Saturday Quiz – February 20, 2010

This Post Has 31 Comments

1. Bill,

   A request. Is it possible to have all comments in posts on the same page, instead of having “Newer Comments”? It’s a personal preference, though I guess – others may prefer status quo.

2. Hi Bill,

   Not sure I understand why 2 is False. In “A Modern Monetary Theory Lullaby,” you state,

   “When foreigners demand less $AUD, its value declines. Prices rise to some extent in the domestic economy but our exports become more competitive. This process has historically had limits in which the fluctuations vary. At worst, it will mean small price rises for imported goods.”

   I assume this would be the case in responding to question 2 as well. Please explain.

   Thanks.

3. Dear SCC

   Thanks for your question – upon reflection the wording is a little ambiguous which is why you might be questioning the answer.

   It is false because despite what the popular media will tell us the Chinese government can only convert back into its currency if there is enough Yuan to do so.

   At any point in time someone currently holding the Yuan has to hold the dollars they shed. There has to be a counterparty.

   The private sector in China is the only possibility but to provide that much Yuan liquidity for them to take the USD would be nigh on impossible in any feasible time frame.

   The absolute correct answer would be “highly unlikely to near impossible” so false is closer than true.

   I hope that helps.

   best wishes
   bill

4. Why do you think the CB can control yields at every maturity? Do you mean, the CB can outlaw certain types of transactions, or retire all the paper and forbid new issues? As long as more can be issued, the intervention amounts to windfalls to the issuers, and for liquid securities, you would need to retire all of them to “set the yield”. For illiquid securities, you can again intervene, but only at the expense of supplying windfalls (when attempting to lower yields) or increase costs when attempting to raise yields.

   The term structure of interest rates is an endogenous phenomenon, arising from the return characteristics of overlapping forms of capital in a growing economy. As the returns of each individual technique fall, the cost of capital for that technique rises, so that the economy as a whole has stable returns, but in aggregate you get a yield curve:

   To get the high cost of capital rate, you need to wait for the issuer’s CoC to rise to that rate. But when it rises to the overall equity return, that technique is already saturated and further expansion occurs with a different technique. That means to get a bond with a high yield, you need to buy it early and wait. On the other hand, if you can get the same rate with a long and short term commitment, then this would mean that the cost of capital for that technique is not falling, and is moreover not expected to fall, and this allows business to roll over short term debts for long periods, adding to that variety of capital indefinitely. That isn’t going to happen, except possibly during exceptional boom times. Nor is this something that should be encouraged, for system stability concerns.

   And this is not solely a philosophical question — the Fed just spent a lot of effort trying to drive down agency MBS, purchasing 1/3 of them outright. Pimco was happy to sell them the paper, rotating into treasuries. Next, maybe the CB should try to lower yields on treasuries, and sell MBS back to Pimco, hitting all the high yields. Btw, the best analysis I’ve seen of the CB intervention is that agency yields fell by 25 bp, but even that is debatable, as the risk characteristics were changed (e.g. The Treasury purchased majority holdings in the agencies, there was a rush to safety, unusually depressed NGDP prospects, and Congress passed legislation funding the agencies.)

5. RSJ,

   Bill is talking of a situation in which the preferences of investors changes and they sell some Treasuries and buy corporate bonds. So he is talking of the exogeneity of the government bond yield curve only.

   The government debt yield curve has been “endogenized” by the people in power, everywhere in the world. However, it can be set to whatever value the Treasury and the Fed like by price announcement and the promise by the central bank to buy/sell at that price in unlimted quantities.

6. In other words, the Fed purchases of assets fixed Q and let P float, whereas it can fix P and let Q float.

7. Also, this is not just an academic point. The Treasury and the Fed set the 25 year point at 2.75% annual yield during the second World War.

8. Ramanan,

   I’m not sure what you mean in your first comment — all yields compete.

   In terms of your second comment, Fixing P and Q refers to goods, not asset prices. You cannot “fix” a growth rate or default risk to be something other than what it is. You cannot “fix” a return that is diminishing to be increasing, or vice versa. The returns are what they are, and the prices follow, albeit haphazardly. Attempts to insure that some asset is priced differently than its no arbitrage price results in windfalls accruing to someone, and not to the poor. There is a fundamental difference between between a prediction about the future, and the cost of an apple.

9. During WW2, they ran a command economy, Economic decisions, including decisions about wages, dividends and investment, were set by panels of government officials. Ration coupons, conscriptions, etc. The government can always make things illegal, of course, and there was a lot of solidarity and mobilization during that war. But they tried to keep that rate after WW2, and got 25% inflation, leading to the treasury accord. Of course they could have “fixed” the yield to be a different number, and then not sell bonds, giving huge windfalls to equity holders when that yield was too low, and give massive windfalls to bondholders when that yield was too high. But it is better to not sell bonds at all than to do so in a way that deviates from the market price. At least this way your own hands are clean that you are not purposefully benefiting either the long or short.

10. RSJ,

    You are talking about Q3, right ?

    I am not sure how you interpreted the question. I am reasonabaly sure that he didn’t mean buying corporate bonds by the central bank.

    Plus in Q3, he is talking of securities purchases only, not labour or production, I think.

    A corporate cannot fix the yield of the long term bond – at best it can do some optimization with maturities or do a crazy swap with Goldman Sachs! A government can! That is what I am trying to say. In QE, they removed some supply out of the market and hence brought the yield curve down. However, they could have announced an explicit peg – hence fixing Q vs fixing P.

11. I did not say that corporates set the yield. The market as a whole sets the yield. And the yield is the same for each maturity, so there is no difference between trying to fix the yield of public, corporate, or financial debt. All these compete.

    The yield of the long term bond is a discount to the equity yield. Very long term bonds are the same as the equity yield, which is the growth rate of the overall economy. Historically long terms bonds and equity have both yielded this rate, with a mean error in the handful of basis points over long time periods. Of course there are 20 year bull and bear markets, typically inverse in equity and long term corporate debt.

    Shorter term yields are discounts to these yields. If you imagine a certain variety of capital– this is an idealization — You start out in the early stages when there is high risk and very rapid growth. Then it starts to mature and growth begins to slow. Eventually it becomes a cash-flow pig, paying the market yield in dividends and holding no debt. And note this is not a corporation, as a corporation holds a portfolio of these, often capitalizing it’s new projects with the cash-flows of the saturated forms of capital. For example photoshop (now made in India) is a cash-flow pig for Adobe, but they are investing in flash, etc.

