interest rates – Page 4

Liquidity Preference and Solidity Preference in the 19th Century

So I’ve been reading Homer and Sylla’s History of Interest Rates. One of the many fascinating things I’ve learned, is that in the market for federal debt, what we today call an inverted yield curve was at one time the norm.

From the book:

Three small loans floated in 1820–1821, principally to permit the continued redemption of high rate war loans, provide an interesting clue to investor preference… These were:

$4.7 million “5s of 1820,” redeemable in 1832; sold at 100 = 5%.
“6s of 1820,” redeemable at pleasure of United States; sold at 102 = 5.88%.
“5s of 1821,” redeemable in 1835; sold at 1051⁄8 =4.50%, and at 108 = 4.25%.

The yield was highest for the issue with early redemption risk and much lower for those with later redemption risks.

Nineteenth century government bonds were a bit different from modern bonds, in that the principal was repaid at the option of the borrower; repayment is usually not permitted until a certain date. [1] They were also sold with a fixed yield in terms of face value — that’s what the “5” and “6” refer to — but the actual yield depended on the discount or premium they were sold at. The important thing for our purposes is that the further away the earliest possible date of repayment is, the lower the interest rate — the opposite of the modern term premium. That’s what the passage above is saying.

The pattern isn’t limited to the 1820-21 bonds, either; it seems to exist through most of the 19th century, at least for the US. It’s the same with the massive borrowing during the Civil War:

In 1864, although the war was approaching its end, it had only been half financed. The Treasury was able to sell a large volume of bonds, but not at such favorable terms as the market price of its seasoned issues might suggest. Early in the year another $100 million of the 5–20s [bonds with a minimum maturity of 5 years and a maximum of 20] were sold and then a new longer issue was sold as follows:

1864—$75 million “6s” redeemable in 1881, tax-exempt; sold at 104.45 = 5.60%.

The Treasury soon made an attempt to sell 5s, which met with a lukewarm reception. In order to attract investors to the lower rate the Treasury extended the term to redemption from five to ten years and the maturity from twenty to forty years:

1864—$73 million “5%, 10–40s of 1864,” redeemable 1874, due in 1904, tax-exempt; sold at 100 = 5%.

Isn’t that striking? The Treasury couldn’t get investors to buy its shorter bonds at an acceptable rate, so they had to issue longer bonds instead. You wouldn’t see that story today.

The same pattern continues through the 1870s, with the new loans issue to refinance the Civil War debt. The first issue of bonds, redeemable in five to ten years sold at an interest rate of 5%; the next issue, redeemable in 13-15 years sold at 4.5%; and the last issue, redeemable in 27-29 years, sold at 4%. And it doesn’t seem like this is about expectations of a change in rates, like with a modern inverted yield curve. Investors simply were more worried about being stuck with uninvestable cash than about being stuck with unsaleable securities. This is a case where “solidity preference” dominates liquidity preference.

One possible way of explaining this in terms of Axel Leijonhufvud’s explanation of the yield curve.

The conventional story for why long loans normally have higher interest rates than short ones is that longer loans impose greater risks on lenders. They may not be able to convert the loan to cash if they need to make some payment before it matures, and they may suffer a capital loss if interest rates change during the life of the loan. But this can’t be the whole story, because short loans create the symmetric risk of not knowing what alternative asset will be available when the loan matures. In the one case, the lender risks a capital loss, but in the other case they risk getting a lower income. Why is “capital uncertainty” a greater concern than “income uncertainty”?

The answer, Leijonhufvud suggests, lies in

Keynes’ … “Vision” of a world in which currently active households must, directly or indirectly, hold their net worth in the form of titles to streams that run beyond their consumption horizon. The duration of the relevant consumption plan is limited by the sad fact that “in the Long Run, we are all dead.” But the great bulk of the “Fixed Capital of the modem world” is very long- term in nature and is thus destined to survive the generation which now owns it. This is the basis for the wealth effect of changes in asset values.

The interesting point about this interpretation of the wealth effect is that it also provides a price-theoretical basis for Keynes’ Liquidity Preference theory. … Keynes’ (as well as Hicks’) statement of this hypothesis has been repeatedly criticized for not providing any rationale for the presumption that the system as a whole wants to shed “capital uncertainty” rather than “income uncertainty.” But Keynes’ mortal consumers cannot hold land, buildings, corporate equities, British consols, or other permanent income sources “to maturity.” When the representative, risk-averting transactor is nonetheless induced by the productivity of roundabout processes to invest his savings in such income sources, he must be resigned to suffer capital uncertainty. Forward markets will therefore generally show what Hicks called a “constitutional weakness” on the demand side.

I would prefer not to express this in terms of households’ consumption plans. And I would emphasize that the problem with wealth in the form of long-lived production processes is not just that it produces income far into the future, but that wealth in this form is always in danger of losing its character as money. Once capital is embodied in a particular production process and the organization that carries it out, it tends to evolve into the means of carrying out that organization’s intrinsic purposes, instead of the capital’s own self-expansion. But for this purpose, the difference doesn’t matter; either way, the problem only arises once you have, as Leijonhufvud puts it, “a system ‘tempted’ by the profitability of long processes to carry an asset stock which turns over more slowly than [wealth owners] would otherwise want.”

The temptation of long-lived production processes is inescapable in modern economies, and explains the constant search for liquidity. But in the pre-industrial United States? I don’t think so. Long-lived means of production were much less important, and to the extent they did exist, they weren’t an outlet for money-capital. Capital’s role in production was to finance stocks of raw materials, goods in process and inventories. There was no such thing, I don’t think, as investment by capitalists in long-lived capital goods. And even land — the long-lived asset in most settings — was not really an option, since it was abundant. The early United States is something like Samuelson’s consumption-loan world, where there is no good way to convert command over current goods into future production. [2] So there is excess demand rather than excess supply for long-lasting sources of income.

The switch over to positive term premiums comes early in the 20th century. By the 1920s, short-term loans in the New York market consistently have lower rates than corporate bonds, and 3-month Treasury bills have rates below longer bonds. Of course the organization of financial markets changed quite a lot in this period too, so one wouldn’t want to read too much into this timing. But it is at least consistent with the Leijonhufvud story. Liquidity preference becomes dominant in financial markets only once there has been a decisive shift toward industrial production by long-lived firm using capital-intensive techniques, and once claims on those firms has become a viable outlet for money-capital.

* * *

A few other interesting points about 19th century US interest rates. First, they were remarkably stable, at least before the 1870s. (This fits with the historical material on interest rates that Merijn Knibbe has been presenting in his excellent posts at Real World Economics Review.)

Second, there’s no sign of a Fisher equation. Nominal interest rates do not respond to changes in the price level, at all. Homer and Sylla mention that in earlier editions of the book, which dealt less with the 20th century, the concept of a “real” interest rate was not even mentioned.

As you can see from this graph, none of the major inflations or deflations between 1850 and 1960 had any effect on nominal interest rates. The idea that there is a fundamentals-determined “real” interest rate while the nominal rate adjusts in response to changes in the price level, clearly has no relevance outside the past 50 years. (Whether it describes the experience of the past 50 years either is a question for another time.)

Finally, there is no sign of “crowding out” of private by public borrowing. It is true that the federal government did have to pay somewhat higher rates during the periods of heavy borrowing (and of course also political uncertainty) in the War of 1812 and the Civil War. But rates for other borrowers didn’t budge. And on the other hand, the surpluses that resulted in the redemption of the entire debt in the 1830s didn’t deliver lower rates for other borrowers. Homer and Sylla:

Boston yields were about the same in 1835, when the federal debt was wiped out, as they were in 1830; this reinforces the view that there was little change in going rates of long-term interest during this five- year period of debt redemption.

If government borrowing really raises rates for private borrowers, you ought to see it here, given the absence of a central bank for most of this period and the enormous scale of federal borrowing during the Civil War. But you don’t.

[1] It seems that most, though not all, bonds were repaid at the earliest possible redemption date, so it is reasonably similar to the maturity of a modern bond.

[2] Slaves are the big exception. So the obvious test for the argument I am making here would be to find the modern pattern of term premiums in the South. Unfortunately, Homer and Sylla aren’t any help on this — it seems the only local bond markets in this period were in New England.

The Interest Rate, the Interest Rate, and Secular Stagnation

In the previous post, I argued that the term “interest rate” is used to refer to two basically unrelated prices: The exchange rate between similar goods at different periods, and the yield on a credit-market instrument. Why does this distinction matter for secular stagnation?

Because if you think the “natural rate of interest,” in the sense of the credit-market rate that brings aggregate expenditure to a desired level in some real-world economic situation, should be the time-substitution rate that would exist in a model that somehow corresponds to that situation, when the two are in fact unrelated — well then, you are going to end up with a lot of irrelevant and misleading intuitions about what that rate should be.

In general, I do think the secular stagnation conversation is a real step forward. So it’s a bit frustrating, in this context, to see Krugman speculating about the “natural rate” in terms of a Samuelson-consumption loan model, without realizing that the “interest rate” in that model is the intertemporal substitution rate, and has nothing to do with the Wicksellian natural rate. This was the exact confusion introduced by Hayek, which Sraffa tore to pieces in his review, and which Keynes went to great efforts to avoid in General Theory. It would be one thing if Krugman said, “OK, in this case Hayek was right and Keynes was wrong.” But in fact, I am sure, he has no idea that he is just reinventing the anti-Keynesian position in the debates of 75 years ago.

The Wicksellian natural rate is the credit-market rate that, in current conditions, would bring aggregate expenditure to the level desired by whoever is setting monetary policy. Whether or not there is a level of expenditure that we can reliably associate with “full employment” or “potential output” is a question for another day. The important point for now is “in current conditions.” The level of interest-sensitive expenditure that will bring GDP to the level desired by policymakers depends on everything else that affects desired expenditure — the government fiscal position, the distribution of income, trade propensities — and, importantly, the current level of income itself. Once the positive feedback between income and expenditure has been allowed to take hold, it will take a larger change in the interest rate to return the economy to its former position than it would have taken to keep it there in the first place.

