantive posts rebuking your argument does not smell exactly like a profile in courage. Besides, the argument is not going to go away.

      Delete
    5. You're not giving me a substantive argument. You're just having fun being disagreeable.

      Delete
    6. I just read your Economic Record paper. I find the idea of a fixed proportion of transactions in the decentralized market silly. Most of your conclusions stem from that assumption. No wonder they are "counter-intuitive" (euphemism for non-sense).

      Delete
    7. "Most of your conclusions stem from that assumption."

      Explain, in detail.

      Delete
    8. Your model rests on the assumption that there is some deep parameter that rules the proportion of transactions that happen in the monitored and non-monitored decentralized market (DM). I find that silly. If an economy is awash in cash and with little interest bearing liquid assets, it will self-organize so more transactions will occur in the cash side of the decentralized market. It must not be hard to add that feature to your model.

      Delete
    9. OK. So add that feature, work it out, and tell me what you get.

      Delete
    10. Sincerely, I have more useful things to do than to tinker with a model that I consider silly to begin with.

      Delete
    11. So, you're telling me you know it's wrong, but you haven't bothered to figure out how it works? That makes a lot of sense.

      Delete
    12. No, I am telling you that your modeling strategy is silly; and so is your model for money and transactional technology. If that was not enough of a waste of time, it appears that you built your model to match stylized facts that are not facts. Oh boy.

      Delete
  2. Thank you for great posting!

    Your argument is convincing, but in terms of quantity, it depends on the degrees of substitution effects between current consumption & saving and between saving instruments with different degrees of liquidity.

    I believe the latter effect is larger than the former, but still have some reservation whether the size of the effect is significant.

    JS

    ReplyDelete
    Replies
    1. Yes, correct. If you're interested in the technical things, you can read the papers (see the three posts linked above) for the details.

      Delete
    2. I should add that "substitution effects between savings instruments" can be more awkward relative to how we think about complements and substitutes in terms of goods and services. But sometimes we can get some clear idea of assets being complements or substitutes in some rough sense.

      Delete
  3. Steve: "Since the liquidity payoffs on money and short-term government debt have gone down, in order to induce asset-holders to hold the money and the short-term government debt, the rates of return on money and short-term government debt must go up. That is, the inflation rate must go down."

    I was following you fine up to that point.

    Start in equilibrium. Then helicopter some more money on the economy. People don't want to hold more money. At the margin, the value they place on liquidity has gone down. So they try to get rid of the extra money by spending it. The excess demand for goods causes the price level to rise. that causes the real value of the stock of money to fall back to its original level. We are now back in equilibrium at a higher price level.

    Helicopter money does not cause deflation.

    ReplyDelete
    Replies
    1. Good. That's important. "Issue more debt." That's in real terms.

      Delete
    2. To expand on that: The Treasury does not issue money. The Fed does not have a helicopter.

      Delete
    3. This seems more significant to me than it might be. My apologies if I am just misunderstanding things:

      Stephen is claiming that the the mechanisms of QE, specifically, can cause deflation in a liquidity trap. But he makes no claim regarding 'helicoptering in' money.

      If this is the case, Stephen's result actually seems fairly plausible to me. (And makes some empirical/intuitive sense.)

      I'm still a little... wary of the paragraph Nick quotes. It feels like something of a black box.

      Delete
    4. The Fed+Treasury can helicopter money.

      But you get the same result if there are helicopter bonds, even if they are real (indexed) bonds.

      The public has a demand for B+(M/P). The Treasury does a helicopter increase in B. There is now an excess supply of B+(M/P) at the existing P. So people try to buy goods. So there is an excess demand for goods. So P rises to restore the original level of B+(M/P). And we are back in equilibrium.

      Delete
    5. Do inflation expectations play a role here?

      I agree with Stephen that the Fed does not do helicopter drops, but merely changes the liquidity profile of money+Treasuries in the hands of the public. That does not resolve the issue of whether inflation must go up or down, as far as I can see.