    During these stages, each form of capital is funded differently, because to do otherwise would allow arbitrage. It makes no sense to sell debt when returns are increasing, you want to capitalize with equity. It makes no sense to sell debt when returns have hit the market rate, as you do not want to increase the size of the capital stock. You sell debt from the start of diminishing returns up until the returns hit the equity yield. And the maturity of the debt is the current time to the equity saturation point. So to be repaid 10 years from now, when your project is mature and yielding 6%, you are willing to pay out a higher yield. For debt that matures when your project is yielding 3%, you will not be able to sell 6% debt — even if you wanted to. You would end up liquidating your project. So the yield that you are willing to pay (or that the market is willing to buy) is such that if you didn’t sell the debt but capitalized with equity, you would be no better or worse off. This is the no arbitrage condition, and fixes the yield, based on (often flawed) expectations. Such a model will give you a yield curve for each flavor of capital, and with a constant stream of such varieties, you get a constant yield curve. Note that there is no room for any P and Q nonsense. We are talking about the profitability of certain enterprises, and asset prices are bets on this.

    Even things like real estate and government debt are set by these expectations, because you always have the choice of investing in the form of debt that delivers a real return, based on the characteristics of the economy. So the yield curve coalesces into an overall view of the economic performance over different time horizons.

    Notice that there are no free variables here — everything is decided by the arc of that variety of capital — how quickly it can capitalize itself, what its ultimate saturation level is, and what returns it will offer as it trends to this level. Of course markets are often wrong, but that just means we can’t predict the future. It doesn’t mean that the government would do a better job trying to predict it for us. Let the government focus on influencing the real economy: jobs, income distribution, education, pensions, etc. If the car is broke, fix the engine, don’t try fiddle with the speedometer as a way of getting yourself moving. At worst, such an approach will break the speedometer, and at best, it will achieve nothing, but the only reliable outcome from this is big bank bonuses and/or dysfunctional funding markets.

12. Dear RSJ

    Thanks again for your input.

    You might like to read the speech that Ben Bernanke made to the National Economists Club in Washington (November 21, 2002) called – Deflation: Making Sure “It” Doesn’t Happen Here.

    Among other things, he outlined a series of measures by which the central bank can influence capital markets – all of which we have now seen in action in the current crisis and which have been described as non-traditional monetary policy. These measures included large-scale purchases of financial assets (bonds) to lower “rates further out along the Treasury term structure” and “announcing explicit ceilings for yields on longer-maturity Treasury debt” and enforcing those “interest-rate ceilings by committing to make unlimited purchases of securities up to two years from maturity at prices consistent with the targeted yields”.

    If you read the BIS paper – Unwinding public interventions after the crisis – that I referred to in Friday’s blog – The vandals are gathering – you will see they discuss the same topic which they call “central banks’ unconventional balance sheet policies”.

    In other words, the paper is addressing the aspects of monetary policy that go beyond setting the short-term interest rate which have emerged in the currenc crisis.

    The BIS said:

    However, the key issue here relates to the potential role of central banks in directly influencing long-term bond yields and credit spreads: market participants should be under no illusion that we are entering into a new permanent accommodative monetary policy regime in which central banks would be able and willing to control the entire length of yield curves as well as credit spreads and mortgage rates. The unconventional measures should not be seen as an additional set of tools that central banks would use in their normal day-to-day conduct of policy.

    I made the point that all the angst about bond markets are going to blow up because traders and ratings agencies will punish government’s with deficits they judge to be “unsustainable” is unnecessary. If any government is paying yields on their debt higher than they think is desirable then they can simply instruct the central bank (or change legislation to allow them to do so) to manage the yield curve.

    The fact that central banks do not do this reflects the ideological position that monetary policy should be targeted at fighting inflation (with no particular regard for the real consequences – given they believe in the NAIRU myth) and that fiscal policy should be a passive partner in this unemployment-creating folly.

    If governments started to use the capacity of their central bank operations to control longer-term yields then the bond markets, the rating agencies and all the rest of these self-serving forces – become totally irrelevant.

    That was the point of the question. For readers to understand that it is a policy choice which makes it a political choice for governments to allow bond traders to hold them to ransom. The government has the capacity to eliminate that coercive power. Most people fuelled by the media do not understand that. They think in some way it is inevitable that the bond markets will drive up yields on public debt and beyond government control so the only thing the government can do is obey. The coercion then manifests in the form of damaging (and unnecessary) fiscal austerity.

    best wishes
    bill

13. RSJ,

    Many things. Analogies with corporates do not apply to the government sector. Even Ben Bernanke knows that. Check this speech Deflation: Making Sure “It” Doesn’t Happen Here where he talks of explicit ceilings on yields and the Fed purchasing unlimited quantities of Treasuries.

    There are economists who do not even believe that the Fed can target the short term rate. In principle, it can of course, set it to arbitrary accuracy. Some central banks other than the Fed are close to doing that and the Fed too has over time been able to have the Fed Funds trade close to the target.

    The government sector is in a different position from the rest of the economy. For a country which doesn’t export in huge quantities, the government is in a position of negative (technical) net worth as far as financial assets are concerned. So a rule of thumb is that governments run deficits forever. It is impossible for other sectors of an economy to be in a deficit forever.

    I remember having a similar debate – I don’t know what is preventing you from at least appreciating the idea, that the Treasury and the Fed can set the yield curve. Bill frequently talks about the tap system which used to exist in Australia. See this: Why history matters

    Central banks follow a Taylor rule policy out of choice, nobody forces them to do so.

14. I understand you, Ramanan. You say that the government can set 5 year bonds to yield 100%, 50% or 0%, because, well they are *government* bonds, and then magic pixie dust comes in to say that people will pay more or less for *government* bonds than for other bonds, on a risk adjusted basis, because the dollars are tastier, and well, it is the “government” and there are Ps and Qs involved. Like Walmart employees, we shout this every morning.