There’s no harm in the term “natural rate of interest” if you understand it to mean “the credit market interest rate that policymakers should target to get the economy to the state they think it should be in, from the state it in now.”And in fact, that is how working central bankers do understand it. But if you understand “natural rate” to refer to some fundamental parameter of the economy, you will end up hopelessly confused. It is nonsense to say that “We need more government spending because the natural rate is low,” or “we have high unemployment because the natural rate is low.” If G were bigger, or if unemployment weren’t high, there would be a different natural rate. But when you don’t distinguish between the credit-market rate and time-substitution rate, this confusion is unavoidable.

Keynes understood clearly that it makes no sense to speak of the “natural rate of interest” as a fundamental characteristic of an economy, independent of the current state of aggregate demand:

In my Treatise on Money I defined what purported to be a unique rate of interest, which I called the natural rate of interest — namely, the rate of interest which, in the terminology of my Treatise, preserved equality between the rate of saving (as there defined) and the rate of investment. I believed this to be a development and clarification of Wicksell’s “natural rate of interest”, which was, according to him, the rate which would preserve the stability if some, not quite clearly specified, price-level.

I had, however, overlooked the fact that in any given society there is, on this definition, a different natural rate of interest for each hypothetical level of employment. And, similarly, for every rate of interest there is a level of employment for which that rate is the “natural” rate, in the sense that the system will be in equilibrium with that rate of interest and that level of employment. Thus it was a mistake to speak of the natural rate of interest or to suggest that the above definition would yield a unique value for the rate of interest irrespective of the level of employment. I had not then understood that, in certain conditions, the system could be in equilibrium with less than full employment.

I am now no longer of the opinion that the concept of a “natural” rate of interest, which previously seemed to me a most promising idea, has anything very useful or significant to contribute to our analysis. It is merely the rate of interest which will preserve the status quo; and, in general, we have no predominant interest in the status quo as such.

EDIT: In response to Nick Edmonds in comments, I’ve tried to restate the argument of these posts in simpler and hopefully clearer terms:

Step 1 is to recognize that in a model like Samuelson’s, “interest rate” just means any contract that allows you to make a payment today and receive a flow of income in the future. It would be the exact same model, capturing the exact same features of the economy, if we wrote “profit rate” or “house price-to-rent ratio” instead of “interest rate.” Any valid intuition the model gives us, applies to ALL asset yields, not just to the the credit-instrument yields that we call “interest rates” in every day life.

Step 2 is to think about the other factors that enter into real-world asset yields, besides the intertemporal exchange rate Samuelson is interested in — risk, liquidity, carrying costs and depreciation, and expected capital gains. Since all real-world asset yields incorporate at least one of these factors, none correspond exactly to Samuelson’s intertemporal interest rate.

Step 3 is to realize that not only are credit-instrument yields not exactly the Samuelson “interest rate,” they aren’t even approximately it. The great majority of credit market transactions we see in real economies are not exchanges of present income for future income, but exchanges of two different claims on future income. So the intertemporal interest rate enters on both sides and cancels out.

At that point, we have established that the “interest rate” the monetary authority is targeting is not the “interest rate” Samuelson is writing about.

Step 4 is then to ask, what does it mean to say that some particular credit-market interest rate is the “natural” one? That is where the dependence on fiscal policy, income distribution, etc. come in. But those factors are not part of the argument for why the credit-market rate is not even approximately the intertemporal rate.

The Interest Rate and the Interest Rate

We will return to secular stagnation. But we need to clear some ground first. What is an interest rate?

Imagine you are in a position to acquire a claim on a series of payments in the future. Since an asset is just anything that promises a stream of payments in the future, we will say you are thinking of buying of an asset. What will you look at to make your decision?

First is the size of the payments you will receive, as a fraction of what you pay today. We will call that the yield of the asset, or y. Against that we have to set the risk that the payments may be different from expected or not occur at all; we will call the amount you reduce your expected yield to account for this risk r. If you have to make regular payments beyond the purchase of the asset to receive income from it (perhaps taxes, or the costs of operating the asset if it is a capital good) then we also must subtract these carrying costs c. In addition, the asset may lose value over time, in which case we have to subtract the depreciation rate d. (In the case of an asset that only lasts one period — a loan to be paid back in full the next period, say — d will be equal to one.) On the other hand, owning an asset can have benefits beyond the yield. In particular, an asset can be sold or used as collateral. If this is easy to do, ownership of the asset allows you to make payments now, without having to waiting for its yield in the future. We call the value of the asset for making unexpected payments its liquidity premium, l. The market value of long-lasting assets may also change over time; assuming resale is possible, these market value changes will produce a capital gain g (positive or negative), which must be added to the return. Finally, you may place a lower value on the payments from the asset simply because they take place in the future; this might be because your needs now are more urgent than you expect them to be then, or simply because you prefer income in the present to income in the future. Either way, we have to subtract this pure time-substitution rate i.

So the value of an asset costing one unit (of whatever numeraire) will be 1 + y – r – c – d + l + g – i.

(EDIT: On rereading, this could use some clarification:

Of course all the terms can take on different (expected) values in different time periods, so they are vectors, not scalars. But if we assume they are constant, and that the asset lasts forever (i.e. a perpetuity), then we should write its equilibrium value as: V = Y/i, where Y is the total return in units of the numeraire, i.e. Y = V(y – r – c + l + g) and i is the discount rate. Divide through both sides by V and we have i = y – r – c + l + g. We can now proceed as below.)

In equilibrium, you should be just indifferent between purchasing and not purchasing this asset, so we can write:

y – r – c – d + l + g – i = 0, or

(1) y = r + c + d – l – g + i

So far, there is nothing controversial.

In formal economics, from Bohm-Bawerk through Cassel, Fisher and Samuelson to today’s standard models, the practice is to simplify this relationship by assuming that we can safely ignore most of these terms. Risk, carrying costs and depreciation can be netted out of yields, capital gains must be zero on average, and liquidity is assumed not to matter or just ignored. So then we have:

(2) y = i

In these models, it doesn’t matter if we use the term “interest rate” to mean y or to mean i, since they are always the same.

This assumption is appropriate for a world where there is only one kind of asset — a risk-free contract that exchanges one good in the present for 1 + i goods in the future. There’s nothing wrong with exploring what the value of i would be in such a world under various assumptions.

The problem arises when we carry equation (2) over to the real world and apply it to the yield of some particular asset. On the one hand, the yield of every existing asset reflects some or all of the other terms. And on the other hand, every contract that involves payments in more than one period — which is to say, every asset — equally incorporates i. If we are looking for the “interest rate” of economic theory in the economic world we observe around us, we could just as well pick the rent-price ratio for houses, or the profit rate, or the deflation rate, or the ratio of the college wage premium to tuition costs. These are just the yields of a house, of a share of the capital stock, of cash and of a college degree respectively. All of these are a ratio of expected future payments to present cost, and should reflect i to exactly the same extent as the yield of a bond does. Yet in everyday language, it is the yield of the bond that we call “interest”, even though it has no closer connection to the interest rate of theory than any of these other yields do.

This point was first made, as far as I know, by Sraffa in his review of Hayek’s Prices and Production. It was developed by Keynes, and stated clearly in chapters 13 and 17 of the General Theory.

For Keynes, there is an additional problem. The price we observe as an “interest rate” in credit markets is not even the y of the bond, which would be i modified by risk, expected capital gains and liquidity. That is because bonds do not trade against baskets of goods. They trade against money. When we see a bond being sold with a particular yield, we are not observing the exchange rate between a basket of goods equivalent to the bond’s value today and baskets of goods equivalent to its yield in the future. We are observing the exchange rate between the bond today and a quantity of money today. That’s what actually gets exchanged. So in equilibrium the price of the bond is what equates the expected returns on the two assets:

(3) y_B – r_B + l_B + g_B – i = l_M – i

(Neither bonds nor money depreciate or have carrying costs, and money has no risk. If our numeraire is money then money also cannot experience capital gains. If our numeraire was a basket of goods instead, then -g would be expected inflation, which would appear on both sides and cancel out.)

What we see is that i appears on both sides, so it cancels out. The yield of the bond is given by:

(4) y_B = r_B – g_B + (l_M – l_B)

The yield of the bond — the thing that in conventional usage we call the “interest rate” — depends on the risk of the bond, the expected price change of the bond, and the liquidity premium of money compared with the bond. Holding money today, and holding a bond today, are both means to enable you to make purchases in the future. So the intertemporal substitution rate i does not affect the bond yield.

(We might ask whether the arbitrage exists that would allow us to speak of a general rate of time-substitution i in real economies at all. But for present purposes we can ignore that question and focus on the fact that even if there is such a rate, it does not show up in the yields we normally call “interest rates”.)

This is the argument as Keynes makes it. It might seem decisive. But monetarists would reject it on the grounds that nobody in fact holds money as a store of value, so equation (3) does not apply. The bond-money market is not in equilibrium, because there is zero demand for money beyond that needed for current transactions at any price. (The corollary of this is the familiar monetarist claim that any change in the stock of money must result in a proportionate change in the value of transactions, which at full employment means a proportionate rise in the price level.) From the other side, endogenous money theorists might assert that the money supply is infinitely elastic for any credit-market interest rate, so l_M is endogenous and equation (4) is underdetermined.

As criticisms of the specific form of Keynes’ argument, these are valid objections. But if we take a more realistic view of credit markets, we come to the same conclusion: the yield on a credit instrument (call this the “credit interest rate”) has no relationship to the intertemporal substitution rate of theory (call this the “intertemporal interest rate.”)

Suppose you are buying a house, which you will pay for by taking out a mortgage equal to the value of the house. For simplicity we will assume an amortizing mortgage, so you make the same payment each period. We can also assume the value of housing services you receive from the house will also be the same each period. (In reality it might rise or fall, but an expectation that the house will get better over time is obviously not required for the transaction to take place.) So if the purchase is worth making at all, then it will result in a positive income to you in every period. There is no intertemporal substitution on your side. From the bank’s point of view, extending the mortgage means simultaneously creating an asset — their loan to you — and a liability — the newly created deposit you use to pay for the house. If the loan is worth making at all, then the expected payments from the mortgage exceed the expected default losses and other costs in every period. And the deposits are newly created, so no one associated with the bank has to forego any other expenditure in the present. There is no intertemporal substitution on the bank’s side either.

(It is worth noting that there are no net lenders or net borrowers in this scenario. Both sides have added an asset and a liability of equal value. The language of net lenders and net borrowers is carried over from models with consumption loans at the intertemporal interest rate. It is not relevant to the credit interest rate.)