      Stephen says: "To restore equilibrium in the money market, P must go down." Nick says: "To restore equilibrium in the money+bonds market, P must go up."

      If the public expects inflation to go up, money demand declines and Nick's high-inflation equilibrium is realized. If the public expects disinflation (or deflation), they will hold on to their money balances and deflation will happen.

      Can the central bank nudge the market into one equilibrium or the other by shaping inflation expectations?

      Is this crazy? Am I way off track?

      Delete
    6. That is not crazy. Expected inflation is p(t+1)/p(t) and p(t) is in large part predetermined because of menu costs, long-term contracts etc. The key is to model the expectations of inflation.

      When we go back the main equation of Professor Williamson’s model:

      Ep(t+1)/p(t) = f(L)

      The right way to look at that is:

      L = h(Ep(t+1)/p(t))

      That is, the causality runs from the right hand side to L. When inflation expectations are high, agents economize on their cash holdings to the point where the marginal benefit of cash is equal to the inflation tax. Liquidity is an epiphenomenon.

      Delete
    7. So would you agree then that the equilibrium selection (i.e. higher inflation or lower inflation) is a function of inflation expectations, and that those expectations are (at least partly) determined by the central bank?

      Delete
    8. Yes, I would agree. Equilibrium selection is where one will find "the action". It depends on lots of stuff, from monetary and fiscal policy to economic institutions and social norms.

      Delete
    9. NIck Rowe and Steve: If the liquidity premiums have gone down and there is a helicopter drop either of bonds or money, then people would either not pick them up or or pick them up and store till their value increases. If the asset does not have storage cost and there is a possibility of its value changing in future why would people spend it right now to get rid of it. So I think in this case even a helicopter drop would be deflationary or disinflationary.

      Delete
  4. Stephen, bro, sir, Anonymous 5:43 here. That crazy anonymous Italian guy said something about supply and demand for financial assets being different from other goods. I guess he deleted it. You were collateral damage

    ReplyDelete
    Replies
    1. I am Brazilian; what I said is pretty standard mainstream economics (so not crazy); and Stephen was the one who deleted it.

      Delete
    2. What I think anonymous means is that you typically start your comments with "you are wrong," then go on to say something incoherent. If you become a negative exernality, I have to do something about it.

      Delete
  5. "So they try to get rid of the extra money by spending it. The excess demand for goods causes the price level to rise."

    No, because the fiscal authority determines the price level. It can raise taxes to absorb the "extra money" if it wants to keep the price level constant. The critics have missed this point, which is why they incorrectly raise issues about the direction of causation, instability, etc.

    I think this debate boils down to a disagreement over what is held constant.

    ReplyDelete
  6. Steve,

    Equilibrium conditions don't provide a sufficient story for causality. Sure, the marginal value of cash holdings must equal the inflation rate in equilibrium (b/c nominal rates are at 0), but that could happen either through a falling inflation rate (Your story), or a drawdown in holdings of cash (the monetarist story). I don't think your dynamic story of how you get to equilibrium makes any sense. I wrote all of this up here:

    http://synthenomics.blogspot.com/2013/12/why-dynamic-stories-are-important.html

    ReplyDelete
  7. You still haven't responded to the Krugman objection. You've given it a good faith try, that's for sure, and it's quite a smart post. I think there's a methodological issue at work here, one that runs deeper than any specific model.

    Consider this excerpt from your post::

    "the liquidity payoffs on money and short-term government debt have gone down, in order to induce asset-holders to hold the money and the short-term government debt, the rates of return on money and short-term government debt must go up. That is, the inflation rate must go down."

    Look at that closely. You spend 1000 words or more on the background to your story, setting up all the relationships, etc. But when it comes time to tell the actual story -- that is, *how does it happen that prices change** -- there are no words, no intuition, no story. Just the invocation of a definitional identity. You're comfortable with that. Krugman, I think Nick Rowe, and others are less so.

    I suppose one could describe the issue in terms of big fancy phrases like "the ontological status of economic semi-observables." LOL. I know there are philosophers of math and philosophers of science? Are there philosophers of econ?