    5 year bonds are all the same, once adjusted for risk. That is why the Fed failed to lower mortgage yields. That is why, for all the talk of controlling rates, the CB is very careful to keep rates low enough to ensure a few hundred billion go to banker’s bonuses each year, but not too low so as to allow others to enjoy below market funding. In this way, those who rely on short term funds have an advantage. This is a delicate tightrope, and the end-game was Bear Stearns financing itself with 75 billion of *overnight* loans that they continually rolled over to purchase 10 YR duration MBS, and then the bondholders were bailed out by the CB. All in the name of controlling rates, which have not changed for people who cannot have the government on the other side of their trade.

    Now you can make any system you want, in which windfalls are available to longs or to shorts, or to your favorite powerful group, but you cannot control rates. All you can do is sell something below cost or buy something above cost. Under the TAPS system when yields were too low, bonds didn’t sell and when they were too high the private sector got a windfall. Our CB was worse prior to ’51, because it bought from the open market as well, so there were huge windfalls, primarily for pension funds, when the fixed rate was both too high and too low. At least, with TAPS, there is only corruption in one case, and the government abstains in the other. But then why not abstain all the time?

    But the ability to withdraw from the market and refuse to sell something does not mean that you can set the rate. The government did not control the rate under TAPS, or under our 1946-1951 experiment — what it controlled was who was subsidized and under what circumstances. That power it keeps, and that is what happens when you sell something for less than the going rate, or buy something for more than the going rate.

15. Hmm, I apologize for the excess snark, Ramanan. The point is, anyone can sell something below the market price, and buy something above that price, and refuse to transact in other cases. The government is not special in that regard. Attempts to sell something at a fixed price only will result in someone buying it for less than they are willing to pay, or selling it for more than they are willing to accept. That may be OK with where the advantages of a uniform price outweigh the disadvantages of variations in preferences, but it is not OK in markets where professionals can earn huge profits from small arbitrage spreads, due to the ability to incur enormous leverage. In that situation, anytime you transact outside of a competitive auction setting, you are giving large amounts of money away to financial professionals.

    As you say, that is fine because the government is not revenue constrained, but why is the government using its massive powers to sell or buy securities at non-market rates? Someone is benefitting from this arbitrage opportunity, and yet there are many unmet fiscal needs that need to be addressed. The U.S. has issued a GDP worth of commitments to back securities, all in some effort to boost asset prices, and yet not a single commitment to employment, distributional issues, or any of the root causes of this 30 year sickness. Absolutely zero financial market intervention vis-a-vis prices is necessary to achieve any public purpose goals.Yet we focus on the asset prices, precisely during the period of time that finance has taken over the economy. The two are not unrelated. We have neither a shortage of money, nor an excess of interest rates; our problems are in the areas of distribution and production, in the real economy. Tax policy, regulation, and public spending can solve these issues independent of any market valuation outcome.

16. Hi Bill,

    I’ve been reading your blog for a few weeks now, and have enjoyed the MMT perspective.

    However, I don’t understand why Question 5 is false – surely a larger deficit is always more expansionary in nominal terms than a smaller?

    Thanks

17. Dear LAJ

    Thanks for your E-mail.

    Q5 is:

    A budget deficit that is equivalent to 5 per cent of GDP is always more expansionary in terms of nominal GDP growth than a budget deficit that is equivalent to 3 per cent of GDP but the actual number should not become the policy focus.

    Note the wording said “more expansionary in terms of nominal GDP growth” – so a deficit equal to 5 per cent of GDP in the face of a very large private spending gap may not be associated with as fast a GDP growth rate than a 3 per cent deficit to GDP ratio associated with a much smaller private spending gap. Yes, the impact of the higher deficit per se will be larger in nominal terms but the expansion of overall nominal GDP growth will not be under all circumstances.

    It was a trick question which might not have been worded very well.

    best wishes
    bill

18. RSJ,

    No need for an apology 🙂 I remember my days in academia, where I used to get friendly with people debating strongly on something or the other.

    The government purcahasing its own security through the central bank is a different thing than the government purchasing some real goods. Plus it is not necessary for the government to start dictating to control the yields. It doesn’t need to declare that yields higher than 5% is a criminal offence.

    An active coordination is needed for the government and the central bank to carry out such an operation, though it is hardly a Herculean task. The government spends by crediting bank accounts. If the government announces that the price of a CUSIP is $x, why would anyone sell it to others for less than $x ? If I am ready to sell my securities to the government at $x, the government simply has to instruct the Treasury to credit my bank account by $x and this can be automated by a software.

    In the present scenario, the government simply lets the market decide the price of the Treasury security through an auction. The government is in no shortage of funds to be forced to do that. In some sense, what the yields should be for various maturities is a political choice.

    The point Bill is trying to make is that the Treasury does not have to be a slave to the market, if the markets decide to bid up the yields on government securities. Imagine – the US Treasury market collapses for some reason or the other. Not likely but one in a million chance because there are enough fear mongerers out there who have been trying to play all sorts of games. A corporate under such circumstances is in a terrible situation. The government, on the other hand, can simply take control of the situation.

    I think in the US, the Treasury has been building a buffer. It’s account at the Fed is around $200b or something and the Treasury also holds accounts at the commercial banks and that amount has increased further by around $150b-$200b (Its usually $1-10b). The Treasury’s account at the Fed cannot be in an overdraft position, by law. The Fed is not allowed by law to buy securities directly from the Treasury. There is a catch, however. When the Treasury securities held by the Fed mature, they are exchanged for new securities. It is allowed by law. So the Fed can buy securities from the market, just when they are about to mature. So even without a change in law, the government sector has a control over the markets. There is no chance that the market loses confidence and starts selling the Treasuries. Even if it does, the Fed can buy them at a lower price!

    The point Bill is trying to make is that if governments understand such things well, they do not have to worry about the government bond market – it doesn’t stop the government from spending. There are simpler ways to tackle the situation than what is happening right now. People working for the Treasury can sleep well!

    Ben Bernanke knows this I guess – but he seems to believe in the “inflation expectation theory” and that is his argument for letting the price of Treasuries float.

19. Ramanan,

    The government spends by crediting bank accounts. If the government announces that the price of a CUSIP is $x, why would anyone sell it to others for less than $x ?