If these transactions are income-positive for all periods for both sides, why aren’t they carried to infinity? One reason is that the yields for the home purchaser fall as more homes are purchased. In general, you will not value the housing services from a second home, or the additional housing services of a home that costs twice as much, as much as you value the housing services of the home you are buying now. But this only tells us that for any given interest rate there is a volume of mortgages at which the market will clear. It doesn’t tell us which of those mortgage volume-interest rate pairs we will actually see.

The answer is on the liquidity side. Buying a house makes you less liquid — it means you have less flexibility if you decide you’d like to move elsewhere, or if you need to reduce your housing costs because of unexpected fall in income or rise in other expenses. You also have a higher debt-income ratio, which may make it harder for you to borrow in the future. The loan also makes the bank less liquid — since its asset-capital ratio is now higher, there are more states of the world in which a fall in income would require it to sell assets or issue new liabilities to meet its scheduled commitments, which might be costly or, in a crisis, impossible. So the volume of mortgages rises until the excess of housing service value over debt service costs make taking out a mortgage just worth the incremental illiquidity for the marginal household, and where the excess of mortgage yield over funding costs makes issuing a new mortgage just worth the incremental illiquidity for the marginal bank. (Incremental illiquidity in the interbank market may — or may not — mean that funding costs rise with the volume of loans, but this is not necessary to the argument.)

Monetary policy affects the volume of these kinds of transactions by operating on the l terms. Normally, it does so by changing the quantity of liquid assets available to the financial system (and perhaps directly to the nonfinancial private sector as well). In this way the central bank makes banks (and perhaps households and businesses) more or less willing to accept the incremental illiquidity of a new loan contract. Monetary policy has nothing to do with substitution between expenditure in the present period and expenditure in some future period. Rather, it affects the terms of substitution between more and less liquid claims on income in the same future period.

Note that changing the quantity of liquid assets is not the only way the central bank can affect the liquidity premium. Banking regulation, lender of last resort operations and bailouts also change the liquidity premium, by chaining the subjective costs of bank balance sheet expansion. An expansion of the reserves available to the banking system makes it cheaper for banks to acquire a cushion to protect themselves against the possibility of an unexpected fall in income. This will make them more willing to hold relatively illiquid assets like mortgages. But a belief that the Fed will take emergency action prevent a bank from failing in the event of an unexpected fall in income also increases its willingness to hold assets like mortgages. And it does so by the same channel — reducing the liquidity premium. In this sense, there is no difference in principle between monetary policy and the central bank’s role as bank supervisor and lender of last resort. This is easy to understand once you think of “the interest rate” as the price of liquidity, but impossible to see when you think of “the interest rate” as the price of time substitution.

It is not only the central bank that changes the liquidity premium. If mortgages become more liquid — for instance through the development of a regular market in securitized mortgages — that reduces the liquidity cost of mortgage lending, exactly as looser monetary policy would.

The irrelevance of the time-substitution rate i to the credit-market interest rate y_B becomes clear when you compare observed interest rates with other prices that also should incorporate i. Courtesy of commenter rsj at Worthwhile Canadian Initiative, here’s one example: the Baa bond rate vs. the land price-rent ratio for residential property.

Both of these series are the ratio of one year’s payment from an asset, to the present value of all future payments. So they have an equal claim to be the “interest rate” of theory. But as we can see, none of the variation in credit-market interest rates (y_B, in my terms) show up in the price-rent ratio. Since variation in the time-substituion rate i should affect both ratios equally, this implies that none of the variation in credit-market interest rates is driven by changes in the time-substitution interest rate. The two “interest rates” have nothing to do with each other.

(Continued here.)

EDIT: Doesn’t it seem strange that I first assert that mortgages do not incorporate the intertemporal interest rate, then use the house price-rent ratio as an example of a price that should incorporate that rate? One reason to do this is to test the counterfactual claim that interest rates do, after all, incorporate Samuelson’s interest rate i. If i were important in both series, they should move together; if they don’t, it might be important in one, or in neither.

But beyond that, I think housing purchases do have an important intertemporal component, in a way that loan contracts do not. That’s because (with certain important exceptions we are all aware of) houses are not normally purchased entirely on credit. A substantial fraction of the price is paid is upfront. In effect, most house purchases are two separate transactions bundled together: A credit transaction (for, say, 80 percent of the house value) in which both parties expect positive income in all periods, at the cost of less liquid balance sheets; and a conceptually separate cash transaction (for, say, 20 percent) in which the buyer foregoes present expenditure in return for a stream of housing services in the future. Because house purchases must clear both of these markets, they incorporate i in way that loans do not. But note, i enters into house prices only to the extent that the credit-market interest rate does not. The more important the credit-market interest rate is in a given housing purchase, the less important the intertemporal interest rate is.

This is true in general, I think. Credit markets are not a means of trading off the present against the future. They are a means of avoiding tradeoffs between the present and the future.

Secular Stagnation, Progress in Economics

It’s the topic of the moment. Our starting point is this Paul Krugman post, occasioned by this talk by Lawrence Summers.

There are two ways to understand “secular stagnation.” One is that the growth rate of income and output will be slower in the future. The other is that there will be a systematic tendency for aggregate demand to fall short of the economy’s potential output. It’s the second claim that we are interested in.

For Krugman, the decisive fact about secular stagnation is that it implies a need for persistently negative interest rates. That achieved, there is no implication that growth rates or employment need to be lower in the future than in the past. He is imagining a situation where current levels of employment and growth rates are maintained, but with permanently lower interest rates.

We could also imagine a situation where full employment was maintained by permanently higher public spending, rather than lower interest rates. Or we could imagine a situation where nothing closed the gap and output fell consistently short of potential. What matters is that aggregate expenditure by the private sector tends to fall short of the economy’s potential output, by a growing margin. For reasons I will explain down the road, I think this is a better way of stating the position than a negative “natural rate” of interest.

I think this conversation is a step forward for mainstream macroeconomic thought. There are further steps still to take. In this post I describe what, for me, are the positive elements of this new conversation. In subsequent posts, I will talk about the right way of analyzing these questions more systematically — in terms of a Harrod-type growth model — and about the wrong way — in terms of the natural rate of interest.

The positive content of “secular stagnation”

1. Output is determined by demand.

The determination of total output by total expenditure is such a familiar part of the macroeconomics curriculum that we forget how subversive it is. It denies the logic of scarcity that is the basis of economic analysis and economic morality. Since Mandeville’s Fable of the Bees, it’s been recognized that if aggregate expenditure determines aggregate income, then, as Krugman says, “vice is virtue and virtue is vice.”

A great deal of the history of macroeconomics over the past 75 years can be thought of as various efforts to expunge, exorcize or neutralize the idea of demand-determined income, or at least to safely quarantine it form the rest of economic theory. One of the most successful quarantine strategies was to recast demand constraints on aggregate output as excess demand for money, or equivalently as the wrong interest rate. What distinguished real economies from the competitive equilibrium of Jevons or Walras was the lack of a reliable aggregate demand “thermostat”. But if a central bank or other authority set that one price or that one quantity correctly, economic questions could again be reduced to allocation of scarce means to alternative ends, via markets. Both Hayek and Friedman explicitly defined the “natural rate” of interest, which monetary policy should maintain, as the rate that would exist in a Walrasian barter economy. In postwar and modern New Keynesian mainstream economics, the natural rate is defined as the market interest rate that produces full employment and stable prices, without (I think) explicit reference to the intertemporal exchange rate that is called the interest rate in models of barter economies. But he equivalence is still there implicitly, and is the source of a great deal of confusion.

I will return to the question of what connection there is, if any, between the interest rates we observe in the world around us, and what a paper like Samuelson 1958 refers to as the “interest rate.” The important thing for present purposes is:

Mainstream economic theory deals with the problems raised when expenditure determines output, by assuming that the monetary authority sets an interest rate such that expenditure just equals potential output. If such a policy is followed successfully, the economy behaves as if it were productive capacity that determined output. Then, specifically Keynesian problems can be ignored by everyone except the monetary-policy technicians. One of the positive things about the secular stagnation conversation, from my point of view, is that it lets Keynes back out of this box.

That said, he is only partway out. Even if it’s acknowledged that setting the right interest rate does not solve the problem of aggregate demand as easily as previously believed, the problem is still framed in terms of the interest rate.

2. Demand normally falls short of potential

Another strategy to limit the subversive impact of Keynes has been to consign him to the sublunary domain of the short run, with the eternal world of long run growth still classical. (It’s a notable — and to me irritating — feature of macroeconomics textbooks that the sections on growth seem to get longer over time, and to move to the front of the book.) But if deviations from full employment are persistent, we can’t assume they cancel out and ignore them when evaluating an economy’s long-run trajectory.

One of the most interesting parts of the Summers talk came when he said, “It is a central pillar of both classical models and Keynesian models, that it is all about fluctuations, fluctuations around a given mean.” (He means New Keynesian models here, not what I would consider the authentic Keynes.) “So what you need to do is have less volatility.” He introduces the idea of secular stagnation explicitly as an alternative to this view that demand matters only for the short run. (And he forthrightly acknowledges that Stanley Fischer, his MIT professor who he is there to praise, taught that demand is strictly a short-run phenomenon.) The real content of secular stagnation, for Summers, is not slower growth itself, but the possibility that the same factors that can cause aggregate expenditure to fall short of the economy’s potential output can matter in the long run as well as in the short run.

Now for Summers and Krugman, there still exists a fundamentals-determined potential growth rate, and historically the level of activity did fluctuate around it in the past. Only in this new era of secular stagnation, do we have to consider the dynamics of an economy where aggregate demand plays a role in long-term growth. From my point of view, it’s less clear that anything has changed in the behavior of the economy. “Secular stagnation” is only acknowledging what has always been true. The notion of potential output was never well defined. Labor supply and technology, the supposed fundamentals, are strongly influenced by the level of capacity utilization. As I’ve discussed before, once you allow for Verdoorn’s Law and hysteresis, it makes no sense to talk about the economy’s “potential growth rate,” even in principle. I hope the conversation may be moving in that direction. Once you’ve acknowledged that the classical allocation-of-scarce-means-to-alternative-ends model of growth doesn’t apply in present circumstances, it’s easier to take the next step and abandon it entirely.