    The way I see it, what happens on the ground is more real than what happens in the clouds. We can observe a merchant raising prices, and understand why he might do so in light of his cost structure, demand functions, and so forth. When we talk about concepts like "real interest rates," we're in the clouds. You can't observe a real rate, not directly, for two reasons: first, it is derived from observable rates and calculated inflation, and second, there is never really "one rate." There are rates on assets, which we relate to each other with models. Maybe the models are good (I have my doubts about some of the simple but widely-used credit spread models), maybe they aren't , but they're models.either way.

    I think another way of phrasing this point is the old chestnut about macro theories being micro-founded. I call it an old chestnut because it's not at all clear what "micro-founded" really means, and I think this debate is bringing the slipperiness of that concept to the forefront.

    What I see in this model is an effort to derive an observable (i.e. price changes) from a set of unobservables (e.g.. liquidity premia). From a theoretical perspective, that's always methodologically suspect -- one that we nonetheless might think of as a useful approximation that we use because economics is hard enough as is. Still, when we do that, we have to sanity check the results -- that's the function of "the story." So when there isn't a sensible story to tell, the results should be viewed with suspicion, in my view.

    On this larger point, it's important, I think, that you realize that you don't have a story in the sense that Krugman and DeLong want.. You have an equation. You have no way to explain how the changes in inflation come about -- just that they do, because the definition of real interest rates says so.

    I think part of your response is that stories aren't important: you've said so before, and here, when pressed for a story, you've given something that looks the part but isn't. I disagree with Brad DeLong's view that the lack of a story means you aren't a card-carrying economist, or whatever silly and disparaging language he used. I mean, maybe he's right, and that models without satisfying stories are useless. But that depends on whether his underlying point is correct, and he seems to think it's obvious beyond debate. I don't think so.

    Still, I think there needs to be some accounting for the relationship between the clouds and the street. If not a story, some way of translating abstract relationships to predictions about things that happen in the world.

    ReplyDelete
    Replies
    1. "You still haven't responded to the Krugman objection. "

      Some people are never satisfied.

      Delete
    2. That is because you did not respond to his objection.

      Delete
    3. When the objection is incoherent, there's not much to say.

      Delete
    4. Incoherent is to state that inflation adjusts to validate the "liquidity premium". That is lack of basic understanding of macroeconomics. That is the tail wagging the dog. You inverted the causality chain. The liquidity premium is a function of the expected inflation. When expected inflation is high and the negative real return on cash is large, agents will reduce their real cash equivalent holdings (which they can easily do because money supply is endogenous). That is valid regardless of the liquidity trap.

      Delete
    5. Some people are never satisfied? That's your response?

      I'm giving you credit for a thoughtful post, and I'm not rejecting the Delong view that there's obviously something deficient with your analysis. I'm chalking it up to a methodological disagreement.

      If I'm incoherent on this point, then so is Karl Smith, I suppose. My views track what he's written about immaculate inflation, not to mention what many others have written (here and elsewhere).

      I am frequently "satisfied." Do you think your long blog post should *automatically* quell objections, simply by the fact that you wrote it? If not, then you should address the substance. Obviously you don't have to do that here, or in response to me, but you should do that at some point.

      It's not just me who had issues with that paragraph. It's leaped off the page to many.

      Delete
    6. Karl Smith is incoherent. Glad I could help.

      Delete
  8. It seems like another way of summarizing Stephen's view (or at least my interpretation of the empirical results Stephen is alleging) is two different views of the same actions - but with different intents:

    Is the FED's QE a pump (pushing cash into banks) or is QE a faucet offering cash only as banks ask for it?

    The NK interpretation is the latter, and that is how open market operations proceed - banks say, "Hi FED, here's some long-term bonds I'd like to exchange for cash." But while banks initiate the transaction, they take the initiative to execute those transactions out of a compulsion derived from market forces. Thus, the mechanism of QE acts like a pump.