    OK, corporate bond A is a 5 yr bond with a (risk-adjusted) yield of 5%, the same yield as the 5 year government bond. The government decides to sell a 5yr bond that yields 4%. Who will be the buyer? Only the CB. It will retire all the instruments. The net effect of this is that the auction fails, or the government sells something to itself, which, from a consolidated accounting view, means that it sells nothing.

    Suppose that there is a large stock of existing government bonds currently yielding 5%. In this case, who will be the seller? Everyone except for the government. They will sell the government bond and use the proceeds to buy A, earning a risk-free profit. In the process of doing this the yield of A does *not* increase, because the financial markets are, to a good approximation, infinitely elastic. As the yield of bond A starts to fall, company A will issue more bonds, shifting its capital structure from equity to debt, as debt is being subsidized. As the ratio of equity to debt decreases, the risk profile of A’s bonds begins to worsen, until the risk-adjusted yield falls to 4%. At that point, the arbitrage opportunity ends.

    So as a result of trying to fix the price of the government cusip, what happens is that wealth is transferred from the holders of bond A, whose risk adjusted return has now fallen from 5% to 4%, to the holders of the government bond, who were able to sell their 5% bond to the government at a profit. But the term structure of interest rates has not changed. An index tracking certain issues may move, but only because the risk characteristics of those issues were changed. In reality, the Fed retired 1/3 of MBS and failed to move the rates in any meaningful way. If it continued and retired the remaining 2/3, then rates would still not move; but removing an instrument from the market place does not mean that you have power over returns.

    The point Bill is trying to make is that the Treasury does not have to be a slave to the market,

    A sovereign government is not influenced by these markets one way or another. It is not funded by financial markets, and therefore does not “lose” money when yields rise, neither does it “save” money when yields fall. Moreover, OMO cannot alter the underlying time-path of returns, or of the size of returns. The government only has *spending* power, it does not have the power to change the time-path of diminishing returns, and therefore it cannot change the rate at which yields rise as a function of time commitment.

    And this is what sets the term structure of interest rates in a no-arbitrage context. The only possibility of government intervention is to to try to break that no-arbitrage condition, and therefore provide arbitrage profits to those who will respond and restore the yields to their competitive level. Now the time required to restore yields to the no arbitrage levels will depend on the type of intervention and the elasticity of the market at that duration, but over the long term, the government will not be able to force an arbitrage opportunity to exist permanently. Something will happen — e.g. the company A will issue so much debt that it’s return characteristic falls, or banks will swell up to consist of 1/3 of the economy, or they will shrink to zero, etc., but eventually the difference between the actual price as set by the rate of diminishing returns and the price the government is willing to pay will go back to zero.

20. RSJ,

    The 30y point on the US yield curve is where it is because the market expects the Fed Funds to average that value over the next 30 years. Another way I look at it is that the yield curve is determined by liquidity preferences of investors.

    The Federal Reserve follows an “inflation-targeting” policy. It is not binding on the Fed to do that. It can just say we are setting the rates to be 100bps forever. The “public sector borrowing requirement” for this fiscal in the US is going to be $1.3/1.7T (?). The Treasury can just say we are issuing 10y Treasuries at 2.5% yield – take it or leave it. If there are no takers, the excess reserves out of government spending just lies with the banks – the Fed “monetizes”.

    I do not know what the arbitrage is for the case we are talking about. Steepeners and flatteners are risky trades, not riskless. For an arbitrage to happen, the forward rates have to be different than the “implied” forward rates. The forward market also needs to be liquid for it to happen.

    Also read Bill’s comments above. I think you are making an assumption about the existence of “free markets”. That may get us into another discussion, but the point is that even the Fed recognizes that fact that the yields on government securities can be fixed easily. It is not easy to fix the price of bread, of course and I am not at all claiming that it can be fixed.

21. Dear Bill,

    Thanks for the explanation to my question; however, it has generated more questions for me. For example, how does one determine there would not be enough Yuan liquidity to support the Chinese governments conversion of its dollars back into Yuan? What figures would prove this?

    Coincidentally, M. Pettis discusses China’s dollar reserves today here:

    http://mpettis.com/2010/02/what-the-pboc-cannot-do-with-its-reserves/

    He states:

    “Of course the PBoC must fund the purchase of these dollars. It does so primarily by borrowing in the domestic money markets, selling PBoC bills or entering into short term repos (although it also issues some longer-term bonds), or by “creating” money by crediting the accounts of the commercial banks who sell it the dollars.

    This means, to simplify, that the PBoC has a balance sheet consisting on one side of dollar assets (and here “dollar” is short-hand for all foreign assets). Against this and on the other side it has a roughly equivalent amount of RMB liabilities (I say “roughly” because when you run a mismatched balance sheet, changes in the relative value of assets and liabilities will create losses or profits).

    Here is where things get interesting. China’s reserves are often thought of as if they were a treasure trove available for spending. They are not. They are simply the asset side of the mismatched balance sheet. If the PBoC wanted to “spend” $100, say for example to recapitalize a bank, it could do so, but this would automatically create a $100 dollar hole in its balance sheet. – it would still owe the RMB that it borrowed originally to purchase the $100. To put it another way, the reserves are not a savings account, free for the PBoC to spend as it likes. Reserves are effectively borrowed money.”

    Does this reasoning play into your earlier explanation? Again, thank you for any clarification.

    Best regards,
    SCC

22. That example by Pettis above seems bizarre and awkward. Since when is a central bank’s recapitalization of a commercial bank an example of “spending”, and why frame the spending question in the context of the activity of a bank? I would frame the actual example he used in the context of central bank portfolio management, not “spending”.

    I think Bill’s point is that the only dollar/Yuan exchange market of any size at all is the domestic Chinese market. And when PBOC sells dollars for Yuan, it will extinguish an equivalent value of PBOC Yuan liabilities. Nobody will be accumulating Yuan assets in that scenario. The Chinese private sector will be exchanging its Yuan assets for dollar assets. China’s net foreign asset position in dollars will be unchanged as a result.

    If Pettis’ example is intended to portray the central bank’s capitalization of a commercial bank with Yuan, then the central bank might exchange dollars for Yuan as per the above. But somebody in China would still end up with dollar assets. However, that doesn’t appear to be the nature of his example. It looks like he’s illustrating capitalization in dollars, in which case the central bank’s dollar/Yuan balance sheet mismatch remains. So I’m not sure why he used this particular example to illustrate anything.