3. Bubbles are functional

One widely-noted claim in the Summers talk is that asset bubbles have been a necessary concomitant of full employment in the US since the 1980s. Before the real estate bubble there was the tech bubble, and before that there was the commercial real estate bubble we remember as the S&L crisis. Without them, the problem of secular stagnation might have posed itself much earlier.

This claim can be understood in several different, but not mutually exclusive, senses. It may be (1) interest rates sufficiently low to produce full employment, are also low enough to provoke a bubble. It may be (2) asset bubbles are an important channel by which monetary policy affects real activity. Or it may be (3) bubbles are a substitute for the required negative interest rates. I am not sure which of these claims Summers intends. All three are plausible, but it is still important to distinguish them. In particular, the first two imply that if interest rates could fall enough to restore full employment, we would have even more bubbles — in the first case, as an unintended side effect of the low rates, in the second, as the channel through which they would work. The third claim implies that if interest rates could fall enough to restore full employment, it would be possible to do more to restrain bubbles.

An important subcase of (1) comes when there is a minimum return that owners of money capital can accept. As Keynes said (in a passage I’m fond of quoting),

The most stable, and the least easily shifted, element in our contemporary economy has been hitherto, and may prove to be in future, the minimum rate of interest acceptable to the generality of wealth-owners.[2] If a tolerable level of employment requires a rate of interest much below the average rates which ruled in the nineteenth century, it is most doubtful whether it can be achieved merely by manipulating the quantity of money. Cf. the nineteenth-century saying, quoted by Bagehot, that “John Bull can stand many things, but he cannot stand 2 per cent.”

If this is true, then asking owners of money wealth to accept rates of 2 percent, or perhaps much less, will face political resistance. More important for our purposes, it will create an inclination to believe the sales pitch for any asset that offers an acceptable return.

Randy Wray says that Summers is carrying water here for his own reputation and his masters in Finance. The case for bubbles as necessary for full employment justifies his past support for financial deregulation, and helps make the case against any new regulation in the future. That may be true. But I still think he is onto something important. There’s a long-standing criticism of market-based finance that it puts an excessive premium on liquidity and discourages investment in long-lived assets. A systematic overestimate of the returns from fixed assets might be needed to offset the systematic overestimate of the costs of illiquidity.

Another reason I like this part of Summers’ talk is that it moves us toward recognizing the fundamental symmetry between between monetary policy conventionally defined, lender of last resort operations and bank regulations. These are different ways of making the balance sheets of the financial sector more or less liquid. The recent shift from talking about monetary policy setting the money stock to talking about setting interest interest rates was in a certain sense a step toward realism, since there is nothing in modern economies that corresponds to a quantity of money. But it was also a step toward greater abstraction, since it leaves it unclear what is the relationship between the central bank and the banking system that allows the central bank to set the terms of private credit transactions. Self-interested as it may be, Summers call for regulatory forbearance here is an intellectual step forward. It moves us toward thinking of what central banks do neither in terms of money, nor in terms of interest rates, but in terms of liquidity.

Finally, note that in Ben Bernanke’s analysis of how monetary policy affects output, asset prices are an important channel. That is an argument for version (2) of the bubbles claim.

4. High interest rates are not coming back

For Summers and Krugman, the problem is still defined in terms of a negative “natural rate” of interest. (To my mind, this is the biggest flaw in their analysis.) So much of the practical discussion comes down to how you convince or compel wealth owners to hold assets with negative yields. One solution is to move to permanently higher inflation rates. (Krugman, to his credit, recognizes that this option will only be available when or if something else raises aggregate demand enough to push against supply constraints.) I am somewhat skeptical that capitalist enterprises in their current form can function well with significantly higher inflation. The entire complex of budget and invoicing practices assumes that over some short period — a month, a quarter, even a year — prices can be treated as constant. Maybe this is an easy problem to solve, but maybe not. Anyway, it would be an interesting experiment to find out!

More directly relevant is the acknowledgement that interest rates below growth rates may be a permanent feature of the economic environment for the foreseeable future. This has important implications for debt dynamics (both public and private), as we’ve discussed extensively on this blog. I give Krugman credit for saying that with i < g, it is impossible for debt to spiral out of control; a deficit of any level, maintained forever, will only ever cause the debt-GDP ratio to converge to some finite level. (I also give him credit for acknowledging that this is a change in his views.) This has the important practical effect of knocking another leg out from the case for austerity. It’s been a source of great frustration for me to see so many liberal, “Keynesian” economists follow every argument for stimulus with a pious invocation of the need for long-term deficit reduction. If people no longer feel compelled to bow before that shrine, that is progress.

On a more abstract level, the possibility of sub-g or sub-zero interest rates helps break down the quarantining of Keynes discussed above. Mainstream economists engage in a kind of doublethink about the interest rate. In the context of short-run stabilization, it is set by the central bank. But in other contexts, it is set by time preferences and technological tradeoff between current and future goods. I don’t think there was ever any coherent way to reconcile these positions. As I will explain in a following post, the term “interest rate” in these two contexts is being used to refer to two distinct and basically unrelated prices. (This was the upshot of the Sraffa-Hayek debate.) But as long as the interest rate observed in the world (call it the “finance” interest rate) behaved similarly enough to the interest rate in the models (the “time-substitution” interest rate), it was possible to ignore this contradiction without too much embarrassment.

There is no plausible way that the “time substitution” interest rate can be negative. So the secular stagnation conversation is helping reestablish the point — made by Keynes in chapter 17 of the General Theory, but largely forgotten — that the interest rates we observe in the world are something different. And in particular, it is no longer defensible to treat the interest rate as somehow exogenous to discussions about aggregate demand and fiscal policy. When I was debating fiscal policy with John Quiggin, he made the case for treating debt sustainability as a binding constraint by noting that there are long periods historically when interest rates were higher than growth rates. It never occurred to him that it makes no sense to talk about the level of interest rates as an objective fact, independent of the demand conditions that make expansionary fiscal policy desirable. I don’t mean to pick on John — at the time it wasn’t clear to me either.

Finally, on the topic of low interest forever, I liked Krugman’s scorn for the rights of interest-recipients:

How dare anyone suggest that virtuous individuals, people who are prudent and save for the future, face expropriation? How can you suggest steadily eroding their savings either through inflation or through negative interest rates? It’s tyranny!
But in a liquidity trap saving may be a personal virtue, but it’s a social vice. And in an economy facing secular stagnation, this isn’t just a temporary state of affairs, it’s the norm. Assuring people that they can get a positive rate of return on safe assets means promising them something the market doesn’t want to deliver – it’s like farm price supports, except for rentiers.

It’s a nice line, only slightly spoiled by the part about “what the market wants to deliver.” The idea that it is immoral to deprive the owners of money wealth of their accustomed returns is widespread and deeply rooted. I think it lies behind many seemingly positive economic claims. If this conversation develops, I expect we will see more open assertions of the moral entitlement of the rentiers.

Don’t Touch the Yield

There’s a widespread idea in finance and economics land that there’s something wrong, dangerous, even unnatural about persistently low interest rates.

This idea takes its perhaps most reasonable form in arguments that the fundamental cause of the Great Financial Crisis was rates that were “far too low for far too long,” and that continued low interest rates, going forward, will only encourage speculation and new asset bubbles. Behind, or anyway alongside, these kinds of claims is a more fundamentally ideological view, that owners of financial assets are morally entitled to their accustomed returns, and woe betide the society or central banker that deprives them of the fruit of their non-labor. You hear this when certain well-known economists describe low rates as the “rape and plunder” of bondowners, or when Jim Grant says that the real victims of the recession are investors in money-market funds.

I want to look today at the “reaching for yield” version of this argument, which Brad Delong flagged as PRIORITY #1 RED FLAG OMEGA for the econosphere after it was endorsed by the Federal Reserve’s Jeremy Stein. [1] In DeLong’s summary:

Bankers want profits. … And a bank has costs above and beyond the returns on its portfolio. For each dollar of deposits it collects, a bank must spend 2.5 cents per year servicing those deposits. In normal times, when interest rates are well above 2.5 percent per year, banks have a normal, sensible attitude to risk and return. They will accept greater risk only if they come with returns higher enough to actually diminish the chances of reporting a loss. But when interest rates fall low enough that even the most sensible portfolio cannot reliably deliver a return on the portfolio high enough to cover the 2.5 cent per year cost of managing deposits, a bank will “reach for yield” and start writing correlated unhedged out-of-the-money puts so that it covers its 2.5 percent per year hurdle unless its little world blows up. Banks stop reducing their risk as falling returns mean that diversification and margin can no longer be counted on to manage them but instead embrace risks.

It is Stein’s judgment that right now whatever benefits are being provided to employment and production by the Federal Reserve’s super-sub-normal interest rate policy and aggressive quantitative easing are outweighed by the risks being run by banks that are reaching for yield.

Now on one level, this just seems like a non-sequitur. “Banks holding more risky assets” is, after all, just another way of saying “banks making more loans.” In fact, it’s hard to see how monetary policy is ever supposed to work if we rule out the possibility of shifting banks’ demand for risky private assets. [1] An Austrian, I suppose, might follow this logic to its conclusion and reject the idea of monetary policy in general; but presumably not an Obama appointee to the Fed.

But there’s an even more fundamental problem, not only with the argument here but with the broader idea — shared even by people who should know better — that low interest rates hurt bank profits. It’s natural to think that banks receive interest payments, so lower interest means less money for the bankers. But that is wrong.

Banks are the biggest borrowers as well as the biggest lenders in the economy, so what matters is not the absolute level of interest rates, but the spread — the difference between the rate at which banks borrow and the rate at which they lend. A bank covers its costs as reliably borrowing at 1 percent and lending at 4, as it does borrowing at 3 percent and lending at 6. So if we want to argue that monetary policy affects the profitability of bank lending, we have to argue that it has a differential effect on banks funding costs and lending rates.

For many people making the low-rates-are-bad-for-banks argument, this differential effect may come from a mental model in which the main bank liabilities are non-interest-bearing deposits. Look at the DeLong quote again — in the world it’s describing, banks pay a fixed rate on their liabilities. And at one point that is what the real world looked like too.