    Recently, Krugman posted a chart of Japanese bank reserves over the decade, showing that in a liquidity trap, they increase. This is empirical evidence that QE is a de facto *PUMP* of cash into banks.

    Stephen is essentially building off this and saying that it matters because a banker being compelled to take on cash (and receiving IOER) will offer loans on those reserves at a higher rate than she would if she was passively choosing to accept the cash because she knows she can loan it out at a greater profit.

    So I see the disagreement hinging on which resource is scarce - customers seeking new loans or bankers with so much cash they can be picky about new debtors. The fact that reserves pile up in a liquidity trap seems to imply that we have more bankers with cash.

    Seen as a flow, the FED is pumping, the banks are accruing a reservoir of cash - but they are exercising a kind of monopoly-rent on the supply of new loans, starving the consumer-side of the flow for cash. That lack of cash on the consumer-side makes cash more valuable to the part of the market that "matters." On the consumer side there is a liquidity constraint akin to Krugman's Baby-Sitting Coop - where the flow (or velocity) of cash is constrained by this lack of available credit.

    The banks have this "monopoly-rent" power because of their "legal" position as middle-men between FED and consumer. If this relationship wasn't imposed - if, for instance, a consumer could exchange a 10-year debt for cash just as easily as a bank, then the ability of banks to control the flow would be undermined.

    This is just ONE way of viewing what Stephen is saying - though I'm not sure I agree 100% with Stephen. But an alternative way of visualizing the system is Gresham's Law (long-term bonds are the over-valued commodity that a price-setting agent has induced the market to eschew in favor of cash - resulting in appreciation, not inflation).

    ReplyDelete
  9. Bravo Stephen! This post is a remarkable effort to broaden “the discussion” and it succeeds, for me at least.

    Your definition/explanation of “liquidity’s extra value” in terms of “its acceptability in exchange” and/or its use “as collateral,” seems too narrow. People sometimes sell bonds (for money) because they believe bond prices will fall, or because they’re less confident in their forecasts. Are they necessarily searching for a greater volume of easy payment facilities and/or additional collateral?

    What explains the different rates of return on assets A, B, etc.? You answer, “if the rate of return on asset A is higher than the rate of return on asset B, then part or all of the explanation for this is that asset A is less liquid than asset B.” Stocks have had higher returns than bonds for a long time, but neither is typically used as collateral, and both can usually be easily “liquidated,” i.e., sold for “money.”

    “The demand for assets comes from the willingness of savers to hold those assets . . . Given demand, an increase in the supply of available assets will cause the rates of return on all assets to go up, so as to induce savers to hold those assets.” The last sentence is a truism, but it only packs a punch if we assume this increase in the supply of financial assets occurs with no change in the demand for financial assets. Perhaps, but what’s the “story” behind state of affairs?

    “Suppose that short-term government debt is more liquid than long-term debt. To be more precise, short-term debt may be a more preferable asset to use as collateral, perhaps because its market price is less volatile.” Short-term debt “may” be better collateral, “perhaps because its market price is less volatile.” Yes, quite.

    “There are various short-run effects of monetary policy that could come into play. However, I think it's fair to argue that any of those short run effects have played themselves out in the financial crisis and its aftermath, and now we're looking at the effects I've described.” Begging the question, no? Assuming the “short run effects have played themselves out” assumes we can ignore actions taken when the system is out of equilibrium, which was one of the main points of dispute.

    Leaving these reservations aside, I thank you for the “translation” and don’t begrudge you the “done” at the conclusion of contribution.

    ReplyDelete
  10. At the end of the day this model still runs counter to the empirical facts, namely that QE creates (at least some) inflation. Theories that do not explain a piece of the real world are fun to do but essentially worthless.

    ReplyDelete
    Replies
    1. "At the end of the day this model still runs counter to the empirical facts, namely that QE creates (at least some) inflation."

      That's not a fact. That's a hypothesis.