    I find the first paragraph quoted from Pettis above quite interesting. It seems to indicate, predictably perhaps, that he does not fully relate to the MMT perspective on banking as it applies to the case of China’s central bank operating in its own currency. PBOC creates Yuan reserve liabilities when it buys dollars from the exporters. It then converts those reserve liabilities to other instruments as desired, with such operations as issuing central bank treasury bills. Pettis talks about PBOC “funding” dollar purchases with Yuan, but that is no more the case than it is when the Fed creates excess dollar reserves by purchasing private sector dollar assets.

23. Ramanan,

    Long post warning!

    The “no arbitrage” condition is such that after selling debt:

    (*) new equity value – old equity value = market value of debt

    If (*) is violated, then equity holders can obtain free money by selling debt, or bondholders can obtain free money by selling bonds. The no arbitrage price is the price at which neither happens.

    And this drives the interest rates for each maturity, because the left hand side is completely determined by the (expectations) of the diminishing returns of the variety of capital employed and the (expected) long term growth rate of the economy, whereas the RHS is determined by the terms of the bond. So the no arbitrage price is endogenous combination of expectations of the performance characteristics of different types of capital over various time horizons.

    FF has *nothing* to do with the no-arbitrage price. FF is only relevant as a hidden variable of the banking system that determines the overall competitive advantage that banks have versus non-banks. If they are at an advantage, then the economy is financialized. If they are at a disadvantaged, then the economy is de-financialized. It is a knife edge, in which you need to offset the (artificial) subsidies you are providing to banks in the form of government guarantees for depositors and unlimited reserves with the (artificial) fees you are imposing on banks in the form of FDIC fees and FF. With the left hand and the discount window you give, and with the right hand via OMO you take away. Neither operation is “natural”, and the only “natural” rate is one in which no arbitrage is possible and returns are determined by expectations of real returns.

    Here is an example. Suppose a variety of capital has an expected return characteristic:

    R(C) = 1/(2C) C>= 1/2

    –i.e. The returns as a function of the size of the capital stock are 1/C$ per year, for ever $1 of capital employed.

    Suppose the equity market yield = m = .05%

    C goes from 1 to 10. At C = 10, R(C) = market return, and additional investment in C will not occur.

    What is the time path, assuming no debt?

    dC/dt = R(C)*C = t (all returns will be re-invested as they are worth more by being invested internally).

    Then,

    C = t/2 for 1 <= t < 20
    C = 20 for 20 t >= 1)
    E(t) = 10 t > 20

    Whether you, as a *market investor* buy the capital at t = 1 or t = 10 or t = 50, you still only get the market return: .05%. In the latter years you will get it as a cash, and in the earlier years you get it as a capital gain.

    Note that E(1)/C(1) = 7.9 and E(10)/C(10) = 1. Over time, the “future” value of the capital is converted into tangible value, by means of re-investment or debt sales. And in fact this “future value” — E-C, is the source of all expansion funding for the business. Capital is *not* funded by financial market participants buying and selling claims — not even new issues. Capital is funded by expectations of future return. Financial market investors determine what those future returns are, and when the value of future returns exceeds the current value of deployed capital, the business is able to convert the difference into expansion funds. When it falls below that value, the capital is liquidated.

    All of this was assuming equity financing, and no debt. Now suppose, at t = 1, the business decides to speed up the time path by raising $1 in bond sales, using the proceeds to increase C by 1. On the one hand, the market consols will be delivered 2 periods sooner, which means that their present value increases. On the other hand, the business has a new liability. The no arbitrage price is such that these two off-set exactly:

    10*(1+m)^(t-18) – 10*(1+m)^(t-20) = Bond liability(t)

    If the (zero coupon) bond has a rate r and matures when t = s, then we have:

    10*(1+m)^(s-18) – 10*(1+m)^(s-20) = (1+r)^(s-1) for s > 1

    We can solve for the arbitrage-free term structure of interest rates for this particular form of capital as a function of the maturity: (= s-1).

    maturity/rate for borrowing $1 of capital at time t = 1.

    1 1.2%
    5 2.3%
    10 2.58%
    18 2.7%

    Note that lim as repayment time –> inf is the market yield, 5%, but you need to be careful to switch to the constant equity formulas when you breach saturation point (if you do not asymptotically approach them). In fact, this is why the long term bond rate becomes the market rate.

    In terms of liquidity preference and risk aversion, these also play a role, of course, but you don’t need these to see a yield curve. It is enough to view the economy as a series of overlapping forms of capital, each with its own term structure, and even though you cannot aggregate the capital, you can aggregate the term structure. You cannot aggregate C, but you *can* aggregate E(C). the emergent effect is a yield curve that does aggregate expectations of return over different time horizons for all forms of heterogeneous capital.

    There is a free parameter here that influences rates: the relationship between the amount borrowed and repayment time. Here Liquidity preference/rollover concerns enters into the picture:

    The arbitrage free price depends on the value obtained from the loan. The value is expressed as decrease in the time to obtaining the market consol: I.e. to borrow enough to save “a” units of time means the principle amount = C(t+a) – C(t), to be repaid at t+s. *If* s < a, then in effect you are repaying by selling the capital stock you just purchased. In that case, the intrinsic value to you is zero — i.e. You borrowed $1, bought something, sold it back for $1 and repay the dollar in zero time. In this case, *apart* from financial arbitrage, it is not possible for a firm to sell productive capital it has deployed without incurring a loss. So a "no-ponzi" condition is needed, in which the repayment time exceeds the time saved. In this case, we get a non-zero 0 day rate, proportional to the slope of the rate of increase in equity value at that time, divided by the slope of the change in capital stock. This corresponds well with the observed behavior of yield curves. We can then talk about the ratio (a/s) as a measure of liquidity preference, and observe how this affects the yield curve.

    There are many other interesting things here (to me), but the point is that what *should* drive rates is a no arbitrage condition combined with the real return characteristics of the economy and expectations of return. Anytime government uses its power to allow financial arbitrage to occur, it is not serving a public purpose, and is promoting income concentration and financialization.