In 1960, non-interest-bearing deposits made up over two-thirds of total bank liabilities. In a system like that, it’s natural to see the effect of monetary policy as mainly on the asset side of bank balance sheets. But today’s bank balance sheets look very different: commercial banks now pay interest on around 80 percent of their liabilities. So it’s much less clear, a priori, why policy changes should affect banks interest income more than their funding costs. Since banks borrow short and lend long (that’s sort of what it means to be a bank), and since monetary policy has its strongest effects at shorter maturities, one might even expect the effect on spreads to go the other way.

And in fact, when we look at the data, that is what we see.

Average interest rate paid (red) and received (blue) by commercial banks. Source: FDIC

The black line with diamonds is the Federal Funds rate, set by monetary policy. The blue line is the average interest rate charged by commercial banks on all loans and leases; the solid red line is their average funding cost; and the dotted red line is the average interest rate on commercial banks’ interest-bearing liabilities. [3] As the figure shows, in the 1950s and ’60s changes in the federal funds rate didn’t move banks’ funding costs at all, while they did have some effect on loan rates; the reach-for-yield story might have made sense then. But in recent decades, as banks’ pool of cheap deposit funding has dried up, bank funding costs have become increasingly sensitive to the policy rate.

Looking at the most recent cycle, the decline in the Fed Funds rate from around 5 percent in 2006-2007 to the zero of today has been associated with a 2.5 point fall in bank funding costs but only a 1.5 point fall in bank lending rates — in other words, a one point increase in spreads. The same relationship, though weaker, is present in the previous two cycles, but not before. More generally, the correlation of changes in the federal funds rate and changes in bank spreads is 0.49 for 1955-1980, but negative 0.38 for the years 1991-2001. So Stein’s argument fails at the first step. If low bank margins are the problem, then “super-sub-normal interest rate policy” is the solution.

Let’s walk through this again. The thing that banks care about is the difference between what it costs them to borrow, and what they can charge to lend. Wider spreads mean lending is more profitable, narrower spreads mean it’s less so. And if banks need a minimum return on their lending — to cover fixed costs, or to pay executives expected bonuses or whatever — then when spreads get too narrow, banks may be tempted to take underprice risk. That’s “reaching for yield.” So turning to the figure, the spread is the space between the solid red line and the solid blue one. As we can see, in the 1950s and ’60s, when banks funded themselves mostly with deposits, the red line — their borrowing costs — doesn’t move at all with the federal funds rate. So for instance the sharp tightening at the end of the 1960s raises average bank lending rates by several points, but doesn’t move bank borrowing rates at all. So in that period, a high federal funds rate means wide bank spreads, and a low federal funds rate means narrower spreads. In that context the “reaching for yield” story has a certain logic (which is not to say it would be true, or important.) But since the 1980s, the red line — bank funding costs — has become much more responsive to the federal funds rate, so this relationship between monetary policy and bank spreads no longer exists. If anything, as I said, the correlation runs in the opposite direction.

Short version: When banks are funded by non-interest bearing deposits, low interest rates can hurt their profits, which makes them have a sad face. But when banks pay interest on almost all their liabilities, as today, low rates make them have a happy face. [4] In which case there’s no reason for them to reach for yield.

Now, it is true that the Fed has also intervened directly in the long end, where one might expect the impact on bank lending rates to be stronger. This is specifically the focus of a speech by Stein last October, where he explicitly said that if the policy rate were currently 3 percent he would have no objection to lowering it, but that he was more worried about unconventional policy to directly target long rates. [5] He offers a number of reasons why a fall in long rates due an expectation of lower short rates in the future would be expansionary, but a fall in long rates due to a lower term premium might not be. Frankly I find all these explanations ad-hoc and hand-wavey. But the key point for present purposes is that unconventional policy does not involve the central bank setting some kind of regulatory ceiling on long rates; rather, it involves lowering long rates via voluntary transactions with lenders. The way the Fed lowers rates on long bonds is by raising their price; the way it raises their price is by buying them. It is true, simply as a matter of logic, that the only way that QE can lower the market rate on a loan from, say, 4 percent to 3.9 percent, is by buying up enough loans (or rather, assets that are substitutes for loans) that the marginal lender now values a 3.9 percent loan the same as the marginal lender valued a 4 percent loan before. If a lender who previously would have considered a loan at 4 percent just worth making, does not now consider a loan at 3.9 percent worth making, then the interest rate on loans will not fall. Despite what John Taylor imagines, the Fed does not reduce interest rates by imposing a ceiling by fiat. So the statement, “if the Fed lowers long rates, bank won’t want to lend” is incoherent: the only way the Fed can lower long rates is by making banks want to lend more.

Stein’s argument is, to be honest, a bit puzzling. If it were true that banks respond to lower rates not by reducing lending or accepting lower profit margins, but by redoubling their efforts to fraudulently inflate returns, that would seem to be an argument for radically reforming the bank industry, or at least sending a bunch of bankers to jail. Stein, weirdly, wants it to be an argument for keeping rates perpetually high. But we don’t even need to have that conversation. Because what matters to banks is not the absolute level of rates, but the spread between their borrowing rate and their lending rate. And in the current institutional setting, expansionary policy implies higher spreads. Nobody needs to be reaching for yield.

[1] The DeLong post doesn’t give a link, but I think he’s responding to this February 7 speech.

[2] As Daniel Davies puts it in comments to the DeLong post:

If the Federal Reserve sets out on a policy of lowering interest rates in order to encourage banks to make loans to the real economy, it is a bit weird for someone’s main critique of the policy to be that it is encouraging banks to make loans. If Jeremy Stein worked for McDonalds, he would be warning that their latest ad campaign carried a risk that it might increase sales of delicious hamburgers.

[3] Specifically, these are commercial banks’ total interest payments from loans and leases divided by the total stock of loans and leases, and total interest payments divided by total liabilities and interest-bearing liabilities respectively.

[4] Why yes, I have been hanging around with a toddler lately.

[5] Interesting historical aside: Keynes’ conclusion in the 1930s that central bank intereventions could not restore full employment and that fiscal policy was therefore necessary, was not — pace the postwar Keynesian mainstream — based on any skepticism about the responsiveness of economic activity to interest rates in principle. It was, rather, based on his long-standing doubts about the reliability of the link from short rates to long rates, plus a new conviction that central banks would be politically unable or unwilling to target long rates directly.

Does the Fed Control Interest Rates?

Casey Mulligan goes to the New York Times to say that monetary policy doesn’t work. This annoys Brad DeLong:

THE NEW YORK TIMES PUBLISHES CASEY MULLIGAN AS A JOKE, DOESN’T IT?

… The third joke is the entire third paragraph: since the long government bond rate is made up of the sum of (a) an average of present and future short-term rates and (b) term and risk premia, if Federal Reserve policy affects short rates then–unless you want to throw every single vestige of efficient markets overboard and argue that there are huge profit opportunities left on the table by financiers in the bond market–Federal Reserve policy affects long rates as well.

Casey B. Mulligan: Who Cares About Fed Funds?: New research confirms that the Federal Reserve’s monetary policy has little effect on a number of financial markets, let alone the wider economy…. Eugene Fama of the University of Chicago recently studied the relationship between the markets for overnight loans and the markets for long-term bonds…. Professor Fama found the yields on long-term government bonds to be largely immune from Fed policy changes…

Krugman piles on [1]; the only problem with DeLong’s post, he says, is that

it fails to convey the sheer numbskull quality of Mulligan’s argument. Mulligan tries to refute people like, well, me, who say that the zero lower bound makes the case for fiscal policy. … Mulligan’s answer is that this is foolish, because monetary policy is never effective. Huh?

… we have overwhelming empirical evidence that monetary policy does in fact “work”; but Mulligan apparently doesn’t know anything about that.

Overwhelming evidence? Citation needed, as the Wikipedians say.

Anyway, I don’t want to defend Mulligan — I haven’t even read the column in question — but on this point, he’s got a point. Not only that: He’s got the more authentic Keynesian position.

Textbook macro models, including the IS-LM that Krugman is so fond of, feature a single interest rate, set by the Federal Reserve. The actual existence of many different interest rates in real economies is hand-waved away with “risk premia” — market rates are just equal to “the” interest rate plus a prmium for the expected probability of default of that particular borrower. Since the risk premia depnd on real factors, they should be reasonably stable, or at least independent of monetary policy. So when the Fed Funds rate goes up or down, the whole rate structure should go up and down with it. In which case, speaking of “the” interest rate as set by the central bank is a reasonable short hand.

How’s that hold up in practice? Let’s see:

The figure above shows the Federal Funds rate and various market rates over the past 25 years. Notice how every time the Fed changes its policy rate (the heavy black line) the market rates move right along with it?

Yeah, not so much.

In the two years after June 2007, the Fed lowered its rate by a full five points. In this same period, the rate on Aaa bonds fell by less 0.2 points, and rates for Baa and state and local bonds actually rose. In a naive look at the evidence, the “overwhelming” evidence for the effectiveness of monetary policy is not immediately obvious.

Ah but it’s not current short rates that long rates are supposed to follow, but expected short rates. This is what our orthodox New Keynesians would say. My first response is, So what? Bringing expectations in might solve the theoretical problem but it doesn’t help with the practical one. “Monetary policy doesn’t work because it doesn’t change expectations” is just a particular case of “monetary policy doesn’t work.”

But it’s not at all obvious that long rates follow expected short rates either. Here’s another figure. This one shows the spreads between the 10-Year Treasury and the Baa corporate bond rates, respectively, and the (geometric) average Fed Funds rate over the following 10 years.

If DeLong were right that “the long government bond rate is made up of the sum of (a) an average of present and future short-term rates and (b) term and risk premia” then the blue bars should be roughly constant at zero, or slightly above it. [2] Not what we see at all. It certainly looks as though the markets have been systematically overestimating the future level of the Federal Funds rate for decades now. But hey, who are you going to believe, the efficient markets theory or your lying eyes? Efficient markets plus rational expectations say that long rates must be governed by the future course of short rates, just as stock prices must be governed by future flows of dividends. Both claims must be true in theory, which means they are true, no matter how stubbornly they insist on looking false.

Of course if you want to believe that the inherent risk premium on long bonds is four points higher today than it was in the 1950s, 60s and 70s (despite the fact that the default rate on Treasuries, now as then, is zero) and that the risk premium just happens to rise whenever the short rate falls, well, there’s nothing I can do to stop you.