      Delete
    2. Actually, it is more than a hypothesis. It has empirical support as shown here: http://macromarketmusings.blogspot.com/2013/12/taking-model-to-data.html

      Delete
    3. In order to properly confront the data, we need a model of how QE works, and then we have to argue that the data is somehow consistent with that. I don't think we would call that serious empirical work in that sense.

      Delete
    4. Of course estimating a bunch of VARs is not superdeep empirical work, it's quick and dirty back of the envelope empirical work. And it is also totally sufficient to reject a theoretical idea which is counter to Macro 101.

      Not that I am against ideas which run counter to what we have known for decades but if you propose something which is theoretically dubious you might wanna take a quick look at the data to check whether your theoretical story roughly matches the data.

      Delete
    5. Stephen, Beckworth's post was a direct response to your claim that inflation was being lowered over the past three years because of QE. You are the one that pointed to simple inflation chart. He stepped up and showed your claims are rejected empirically. Have some courage and accept it.

      Delete
    6. No, that's not what I said. Read my first post on this again, then read my comment on Beckworth's post.

      Delete
  11. "Most of us are in agreement that the problem was financial."

    Why? What about the housing bubble? I agree with Dean Baker that what ails us is the lost demand from the housing bubble not a "financial" crisis.

    See this link:

    http://krugman.blogs.nytimes.com/2012/10/20/the-financial-industry-and-financial-crises/?_r=0

    If your solution is to issue more government debt, then doesn't that entail the government spending more money?

    ReplyDelete
    Replies
    1. "What about the housing bubble?"

      So, tell me about that. What was the housing bubble, what caused it, etc.?

      Delete
    2. Have you read this?

      http://www.cepr.net/index.php/publications/reports/the-run-up-in-home-prices-is-it-real-or-is-it-another-bubble/

      Delete
    3. That's not an answer. I want you to tell me about it. Put it in your own words.

      Delete
    4. Stephen, I'm not an economist. I'm not sure how you expect me to explain it better than Baker did in his paper.

      It's a short paper, but I'm willing to put together an answer if you're genuinely interested, but if you already have some objection in mind, I'd rather not bother.

      Especially since you can't even say whether you've read it or not.

      And note, you never said why you "agree the problem was financial" other than to appeal to consensus.

      Delete
    5. Here I will have to side with Prof Williamson. Dean Baker is more of an activist than an economist. While exceptions are possible but rare, as a general rule you gain nothing and may even cause damage to your neural networks by reading stuff from cepr.net

      Delete
    6. Ad hominem? I don't see the value in your comment Anonimo.

      I've gained nothing from it and suspect it damaged my neural network.

      Delete
    7. "Stephen, I'm not an economist. I'm not sure how you expect me to explain it better than Baker did in his paper."

      I can't read everything. I read some Dean Baker once and thought he didn't have much to say. Obviously you thought it was convincing - there was some sense in which you understood it - so you should be able to articulate it.

      Delete
    8. Stephen, is their a difference between reading my comment and reading his paper?

      But again, I'd be willing to get into it if I had more of an idea about where you're coming from.

      You said you agree (why?) with everyone else (who?) that the problem was (is?) financial?

      I'm not even sure what you think the problem (unemployment? idle resources?) is or why it's important.

      All I (think) I know is you (maybe?) want the government to issue more debt (do you want them to spend more too or are they raising taxes in lockstep in your model)?

      Anyway, wasn't the housing bubble crash the event that strained the financial system? That led to the bailouts?

      http://www.cepr.net/documents/publications/housing_fact_2005_07.pdf

      Delete
  12. "Since the liquidity payoffs on money and short-term government debt have gone down, in order to induce asset-holders to hold the money and the short-term government debt, the rates of return on money and short-term government debt must go up. That is, the inflation rate must go down. Going in the other direction, a reduction in the aggregate stock of liquid assets makes the inflation rate go up."

    This is the part where Krugman wants more intuition.

    Here's a question to focus on: what would happen if the inflation rate DIDN'T adjust following a fall in the liquidity premium?