    In terms of the effectiveness of government manipulation of rates, in those markets that are very elastic, intervention will fail to move rates, but will succeed in shifting capital structures. In those areas where the market is less elastic, intervention will succeed, in the short term, but will eventually cause capital structures to shift as well (towards more leverage, or less leverage, depending on whether rates are lower or higher than the no-arbitrage level). In both cases, FF is as relevant to interest rates as the color of Jamie Dimon's Bentley. There is a connection in the very short term markets, but you cannot both have a system in which FF moves rates and one in which banks serve a public purpose. By definition a public purpose banking system is one in which no one is getting free money in the financial markets. This is the real rentier profits that you should be striving to eliminate, instead of arguing that they should be increased.

24. RSJ,

    While I will come back to you on the main points of your comment, I don’t see the relevance directly here. That is ok – I search out your and other commenters’ thoughtful notes sometimes from old posts. Plus it will require repeated reading (-:

    Are you assuming that the US Treasury is an independent institution from the Federal Reserve ? I think this is how economists and bankers in the US think, but they are just two different balance sheets of one government – it cannot be otherwise! Its not the way it is seen in my country(independance and all that), except by economists who have studied in the United States.

    The balance sheet of the government sector (the Treasury and the Fed) should not be looked at like a corporate. Similarly, the investment analysis on securities issued by them, which includes the reserves as well.

    Further, there are two things you have pointed out here and comments on other posts. The effect of FF on the economy and the effect on other rates. While discussing this, you will agree that there are a lot of points on Political Economy here and you have pointed this out yourself several times. While we can discuss the political implication of controlling the interest rates, the more important point is that the yields on government securities can be made exogenous, but the policy makers have chosen to have it become endogenous. We are talking of the government’s control of its the government yield curve, not the corporate bond market. The discussion of the effect of FF on the economy and loan rates is a slightly different topic.

    Because, the politicians and central bankers have made this choice, there is a consequence called fiscal austerity, which leads to damaging effects on the citizens of the economy. (Yes, I am entering into Political Economy now). Consider this – when interest payments on the government bonds go up, thee is a pressure from various sources on the government to cut down its pure expenditures. Even if the politicians do not face pressure, they behave as if they ought to reduce expenditures and do it as well. When deficits are “high”, rating agencies start making threats, causing yields to move and become volatile – which leads to typical scenarios where the FM gives an interview and promises various things.

    Should I take your stand on this as something like the following: The government can control the yield curve, but it will favor the markets/bankers more than it favors others ?

    For the effect of FF on loan rates, will come to you on this. The effect of changing FDTR on the economy is also another discussion to look forward to – maybe we can discuss both here itself.

25. Hi R,

    Many issues here, so I’ll try to address some and summarize. Hopefully the pixels are not being burned in vain.

    Let’s forget the political economy for a moment and talk about fixing prices.

    First, financial markets are not consumption markets. To control the price, you need to be both the marginal buyer as well as the marginal seller. That is why nations that peg currencies accumulate large foreign reserves. It is not enough for them to “sell freely” their currency at a fixed rate. They also need to “buy freely” if they want to put a floor under the exchange rate. So the intuition of a class of producers selling goods to a class of consumers does not hold in the financial markets.

    Also, even if you could be both the marginal seller and buyer, still you could only fix one price within each market. You can’t peg against both the yen and the dollar. In the same way, you can’t arbitrarily control the price of the 7 year bond and the 30 year bond, as setting the price of one will determine the price of the other. In fact, it will determine the price of all other instruments in that market. A “market” is defined as those assets that are substitutable for each other. This is a question of degrees and time-frames, but in the financial markets assets are highly substitutable, as all you care about is return.

    Assets of the same maturity are instantaneously substitutable. So there is no “corporate” yield curve that is different from the “government” yield curve. There is just one yield curve.

    And I would go further and say that when looking at business cycles it’s reasonable to to think of three markets: the capital market (bonds and equity), the money market (repos, bills, commercial paper), and the interbank market (reserves). These markets can be viewed as distinct, as the money market is a rigid market for short term cash-flow management that cares less about return than about safety. It is like your savings account. Whereas the capital market is for longer term investing, and the interbank market is a closed market.

    Only in the interbank market is the government both the marginal seller and buyer of reserves. And banks cannot substitute some other instrument for their cash needs in this market. So the government has the power to control FF.

    In both the capital markets and the money markets, the government is neither the marginal buyer nor the marginal seller, as anyone can sell substitutable assets and anyone can buy them. There is no possibility for the government to fix prices in these markets by buying or selling securities.

    So the only way that the government can control returns is via FF — indirect control via spillover effects.

    I think because of this, the common paradigm for both PK/Neo-classicals is a “monetarist” channel (no offense intended)

    interbank market –> money market –> capital market –> economy

    In this paradigm “The rate of interest” is FF, since that is the knob they turn, and by assumption the other rates move in response to it.

    In my post, I was advocating for a different channel, call it the “return channel”

    economy –> capital markets –> money markets

    In this channel “The rate of interest” is the long term equity yield, which I argue is the GDP growth rate. The “knob” to turn is fiscal policy that affects the economy, which in term determines the returns in the capital markets and the term structure of rates — at least capital market rates. What happens to the money markets is irrelevant in my channel — so I assume that there are no liquidity crises.

    I naturally do not think of supply and demand as setting prices, because capital markets are highly elastic. The arrows in my channel are expectations of return for different forms of capital over different time horizons, combined with no-arbitrage requirements. In this way, longer term rates determine shorter term rates + economic return characteristics determine shorter term rates. I view FF as a synthetic fee that, to the degree that it gives different results in the capital markets than that predicted by the return model, allows for arbitrage.

    The monetarist channel starts with FF. To the monetarists this is “the rate of interest” from which all else flows. Their first arrow hits the money market which is highly inelastic, and therefore they view yields as set by the quantity of bonds/reserves supplied. To them, the role of banks is to be the government’s chosen “instrument” to arbitrage the yield curve. The arrows in the monetarist channel are therefore changes in the supply of bonds or reserves affecting short term yields, which via banks rolling over debt determine long term yields, which end up affecting the economy.

    Who is right? Well, it depends on how elastic you think the capital markets are. I don’t believe the data supports the notion of FF controlling capital market yields, or that long term rates can be calculated from averaging expected short term rates. But I was arguing that you should not use FF type manipulations, because to the degree that they succeed in affecting capital market yields, they will violate some no-arbitrage condition, which is to be avoided.