But what’s the alternative? Am I really saying that players in the bond market are leaving huge profit opportunities on the table? Well, sometimes, maybe. But there’s a better story, the one I was telling the other day.

DeLong says that if rates are set by rational, profit-maximizing agents, then — setting aside default risk — long rates should be equal to the average of short rates over their term. This is a standard view, everyone learns it. but it’s not strictly correct. What profit-maximizing bond traders do, is set long rates equal to the expected future value of long rates.

I went through this in that other post, but let’s do it again. Take a long bond — we’ll call it a perpetuity to keep the math simple, but the basic argument applies to any reasonably long bond. Say it has a coupon (annual payment) of $40 per year. If that bond is currently trading at $1000, that implies an interest rate of 4 percent. Meanwhile, suppose the current short rate is 2 percent, and you expect that short rate to be maintained indefinitely. Then the long bond is a good deal — you’ll want to buy it. And as you and people like you buy long bonds, their price will rise. It will keep rising until it reaches $2000, at which point the long interest rate is 2 percent, meaning that the expected return on holding the long bond and rolling over short bonds is identical, so there’s no incentive to trade one for the other. This is the arbitrage that is supposed to keep long rates equal to the expected future value of short rates. If bond traders don’t behave this way, they are missing out on profitable trades, right?

Not necessarily. Suppose the situation is as described above — 4 percent long rate, 2 percent short rate which you expect to continue indefinitely. So buying a long bond is a no-brainer, right? But suppose you also believe that the normal or usual long rate is 5 percent, and that it is likely to return to that level soon. Maybe you think other market participants have different expectations of short rates, maybe you think other market participants are irrational, maybe you think… something else, which we’ll come back to in a second. For whatever reason, you think that short rates will be 2 percent forever, but that long rates, currently 4 percent, might well rise back to 5 percent. If that happens, the long bond currently trading for $1000 will fall in price to $800. (Remember, the coupon is fixed at $40, and 5% = 40/800.) You definitely don’t want to be holding a long bond when that happens. That would be a capital loss of 20 percent. Of course every year that you hold short bonds rather than buying the long bond at its current price of $1000, you’re missing out on $20 of interest; but if you think there’s even a moderate chance of the long bond falling in value by $200, giving up $20 of interest to avoid that risk might not look like a bad deal.

Of course, even if you think the long bond is likely to fall in value to $800, that doesn’t mean you won’t buy it for anything above that. if the current price is only a bit above $800 (the current interest rate is only a bit below the “normal” level of 5 percent) you might think the extra interest you get from buying a long bond is enough to compensate you for the modest risk of a capital loss. So in this situation, the equilibrium price of the long bond won’t be at the normal level, but slightly below it. And if the situation continues long enough, people will presumably adjust their views of the “normal” level of the long bond to this equilibrium, allowing the new equilibrium to fall further. In this way, if short rates are kept far enough from long rates for long enough, long rates will eventually follow. We are seeing a bit of this process now. But adjusting expectations in this way is too slow to be practical for countercyclical policy. Starting in 1998, the Fed reduced rates by 4.5 points, and maintained them at this low level for a full six years. Yet this was only enough to reduce Aaa bond rates (which shouldn’t include any substantial default risk premium) by slightly over one point.

In my previous post, I pointed out that for policy to affect long rates, it must include (or be believed to include) a substantial permanent component, so stabilizing the economy this way will involve a secular drift in interest rates — upward in an economy facing inflation, downward in one facing unemployment. (As Steve Randy Waldman recently noted, Michal Kalecki pointed this out long ago.) That’s important, but I want to make another point here.

If the primary influence on current long rates is the expected future value of long rates, then there is no sense in which long rates are set by fundamentals. There are a potentially infinite number of self-fulfilling expected levels for long rates. And again, no one needs to behave irrationally for these conventions to sustain themselves. The more firmly anchored is the expected level of long rates, the more rational it is for individual market participants to act so as to maintain that level. That’s the “other thing” I suggested above. If people believe that long rates can’t fall below a certain level, then they have an incentive to trade bonds in a way that will in fact prevent rates from falling much below that level. Which means they are right to believe it. Just like driving on the right or left side of the street, if everyone else is doing it it is rational for you to do it as well, which ensures that everyone will keep doing it, even if it’s not the best response to the “fundamentals” in a particular context.

Needless to say, the idea that that long-term rate of interest is basically a convention straight from Keynes. As he puts it in Chapter 15 of The General Theory,

The rate of interest is a highly conventional … phenomenon. For its actual value is largely governed by the prevailing view as to what its value is expected to be. Any level of interest which is accepted with sufficient conviction as likely to be durable will be durable; subject, of course, in a changing society to fluctuations for all kinds of reasons round the expected normal.

You don’t have to take Keynes as gospel, of course. But if you’ve gotten as much mileage as Krugman has out of the particular extract of Keynes’ ideas embodied in the IS-LM mode, wouldn’t it make sense to at least wonder why the man thought this about interest rates, and if there might not be something to it.

Here’s one more piece of data. This table shows the average spread between various market rates and the Fed Funds rate.

Spreads over Fed Funds by decade
	10-Year Treasuries	Aaa Corporate Bonds	Baa Corporate Bonds	State & Local Bonds
1940s		2.2	3.3
1950s	1.0	1.3	2.0	0.7
1960s	0.5	0.8	1.5	-0.4
1970s	0.4	1.1	2.2	-1.1
1980s	0.6	1.4	2.9	-0.9
1990s	1.5	2.6	3.3	0.9
2000s	1.5	3.0	4.1	1.8

Treasuries carry no default risk; a given bond rating should imply a fixed level of default risk, with the default risk on Aaa bonds being practically negligible. [3] Yet the 10-year treasury spread has increased by a full point and the corporate bond rates by about two points, compared with the postwar era. (Municipal rates have risen by even more, but there may be an element of genuine increased risk there.) Brad DeLong might argue that society’s risk-bearing capacity has decline so catastrophically since the 1960s that even the tiny quantum of risk in Aaa bonds requires two full additional points of interest to compensate its quaking, terrified bearers. And that this has somehow happened without requiring any more compensation for the extra risk in Baa bonds relative to Aaa. I don’t think even DeLong would argue this, but when the honor of efficient markets is at stake, people have been known to do strange things.

Wouldn’t it be simpler to allow that maybe long rates are not, after all, set as “the sum of (a) an average of present and future short-term rates and (b) [relatively stable] term and risk premia,” but that they follow their own independent course, set by conventional beliefs that the central bank can only shift slowly, unreliably and against considerable resistance? That’s what Keynes thought. It’s what Alan Greenspan thinks. [4] And also it’s what seems to be true, so there’s that.

[1] Prof. T. asks what I’m working on. A blogpost, I say. “Let me guess — it says that Paul Krugman is great but he’s wrong about this one thing.” Um, as a matter of fact…

[2] There’s no risk premium on Treasuries, and it is not theoretically obvious why term premia should be positive on average, though in practice they generally are.

[3] Despite all the — highly deserved! — criticism the agencies got for their credulous ratings of mortgage-backed securities, they do seem to be good at assessing corporate default risk. The cumulative ten-year default rate for Baa bonds issued in the 1970s was 3.9 percent. Two decades later, the cumulative ten-year default rate for Baa bonds issued in the 1990s was … 3.9 percent. (From here, Exhibit 42.)

[4] Greenspan thinks that the economically important long rates “had clearly delinked from the fed funds rate in the early part of this decade.” I would only add that this was just the endpoint of a longer trend.

Interest Rates and Expectations: Responses and Further Thoughts

Some good questions asked in comments to yesterday’s post.

Random Lurker doubts whether there is a strict inverse relationship between interest rates and bond values. Indeed there is not, apart from perpetuities (bonds with an infinite maturity, where the principle is never repaid.) I should have been clearer in the post, I was talking about perpetuities just as a simplification of the general case of long assets. But I would argue it’s a reasonable simplification. If you think that the importance of interest rates is primarily for the valuation (rather than the financing) of capital goods, and you think that capital goods are effectively infinitely lived, then an analysisis in terms of perpetuities is the strcitly correct way to think about it. (Both assumptions are defensible, as a first approximation, and Keynes seems to have held both.) On the other hand, if you are thinking in terms of financing conditions for long but not infinitely lived assets, the perpetuity is only an approximation, but for long maturities it’s a reasonably close one. For example, a 30 year bond loses 14% of its value when interest rates rise from 5% to 6%, compared with a 20% loss for a perpetuity. Qualitatively the story will hold as long as the interest rates that matter are much longer than the timescale of business cycles.

Max is confused about my use of “bull” and “bear.” Again, I should have been clearer: I am using the terms in the way that Keynes did, to refer to bullishness and bearishness about bond prices, not about the economy in general.

Finally, the shortest but most substantive comment, from Chris Mealy:

Forcing Bill Gross to lose billions in slow motion is a crazy way to get to full employment.

It is! And that is kind of the point.

I wrote this post mainly to clarify my own thinking, not to make any policy or political argument. But obviously the argument that comes out of this is that while monetary policy can help stabilize demand, it’s very weak at restoring demand once it’s fallen – and not just because short rates can’t go below zero, or because central banks are choosing the wrong target. (Although it is certainly true, and important, that central bankers are not really trying to reduce unemployment.)

Here is the thing: expectations of returns on investment are also conventional and moderately elastic. Stable full employment requires both that expected sales are equal to expenditure at full employment, and that interest rates are such that the full employment level of output is chosen by profit-maximizing businesses. But once demand has fallen – and especially if it has remained depressed for a while – expected sales fall, so the interest rate that would have been low enough to prevent the fall in activity is no longer low enough to reverse it. This is why you temporarily need lower rates than you will want when the economy recovers. But the expectation of long rates returning to their old level will prevent them from falling in the first place. “The power of the central bank to affect the long rate is limited by the opinions about its normal level inherited from the past.” This is why monetary policy cannot work in a situation like this without Bill Gross first losing billions – it’s the only way to change his opinion.

Leijonhufvud:

Suppose that a situation arises in which the State of Expectation happens to be “appropriate”… but that the long rate is higher than “optimal,” so that asset demand prices are too low for full employment… Then it seems quite reasonable to demand that the Central Bank should go to great lengths in trying to reduce the interest rate… If, however, the actual interest rate equals the “optimal” rate consistent with the suggested “neutral state,” while asset prices are too low due to a State of Expectation which is “inappropriately pessimistic”-what then?