    Presumably if the real return on liquid nominal assets is too low, agents will sell them and buy real assets or consumption goods. We normally think of higher demand as bidding UP prices of real goods and causing inflation.

    But you are claiming that instead we get deflation. How does this happen?

    ReplyDelete
    Replies
    1. Good post. Succinct and to the point. This is the question everyone is asking.

      Delete
  13. I think I agree with most of the previous comments: very nice post, but does not really answer the question.

    I think was clear from the previous posts that inflation is determined by an asset pricing condition. The question is how this works exactly.

    Let me try to rephrase the question, perhaps this will clarify things: how do firms know they have to decrease they price after the increase in the collateral? Does the wage change? Does something happen with labor supply?

    I think that understanding these questions would help us understand your model better and, perhaps, trust more the answers it provides.

    ReplyDelete
    Replies
    1. No, you didn't get it. All the other stuff is working in the background of course. There's a labor market, a market-clearing real wage, and in the notes I posted firms are selling goods in exchange for assets (a stand-in, if you like, for people buying goods with bank deposits which are backed by assets). That's all just normal stuff. But look at the title of the post. That says it all.

      Delete
  14. That was a useful explanation, if anything so we can identify the ways you are wrong.
    Good to start with the assumption that we are in the real world. First, the nominal return on currency is zero. Second, the nominal return on short-term risk-free assets is r, a number determined by the central bank (unless your country is called Zimbabwe, does not have its own currency etc).

    Hence, in the real world the difference between the return on T-Bills and currency is not determined by a liquidity premium. It is the other way around. The difference between the return on T-Bills and currency determines a liquidity premium, which is validated by agents’ portfolio decisions and accommodated by the banking system. For instance, if the liquidity premium is too high, agents will economize on cash holdings etc.

    So, now say we are in a liquidity trap. The short-term interest r is equal to zero (because so says the central bank). The long-term interest rate is determined by properly discounted cumulative short-term interest rates plus some liquidity service (or preferred habitat effects). A swap between short-term bills for long-term bonds reduces the yield on the long-term bond due to a preferred habitat effect, so it lowers the long-term interest rate. The liquidity yield of T-bills and currency will in equilibrium be equal to the inflation rate (*) as agents adjust their portfolio in order not to have too much or too little of currency and T-bills.

    (*) which is, obviously, determined in the goods and labor market and largely sticky due to way real life people and firms enter into contractual relationships etc.

    To the extent that QE affects output (say, because it increases the value of illiquid wealth, alleviating borrowing constraints of households), it may set in motion demand-side pressures that would cause prices to rise.

    ReplyDelete
  15. Did you see Nick Rowe's update of his post, with the quote by Bob Murphy?

    worthwhile.typepad.com/worthwhile_canadian_initi/2013/12/helicopter-money-does-not-cause-deflation.html

    ReplyDelete
    Replies
    1. I'm not going to try to parse that.

      Delete
    2. Parse what? The URL or that quote from Bob Murphy?

      Delete
  16. Steven, you should try explaining it again as plenty of people still have no idea what you're talking about.

    Give it one more try, please.

    ReplyDelete
  17. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. (fixing typos)

      There is something a little peculiar about saying that increasing the supply of something (money) increases its price (deflation).

      Saying that in order to induce people to hold money, there must be a prospect of deflation, begs the question of how printing more money would create that prospect of deflation.

      Otherwise the more money one prints, the more deflation there is, until logically one prints an infinite amount of money and its value is unbounded.

      Rather, one might say that the printing of money and a credible inflation target induces people to consume and purchase real assets, leading to the desired level of inflation. Why does this not happen, and what is the alternative story that leads to deflation via less demand/more supply?

      At a minimum, it would seem that having the central bank purchase bonds from the government which then spends on real goods and services would counter any deflationary tendency and have the desired stimulative effect...perhaps this is what you are advocating?

      Delete
  18. "there is a sense in which T-bills could be more liquid than currency in particular uses. For example T-bills are much more convenient for use as collateral in overnight repurchase agreements."