26. RSJ,

    Very nice analysis. Good that you made the distinction and divided the markets into 3 segments. I was also going to do something like that.

    I don’t think PK looks the way you summarized it as

    interbank market -> money market -> capital market -> economy

    It is exactly the opposite! Money is everywhere and anywhere an endogenous phenomenon. The extreme form is called Horizontalism and I do not find that extreme at all. Economic activity is demand-led. Even in the gold-standard era, it was demand-led and money was endogenous. Except setting the overnight rates, the central bank cannot do much. It is not to say that monetarism experiments are not possible but it can be damaging. Demand for credit is not very interest sensitive. Of course mortgages are. Fiscal channel is highly effective in determining demand but the public debt is endogenous and out of the control of the government. Again, the choice of fiscal austerity is consistent with my previous statement. So I am saying – monetarist experiments and fiscal austerity will not make money less endogenous than what it is.

    I agree with you- the effect of the “central bank reaction function” is indeterminable and weak. The fiscal channel is much much cleaner.

    I generally agree with you about money market rates other than the overnight rates. In PK literature, they are endogenous markups over the overnight rates. The T-bill rates are “administered”, bank lending rates are “set” and the commercial paper market I guess is closely related to banks’ lending. Banks by themselves set the price on loans based on how capitalized they are and how much dividends they want to hand out etc., but there is some dependence on the interaction between the non-free market (my terminology) and the free markets. So there is a mix of exogeneity and endogeneity, there. Plus financial firms have their own borrowing route through commercial paper and it is huge in the US. Of course, the non-financial firms have total internal funds, which you reminded me in an old discussion.

    The capital markets on the other hand, is a complicated thing. The equities market may be described in the long run, the way you have described, but the previous decade is a proof that its a crazy market! The yields on the corporate debt maybe described the way you have – I don’t have any idea on this one.

    But, I generally agree with you – Income determines everything, not FF – if that is what you are saying.

    Now I come to our main point of debate – the Treasury yield curve. You compared it with the central bank intervention in the FX market, but I would argue that it is not a good comparison. The Fed/Treasury cannot run out of dollars in defending the price of all securities till the last decimal place. It just needs an announcement everyday in the morning. The bond market will automatically move to that price. Why would anyone sell/buy at a different rate than what is announced ? A 7y price/yield will not determine the 80y price/yield because there is no formula for this. It depends on what the Fed/Treasury will do between the next 7 years and the next 30 years. Nobody can predict this and neither will the Fed/Treasury have an idea. The Fed/Treasury can tell the market that they will move these rates randomly according to some Libor Market Model or a CIR model. In fact, a simpler way to look at this exogeneity is to just consider a perpetuity which pays $1 once per time period.

27. It is exactly the opposite!

    Every time someone says that “higher rates reflect expectations of future short term rates”, which both you and Scott F. have argued, then you are arguing for the monetarist channel. That is the only channel by which FF can affect rates.

    The T-bill rates are “administered”[…] A 7y price/yield will not determine the 80y price/yield because there is no formula for this.

    I provided a formula for the term structure in my first reply, applicable to a single form of capital at any point in it’s life-cycle. Average that by equity value for all forms of capital to get the term structure.

    In PK literature, they are endogenous markups over the overnight rates. The T-bill rates are “administered”

    I know, but the PK literature is wrong in this regard

    1. They didn’t realize that the real economy requires a certain yield curve because returns diminish with time — they just didn’t make the connection, but anyone who has developed capital understands intuitively that yields paid increase with the duration of the capital commitment, because the relative trade-off of paying out a coupon or re-investing changes over time, due to the diminishing returns of your investment. As you can always shift your capital structure from equity to debt, this forces a different interest rate based on the maturity of the bond.

    2. They also didn’t realize that the equity yields need to be the long term growth rate of the economy — something that any equity investor could tell you.

    3. Then, they didn’t realize that the financial markets need to hold to a replicating returns principle, in which the price of one asset is equal to the cost of replicating those returns in the universe of investments not containing the asset — something than any market participant could tell you.

    As a result, the price of government bonds is completely determined by the price of non-government bonds of the same maturity, and the price bonds are completely determined by equity returns over that maturity, as each business has the option of capitalizing itself with equity or with debt. And the price of equity is determined by the overall growth rate of the economy.

    If you combine the above you get the result that all rates are determined endogenously, at least to the degree that the no-arbitrage principle holds. And one consequence of this is that, in an economy with a positive long term growth rate, the risk free return is that growth rate. Only if the economy has zero long term growth does the risk free return equal zero.

    But the PK’ers were ignorant of 1-3, and so decided that because they couldn’t explain how these rates came about, that the rates must therefore be “exogenous” and could be “administered” based on their own particular moral principles.

    Then, the PKers fell victim to some simple aggregation fallacies and decided that low yields correspond to higher wage shares — and so decided that government should attempt to manipulate yields in order to keep profit rates low.

    But that is the exact opposite of how things work. When the cost of capital — the required return — is below the actual return on capital, the surplus profits do not fall to labor, but to capital.

    In the same way, when the cost of capital exceeds the return on capital, then the owners of capital take the biggest hit. Their returns fall to zero, and often the capital is liquidated (e.g. unemployment). But wage shares do not fall when there is high unemployment, they increase. Profit rates fall during downturns, and rise during booms, and wage rates move in the opposite direction.

    And if you were to succeed in consistently suppressing the cost of capital below the returns on capital (which is what our recent macro policy has been about), then you would only succeed in reducing the wage rates, not increasing them, and this would not get you out of your demand shortfall.

    Moreover, because the PK’ers weren’t aware of the replicating returns principle, they didn’t realize that returns arise not from bearing risk of default, but from opportunity cost of investing in something else. Therefore your returns are based on what else you could have done — and that means, in a growing economy, you will always get positive risk-free returns if you make a positive time commitment.

    And the coup-de-grace is that because the PKers confused the interbank market with the capital market, they decided that FF corresponds to a risk-free return that an investor receives for a zero day commitment — which is not the case — and so combining this confusion with the “moral principle” approach to yields, they decided the best thing to do was a zero FF.