Consider what would happen if, in this situation, the long bond rate were forced down to whatever level was necessary to equate ex ante rates of saving and investment at full employment. This would mean that prices of bonds-assets with contractually fixed long receipt streams-would shoot up while equity prices remained approximately constant instead of declining. Through a succession of short periods, with aggregate money expenditures at the full employment level, initial opinions about the future yield on capital would be revealed as too pessimistic. Anticipated returns to capital go up. The contractually fixed return streams on bonds remain the same, and it now becomes inevitable that bond-holders take a capital loss (in real terms).

The Central Bank now has two options. (a) It may elect to stand by [leaving rates at very low levels.] … Since the situation is one of full employment, inflation must result and the “real value” of nominally fixed contracts decline. (b) It may choose … to increase market rate sufficiently to prevent any rise in [inflation]. Bond-holders lose again, since this means a reduction in the money value of bonds.

In other words, in our world of long-lived assets, if you rely only on monetary policy to get you out of depression, Bill Gross has to lose money. On a theoretical level, the fact that the lifetime of capital goods is long relative to the period over which we can reliably treat “fundamentals” as fixed means that the Marshallian long run, in which the capital stock is fully adjusted, does not apply to any actual economy. (This fact has many important implications beyond the scope of these posts.)

The key point for our purposes is that, in the slump, investment demand is lower than it will be once the economy recovers. So if the interest rate falls enough to end the recession, then you must have either a rise in rates or inflation once the slump ends. But either of those will mean losses for bondholders, anticipation of which will prevent long rates from falling the first place. Only if you successfully fool bond market participants can monetary policy produce recovery on a timescale significantly less than average asset life. The alternative is to prove the pessimistic expectations of entrepreneurs wrong by directly raising incomes, but that seems to be off the table.

This point is obvious, but it’s strangely ignored, perhaps because discussion of monetary policy is almost entirely focused on how optimal policy can prevent slumps from occurring in the first place. The implicit assumption of Krugman’s ISLM analysis, for instance, is that investment demand has permanently fallen, presumably unrelatedly to demand conditions themselves. So the new low rate is permanently appropriate. But — I feel it’s it’s safe to say — Krugman, and certainly market participants, don’t really believe this. But if policy is going to be reversed, on a timescale significantly shorter than the duration of the assets demand for which is supposed to be affected by monetary policy, then policy will not work at all.

At this point, though, it would seem that we have proven too much. The question becomes not, why isn’t monetary policy working now, but, How did monetary policy ever work? I can think of at least four answers, all of which probably have some truth to them.

1. It didn’t. The apparent stability of economies with active central banks is due to other factors. Changes in the policy not been stabilizing, or have even been destabilizing. This is consistent with the strand of the Post Keynesian tradition that emphasizes the inflationary impact of rate increases, since short rates are a component of marginal costs; but it is also basically the view of Milton Friedman and his latter-day epigones in the Market Monetarist world. I’m sympathetic but don’t buy it; I think the evidence is overwhelming that high interest rates are associated with low income/output, and vice versa.

2. The focus on long-lived goods is a mistake. The real effect of short rates is not via long rates, but on stuff that is financed directly by short borrowing, particularly inventories and working capital. I’m less sure about this one, but Keynes certainly did not think it was important; for now let’s follow him. A variation is income distribution, including corporate cashflow. Bernanke believes this. I’m doubtful that it’s the main story, but I presume there is something in it; how much is ultimately an empirical question.

3. The answer suggested by the analysis here: Monetary policy works well when the required interest rate variation stays within the conventional “normal” range. In this range, there are enough bulls and bears for the marginal bond buyer to expect the current level of interst to continue indefinitely, so that bond prices are not subject to stabilizing speculation and there is no premium for expected capital losses or gains; so long rates should move more or less one for one with short rates. This works on a theoretical level, but it’s not obvious that it particularly fits the data.

4. The most interesting possibility, to me: When countercylclical monetary policy seemed effective, it really was, but on different principles. Autonomous demand and interest rates were normally at a level *above* full employment, and stabilization was carried out via direct controls on credit creation, such as reserve requirements. A variation on this is that monetary policy has only ever worked through the housing market.

Regardless of the historical issue, the most immediately interesting question is how and whether monetary policy can work now. And here, we can safely say that channels 2,3 and 4, even if real, are exhausted. So in the absence of fiscal policy, it really does come down to the capacity of sustained low short rates to bring expected long rates down. Sorry, Bill Gross!

UPDATE: I was just reading this rightly classic paper by Chari, Kehoe and McGrattan. They’re pure freshwater, everything I hate. But New Keynesians are just real business cycle theorists with a bad conscience, which means the RBCers pwn them every time in straight-up debate. As here.I’m not interested in that, though, though the paper is worth reading if you want the flavor of what “modern macro” is all about. Rather, I’m interested in this subsidiary point in their argument:

as is well-known, during the postwar period, short rates and long rates have a very similar secular pattern. … Second, a large body of work in finance has shown that the level of the long rate is well-accounted for by the expectations hypothesis. … Combining these two features of the data implies that when the Fed alters the current short rate, private agents significantly adjust their long-run expectations of the future short rate, say, 30 years into the future. At an intuitive level, then, we see that Fed policy has a large random walk component to it.

In what sense this is true, I won’t venture to guess. It seems, at least, problematic, given that they also think that “interest rates … should be kept low on average.” The important point for my purposes, tho, is just that even the ultra-orthodox agree, that for a change in monetary policy to be effective, it has to be believed to be permanent. “If that which is at all were not forever…”

Interest Rates and (In)elastic Expectations

[Apologies to any non-econ readers, this is even more obscure than usual.]

Brad DeLong observed last week that one of the most surprising things about the Great Recession is how far long-term interest rates have followed short rates toward zero.

I have gotten three significant pieces of the past four years wrong. Three things surprised and still surprise me: (1.) The failure of central banks to adopt a rule like nominal GDP targeting, or it’s equivalent. (2.) The failure of wage inflation in the North Atlantic to fall even farther than it has–toward, even if not to, zero. (3.) The failure of the yield curve to sharply steepen: federal funds rates at zero I expected, but 30-Year U.S. Treasury bond nominal rates at 2.7% I did not.

… The third… may be most interesting.

Back in March 2009, the University of Chicago’s Robert Lucas confidently predicted that within three years the U.S. economy would be back to normal. A normal U.S. economy has a short-term nominal interest rate of 4%. Since the 10-Year U.S. Treasury bond rate tends to be one percentage point more than the average of expected future short-term interest rates over the next decade, even five expected years of a deeply depressed economy with essentially zero short-term interest rates should not push the 10-Year Treasury rate below 3%. (And, indeed, the Treasury rate fluctuated around 3 to 3.5% for the most part from late 2008 through mid 2011.) But in July of 2011 the 10-Year U.S. Treasury bond rate crashed to 2%, and at the start of June it was below 1.5%. [

The possible conclusions are stark: either those investing in financial markets expect … [the] current global depressed economy to endure in more-or-less its current state for perhaps a decade, perhaps more; or … the ability of financial markets to do their job and sensibly price relative risks and returns at a rational level has been broken at a deep and severe level… Neither alternative is something I would have or did predict, or even imagine.

I also am surprised by this, and for similar reasons to DeLong. But I think the fact that it’s surprising has some important implications, which he does not draw out.

Here’s a picture:

The dotted black line is the Federal Funds rate, set, of course, by the central bank. The red line is the 10-year Treasury; it’s the dip at the far right in that one that surprises DeLong (and me). The green line is the 30-year Treasury, which behaves similarly but has fallen by less. Finally, the blue line is the BAA bond rate, a reasonable proxy for the interest rate faced by large business borrowers; the 2008 financial crisis is clearly visible. (All rates are nominal.) While the Treasury rates are most relevant for the expectations story, it’s the interest rates faced by private borrowers that matter for policy.

The recent fall in 10-year treasuries is striking. But it’s at least as striking how slowly and incompletely they, and corporate bonds, respond to changes in Fed policy, especially recently. It’s hard to look at this picture and not feel a twinge of doubt about the extent to which the Fed “sets” “the” interest rate in any economically meaningful sense. As I’ve mentioned here before, when Keynes referred to the “liquidity trap,” he didn’t mean the technical zero lower bound to policy rates, but its delinking from the economically-important long rates. Clearly, it makes no difference whether or not you can set a policy rate below zero if there’s reason to think that longer rates wouldn’t follow it down in any case. And I think there is reason to think that.

The snapping of the link between monetary policy and other rates was written about years ago by Benjamin Friedman, as a potential; it figured in my comrade Hasan Comert’s dissertation more recently, as an actuality. Both of them attribute the disconnect to institutional and regulatory changes in the financial system. And I agree, that’s very important. But after reading Leijonhufvud’s On Keynesian Economics and the Economics of Keynes [1], I think there may be a deeper structural explanation.

As DeLong says, in general we think that long interest rates should be equal to the average expected short rates over their term, perhaps plus a premium. [2] So what can we say about interest rate expectations? One obvious question is, are they elastic or inelastic? Elastic expectations change easily; in particular, unit-elastic expectations mean that whatever the current short rate is, it’s expected to continue indefinitely. Inelastic expectations change less easily; in the extreme case of perfectly inelastic interest rate expectations, your prediction for short-term interest rates several years from now is completely independent of what they are now.

Inelastic interest-rate expectations are central to Keynes’ vision of the economy. (Far more so than, for instance, sticky wages.) They are what limit the effectiveness of monetary policy in a depression or recession, with the liquidity trap simply the extreme case of the general phenomenon. [3] His own exposition is a little hard to follow, but the simplest way to look at it is to recall that when interest rates fall, bond prices rise, and vice versa. (In fact they are just two ways of describing the same thing.) So if you expect a rise in interest rates in the future that means you’ll expect a capital loss if you hold long-duration bonds, and if you expect a fall in interest rates you’ll expect a capital gain. So the more likely it seems that short-term interest rates will revert to some normal level in the future, the less long rates should follow short ones.