    Do you not know what a repo is? What would be the point of using currency as collateral to borrow currency in the overnight market?

    ReplyDelete
    Replies
    1. It's not so silly. Suppose I'm a pension fund, and I have some "cash" to park overnight. Of course it's not literally currency, it's an account balance with some financial intermediary. Then, suppose that another financial institution offers me an overnight repo contract. A repo is essentially a secured loan that I am going to make to that financial institution. So, suppose I am going to lend $99 million to the financial institution overnight with the promise of repayment of $100 million the next day. The financial institution offers me a choice between two alternatives: (i) the collateral will be $100 million in Treasury bills; (ii) the collateral will be $100 million in currency that the financial institution has in the vault. Question: Which deal will I prefer?

      Delete
    2. Why would an institution with $100 million in the vault want to borrow $99 million from you overnight at a positive interest rate, pledging the $100 million as collateral? It will end up tomorrow with just $99 million in the vault. Makes no sense at all.

      Delete
  19. "... the economy is in a state where the rates of return on money and short-term government debt are the same. The nominal interest rate is zero. Further, the rate of return on both of these assets is equal to minus the inflation rate. What is happening with liquidity premia in this liquidity trap? "

    This bit still doesn't make sense to me. If the nominal interest rate is 0 (to make things more concrete, say we're talking about the overnight rate), then central bank reserve balances are providing no liquidity services on the margin. Banks will rent balances out overnight and accept a 0% return since they won't be foregoing any valuable liquidity services.

    In this state, there is no liquidity premium on money. There can't be. If there was, then the nominal rate, the overnight rate, would be positive, indicating that owners of central bank balances required a return to compensate them for foregone liquidity services. But then we wouldn't be at the zero lower bound.

    So in sum, the scenario you envision of zero nominal rates and the existence of a liquidity premium can't happen.

    ReplyDelete
  20. Steve, isn't this model very very, deeply non-Keynesian by assuming that the liquidity premium is independent of the level of interest rates? In other words, it has a precautionary and transactional demand for "liquidity serivces" but no speculative demand - in other words, no liquidity preference. I think this matters a lot as your model very much needs the liquidity premium on money to be constant when QE is carried out, but actually there is good reason to believe it will rise - if you know that there is a large holder of long dated government bonds who is committed to selling at some time in the future, then the value of being in cash rather than bonds is definitely bigger, particularly when yields are close to zero, and so your upside in the long-govies trade is capped and your downside at its greatest.

    ReplyDelete
  21. I finally have a chance to get to this. Thank you for your efforts to convey the intuition. So far, I'm up to:

    "Thus, in the liquidity trap we are in, the liquidity premia on money and short-term government debt are the same, and positive."

    I think you're missing something here. The liquidity premium doesn't change in the liquidity trap. It's still just as inconvenient and costly to buy something when you have T-bills, compared to when you have money. Money is still better liquidity-wise, and just as risk-free. But then why don't T-bills have a higher interest rate than money? I'd say because money is only just as risk-free as T-bills up to relatively low amounts. Then it becomes riskier than T-bills. Money is only risk free in a bank account up to $250k per bank. Anything over is riskier than T-bills. And of course it's risky to hold a lot of cash – look at the gigantic premium drug lords pay to move their wealth out of cash, although that's largely to avoid law enforcement. It's just if you want risk-free, or "risk-free", and you have a lot of wealth, money is *not* an option.

    So there's an incentive to move large amounts of money, to not let them sit. They're not fully safe in a bank account, so there's an incentive to spend them, or to put them in non-fully-safe assets other than a bank account, if you're not going to put it into T-bills, or if you just sold T-bills.

    ReplyDelete
    Replies
    1. There is another explanation why a zero nominal interest on TBills does not imply they generate the same liquidity yield as cash. The return on cash is zero. The return on a short-term TBills is whatever the central bank decides (unless there is default risk or other complications that do not apply to the case in point). If the monetary policy committee decides to peg the return on TBills at 0.25 or 1, surprise surprise, there is now a gap between the return on TBills and cash. This gap has a name: short-term interest rate; it is a policy variable and not an equilibrium outcome.