    And in this, they agree with the monetarists who argue that rates should always be cut in downturns. But short term yields already fall during downturns, as short term return prospects are falling. So what they are really doing by cutting the short term rates further is excessively steepening the yield curve during busts, and then excessively shortening the curve during booms. This exacerbates the existing credit-cycle, leading not to a “great moderation” but to a “great financialization”.

    The capital markets on the other hand, is a complicated thing.

    Yes, it is complicated and messy, because you are asking someone to incur a long term liability based on expectations of future return. Borrowers in the capital markets are “taking the leap” into the unknown. The role of government here is to regulate the business cycle in such a way as to deficit spend (which increases profit rates) when the return on capital is below the cost of capital, and to tax away profit shares when the return on capital exceeds the cost of capital. In this way, total deficit spending will remain a relatively constant share of GDP, demand “leakages” due to wealth accumulation will be plugged, and the underlying economy will face stable returns. This is much better than attempting to intervene by trying to manipulate the capital markets, which are merely indicators of sentiment. You intervene by manipulating the economy.

    Now I come to our main point of debate – the Treasury yield curve. […] Why would anyone sell/buy at a different rate than what is announced ?

    If you are offering an above market price, then *everyone* would sell to you, and *no one* would buy from you. You become the only buyer until your funding runs out, or until there is nothing left that you do not own. This is unlike the MM, which are inelastic. Capital markets are extremely elastic. The CB recently retired 1/3 of the MBS outstanding without altering yields.

    This is why nations that peg currencies find it easy to *lower* their currency below the market rate — they can always make more of their own currency in relation to the dollar. But they can only accumulate a finite stock of dollar reserves. Even China, with all of it’s dollar reserves, has less than 1 day’s volume of dollar assets on the forex markets. No nation pegs in order to drive yields *lower*, but that is what you are proposing.

    So this case, everyone would sell their holdings to you at whatever above market price you wanted to pay, and you would need to buy all of it.

    But after you have retired the government debt, there would still be a risk-free rate that is set by the price of non-government bonds of the same maturity, adjusted for their own default risk. That is the rate at which government debt would sell for if you decided to issue it. To move that rate, you would need to also support all corporate bond prices, and retire all corporate debt. But then businesses would shift their capital structure by converting equity to debt (e.g. issuing new bonds), so that you would also need to buy up all the equity, too, bit by bit, as it was converted into bonds and then sold to you.

    So if your goal was to lower risk-free yields from 4% to 2%, that would cost you about 120 Trillion dollars in the U.S., with a wealth transfer of 60 Trillion dollars from non-asset holders to asset holders. And unless you are willing to buy up all the capital, you cannot “tell” the capital markets anything.

28. RSJ,

    I am not a PKer but if there is any chance to defend it, I pounce! There is bound to be misrepresentation.

    I am not so sure why you mention Monetarism in your discussion. Nobody here claims that the Fed targets money supply! Nobody here claims that it should target money supply. Money aggregates are not so meaningful or important. Just numbers in balance sheets of various sectors of the economy.

    2. They also didn’t realize that the equity yields need to be the long term growth rate of the economy – something that any equity investor could tell you.

    The way the S&P moved in the Noughties means equity investors simply do not know anything.

    But the PK’ers were ignorant of 1-3, and so decided that because they couldn’t explain how these rates came about, that the rates must therefore be “exogenous” and could be “administered” based on their own particular moral principles.

    They claimed that the short term rate is exogenous when the whole world was either saying it is not or talking of some natural rate of interest.

    And the coup-de-grace is that because the PKers confused the interbank market with the capital market, they decided that FF corresponds to a risk-free return that an investor receives for a zero day commitment – which is not the case – and so combining this confusion with the “moral principle” approach to yields, they decided the best thing to do was a zero FF.

    They understand that the government occupies a special place in the economy. Also please read Scott Fullwiler’s papers on who gets what in the FF market. The yield curve constructed out of prices of government securities can be made exogenous – even Bernanke says that. Yes you can argue there are thousands of other things he has said which are not true but as I had mentioned earlier, read about Australia’s tap system.

    There is only one no-arbitrage condition with government securities: implied forward rate = forward rate. If the forward market is not liquid, any bet on the yield curve is a bet, not an arbitrage. Plus the forward market is not liquid. Plus if the yield curve is set, the forward market will move, not the other way round.

    If you are offering an above market price, then *everyone* would sell to you, and *no one* would buy from you

    Donno – talking around with words. One of the counterparties in a transaction where the yield is different from an announced price is making a loss because the government has announced to buy/sell in whatever quantities. MBS purchases didn’t have an explicit announcement about the price/yield, just an announcement on quantities – $1.2T or something.

    It wouldn’t cost $120T to peg the yield curve. Simple announcement would do.

    At any rate, I do not know what your other things mean, please provided references.

29. RSJ,

    The other thing is that I think you are looking at the government as a user of the currency and not the issuer.

    The image/clip people have in their minds is that the government issued currency in the past and then it became a user just like a corporate. No explanation is given for how did the government issue currency in the past etc. However, this model is completely incorrect!

    You can think of the government spending as an efflux of currency and taxing/”borowing” as a reflux. This is how the US government spends all the time – its not a toy model. Now I have really not detailed anything in the previous statement, but once you see that, things become much more clear. There is a lot of stress on the central banking operation in this blog because the transactions which happen really confuses economists and the central banks make it look as if the government is a corporate.

30. Simple announcement would do.

    LOL,

    Play with this first:

    http://www.quantwolf.com/calculators/bonddefaultcalc.html?price=950&dp1=&face=1000&dp3=&interest=6&dp6=&dp12=&rfi1=0.08&rfi3=0.12&rfi6=0.19&rfi12=0.34

    Just an announcement, huh? Cute.

    Tell that to the Fed that just spend 1.5 Trillion buying 1/3 of MBS and failed to move the rates. They should have made an announcement instead! 😛 Maybe they used the wrong voice, or forgot the magic ink. You should have explained to them that forward rates were all that mattered, and that the government was a currency issuer, therefore MBS should always be whatever the government “announced”.

31. RSJ,

    Btw, there are several PKs who would argue like you – higher economic activity leading to higher rates etc. The data of my country points to exactly the opposite.

    Again, as I had mentioned – it is about P, not Q.

    I would buy a lot of bonds from you and sell it to the Fed if you refuse to follow the Fed in case such a thing happens 🙂