      Delete
  22. What if the Fed made T-bills and then sold them for T-bonds? The interest rate on T-bonds would go down, and the interest rate on T-bills would go up. And the "liquidity premium" (This seems more like the liquidity-and-risklessness premium.) would go down. It would be less desirable to hold money. So, two things could happen, it seems, not just one:

    i) Inflation could go down to induce people to hold money as much as before.

    or

    ii) Inflation could not go down, and people would want to hold money less, and so would get rid of it more quickly by spending it faster. It would become a hot potato, or more of a hot potato. In fact, I pretty much consider money a hot potato now. The "liquidity premium" is not high enough for me except for relatively very small amounts. Anything more I spend very quickly, mostly on stocks or real estate.

    ReplyDelete
  23. I'm kind of confused. How can money and t-bills be perfect substitutes, yet still be exchanged for one another. If people are indifferent between holding T-Bills and holding money, then how is the treasury going to sell its new issues? Why would I bother? Similarly, who is going to sell their T-bills to the Fed if what they get in return is no better than what they are giving up? Why would anyone do a repo? Tbills and cash are exchanged every day, probably several times a day (I would assume. I have very little of either so I can't be certain), so they can't be perfect substitutes.

    The treasury doesn't need to convince people there will be deflation to get them to take the extra treasuries, it just needs to find people that want treasuries instead of money, and there are piles of them, as evidenced by the fact that they are exchanged constantly. Now the stock of money plus bonds has gone up. The public doesn't like the return it's getting on money? Tough shit, they have it now. The only way for them to get rid of their stock of money is to exchange it for goods or services or invest it in something that isn't money or T-bills. And wasn't that the point in the first place?

    ReplyDelete
  24. Nick Rowe has a good explanation of why this is all nonsesnse:

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/12/does-house-building-cause-house-price-inflation-the-sokal-hoax.html

    To be more specific, Nick explains, in a way I could finally understand, what the fallacy is that allows Williamson to conclude that expanding the money supply can be deflationary. I won't attempt to re-create the logic here, except to say that Williamson's logic effectively requires that markets expect that not only QE will end, but when it ends, the money supply will be smaller than it was before QE started.

    Kenneth Duda
    Menlo Park, CA
    kjd@duda.org

    ReplyDelete
    Replies
    1. Nick Rowe doesn't have a good explanation for anything.

      Delete
  25. "But in a liquidity trap, since money and short-term government debt are equally liquid, if the central bank swaps one asset for another then this has no effect."

    But money is not equal to short term-debt in "liquidity-and-risklessness", and it seems that what you call "liquidity" is actually "liquidity-and-risklessness", because there are assets that are just about as liquid as short-term government bonds that are risky, or very risky. Stocks, it seems to me, are just about as liquid as T-bills, in fact, about as liquid as you can get. The sale is almost instant, with close to zero sales cost, and at the market rate.

    So, it seems like the characteristic at issue here is "liquidity-and-risklessness". And with regard to that, short-term government bonds and money are significantly different. Because short-term government bonds are riskless for as many as you own. With money, the risklessness is limited or costly. If a party sells $100 million in T-bills for money, that money is risky. It's only insured to $250k in a bank account. The owner will want to move it relatively fast. He might buy longer term bonds, or something else, then the next owner will want to move the new money, at least anything over $250k per bank account, and so on. It seems the velocity of money would increase.

    Do you still believe in MV = PY?

    Oh, and to any asinine attack annomi, this is all just implied questions, why is it wrong, if it is? So if you could please spare me the, you're just an idiot, or ignoramus, or both, and perhaps try to answer the questions.

    ReplyDelete
    Replies
    1. You're just an idiot, or ignoramus, or both.

      Delete
    2. Obviously predictable, as is you're inability to say why it's wrong.

      Delete
    3. If "your" going to snark, you should at least get "you're" grammar right.

      Delete