This effect gets stronger as we consider longer maturities. In the limiting case of a perpetuity — a bond that makes a fixed dollar period every period forever — the value of the bond is just p/i, where p is the payment in each period and i is the interest rate. So when you consider buying a bond, you have to consider not just the current yield, but the possibility that interest rates will change in the future. Because if they do, the value of the bonds you own will rise or fall, and you will experience a capital gain or loss. Of course future interest rates are never really known. But Keynes argued that there is almost always a strong convention about the normal or “safe” level of interest.

Note that the logic above means that the relationship between short and long rates will be different when rates are relatively high vs. when they are relatively low. The lower are rates, the greater the capital loss from an increase in rates. As long rates approach zero, the potential capital loss from an increase approaches infinity.

Let’s make this concrete. If we write i_s for the short interest rate and i_l for the long interest rate, B for the current price of long bonds, and BE for the expected price of long bonds a year from now, then for all assets to be willing held it must be the case that i_l = i_s – (BE/B – 1), that is, interest on the long bond will need to be just enough higher (or lower) than the short rate to cancel out the capital loss (or gain) expected from holding the long bond. If bondholders expect the long run value of bond prices to be the same as the current value, then long and short rates should be the same. [*] Now for simplicity let’s assume we are talking about perpetuities (the behavior of long but finite bonds will be qualitatively similar), so B is just 1/i_l. [4] Then we can ask the question, how much do short rates have to fall to produce a one point fall in long rates.

Obviously, the answer will depend on expectations. The standard economist’s approach to expectations is to say they are true predictions of the future state of the world, an approach with some obvious disadvantages for those of us without functioning time machines. A simpler, and more empirically relevant, way of framing the question, is to ask how expectations change based on changes in the current state of the world — which unlike the future, we can observe. Perfectly inelastic expectations mean that your best guess about interest rates at some future date is not affected at all by the current level of interest rates; unit-elastic expectations mean that your best guess changes one for one with the current level. An of course there are all the possibilities in between. Let’s quantify this as the subjective annual probability that a departure of interest rates from their current or “normal” level will subsequently be reversed. Now we can calculate the exact answer to the question posed above, as shown in the next figure.

For instance, suppose short rates are initially at 6 percent, and suppose this is considered the “normal” level, in the sense that the marginal participant in the bond market regards an increase or decrease as equally likely. Then the long rate will also be 6 percent. Now we want to get the long rate down to 5 percent. Suppose interest rate expectations are a bit less than unit elastic — i.e. when market rates change, people adjust their views of normal rates by almost but not quite as much. Concretely, say that the balance of expectations is that there is net 5 percent annual chance that rates will return to their old normal level. If the long rate does rise back to 6 percent, people who bought bonds at 5 percent will suffer a capital loss of 20 percent. A 5 percent chance of a 20 percent loss equals an expected annual loss of 1 percent, so long rates will need to be one point higher than short rates for people to hold them. [5] So from a starting point of equality, for long rates to fall by one point, short rates must fall by two points. You can see that on the blue line on the graph. You can also see that if expectations are more than a little inelastic, the change in short rates required for a one-point change in long rates is impossibly large unless rates are initially very high.

It’s easy enough to do these calculations; the point is that unless expectations are perfectly elastic, we should always expect long rates to change less than one for one with short rates; the longer the rates considered, the more inelastic expectations, and the lower initial rates, the less responsive long rates will be. At the longest end of the term structure — the limiting case of a perpetuity — it is literally impossible for interest rates to reach zero, since that would imply an infinite price.

This dynamic is what Keynes was talking about when he wrote:

If . . . the rate of interest is already as low as 2 percent, the running yield will only offset a rise in it of as little as 0.04 percent per annum. This, indeed, is perhaps the chief obstacle to a fall in the rate of interest to a very low level . . . [A] long-term rate of interest of (say) 2 percent leaves more to fear than to hope, and offers, at the same time, a running yield which is only sufficient to offset a very small measure of fear.

Respectable economists like DeLong believe that there is a true future path of interest rates out there, which current rates should reflect; either the best current-information prediction is of government policy so bad that the optimal interest rate will continue to be zero for many years to come, or else financial markets have completely broken down. I’m glad the second possibility is acknowledged, but there is a third option: There is no true future course of “natural” rates out there, so markets adopt a convention for normal interest rates based on past experience. Given the need to take forward-looking actions without true knowledge of the future, this is perfectly rational in the plain-English sense, if not in the economist’s.

A final point: For Keynes — a point made more clearly in the Treatise than in the General Theory — the effectivness of monetary policy depends critically on the fact that there are normally market participants with differing expectations about future interest rates. What this means is that when interest rates rise, people who think the normal or long-run rate of interest is relatively low (“bulls”) can sell bonds to people who think the normal rate is high (“bears”), and similarly when interest rates fall the bears can sell to the bulls. Thus the marginal bond will be held held by someone who thinks the current rate of interest is the normal one, and so does not require a premium for expected capital gains or losses. This is the same as saying that the market as a whole behaves as if expectations are unit-elastic, even though this is not the case for individual participants. [6] But when interest rates move too far, there will no longer be enough people who think the new rate is normal to willingly hold the stock of bonds without an interest-rate risk premium. In other words, you run out of bulls or bears. Keynes was particularly concerned that an excess of bear speculators relative to bulls could keep long interest rates permanently above the level compatible with full employment. The long rate, he warned,

may fluctuate for decades about a level which is chronically too high for full employment; – particularly if it is the prevailing opinion that the rate of interest is self-adjusting, so that the level established by convention is thought to be rooted in objective grounds much stronger than convention, the failure of employment to attain an optimum level being in no way associated, in the minds either of the public or of authority, with the prevalence of an inappropriate range of rates of interest’.

If the belief that interest rates cannot fall below a certain level is sufficiently widespread, it becomes self-fulfilling. If people believe that long-term interest rates can never persistently fall below, say, 3 percent, then anyone who buys long bonds much below that is likely to lose money. And, as Keynes says, this kind of self-stabilizing convention is more likely to the extent that people believe that it’s not just a convention, but that there is some “natural rate of interest” fixed by non-monetary fundamentals.

So what does all this mean concretely?

1. It’s easy to see inelastic interest-rate expectations in the data. Long rates consistently lag behind short rates. During the 1960s and 1970s, when rates were secularly rising, long rates were often well below the Federal Funds rate, especially during tightening episodes; during the period of secularly falling rates since 1980, this has almost never happened, but very large term spreads have become more common, especially during loosening episodes.

2. For the central bank to move long rates, it must persuade markets that changes in policy are permanent, or at least very persistent; this is especially true when rates are low. (This is the main point of this post.) The central bank can change rates on 30-year bonds, say, only by persuading markets that average rates over the next 30 years will be different than previously believed. Over small ranges, the existence of varying beliefs in the bond market makes this not too difficult (since the central bank doesn’t actually have to change any individual’s expectations if bond sales mean the marginal bondholder is now a bull rather than a bear, or vice versa) but for larger changes it is more difficult. And it becomes extremely difficult to the extent that economic theory has taught people that there is a long run “natural” rate of interest that depends only on technology and time preferences, which monetary policy cannot affect.

Now, the obvious question is, how sure are we that long rates are what matters? I’ve been treating a perpetual bond as an approximation of the ultimate target of monetary policy, but is that reasonable? Well, one point on which Keynes and today’s mainstream agree is that the effect of interest rates on the economy comes through demand for long-lived assets — capital goods and housing. [7] According to the BEA, the average current-cost age of private fixed assets in the US is a bit over 21 years, which implies that the expected lifetime of a new fixed asset must be quite a bit more than that. For Keynes (Leijonhufvud stresses this point; it’s not so obvious in the original texts) the main effect of interest rates is not on the financing conditions for new fixed assets, as most mainstream and heterodox writers both assume, but on the discount rate used of the assets. In that case the maturity of assets is what matters. On the more common view, it’s the maturity of the debt used to finance them, which may be a bit less; but the maturity of debt is usually matched to the maturity of assets, so the conclusion is roughly the same. The relevant time horizon for fixed assets is long enough that perpetuities are a reasonable first approximation. [8]

3. So if long rates are finally falling now, it’s only because an environment of low rates is being established as new normal. There’s a great deal of resistance to this, since if interest rates do return to their old normal levels, the capital losses to bondholders will be enormous. So to get long rates down, the Fed has to overcome intense resistance from bear speculators. Only after a great deal of money has been lost betting on a return of interest rates to old levels will market participants begin to accept that ultra-low rates are the new normal. The recent experience of Bill Gross of PIMCO (the country’s largest bond fund) is a perfect example of this story. In late 2010, he declared that interest rates could absolutely fall no further; it was the end of the 30-year bull market in bonds. A year later, he put his money where his mouth was and sold all his holdings of Treasuries. As it turned out, this was just before bond prices rose by 30 percent (the flipside of the fall in rates), a misjudgment that cost his investors billions. But Gross and the other “bears” had to suffer those kinds of losses for the recent fall in long rates to be possible. (It is also significant that they have not only resisted in the market, but politically as well.) The point is, outside a narrow range, changes in monetary policy are only effective when they cease to be perceived as just countercyclical, but as carrying information about “the new normal.” Zero only matters if it’s permanent zero.

4. An implication of this is that in a world where the lifespan of assets is much longer than the scale of business-cycle fluctuations, we cannot expect interest rates to be stationary if monetary policy is the main stabilization tool. Unless expectations are very elastic, effective monetary policy require secular drift in interest rates, since each short-term stabilization episode will result in a permanent change in interest rates. [9] You can see this historically: the fall in long rates in the 1990 and 2000 loosenings both look about equal to the permanent components of those changes. This is a problem for two reasons: First, because it means that monetary policy must be persistent enough to convince speculators that it does represent a permanent change, which means that it will act slower, and require larger changes in short rates (with the distortions those entail) than in the unit-elastic expectations case. And second, because if there is some reason to prefer one long-ru level of interest rates to another (either because you believe in a “natural” rate, or because of the effects on income distribution, asset price stability, etc.) it would seem that maintaining that rate is incompatible with the use of monetary policy for short-run stabilization. And of course the problem is worse, the lower interest rates are.

5. One way of reading this is that monetary policy works better when interest rates are relatively high, implying that if we want to stabilize the economy with the policy tools we have, we should avoid persistently low interest rates. Perhaps surprisingly, given what I’ve written elsewhere, I think there is some truth to this. If “we” are social-welfare-maximizing managers of a capitalist economy, and we are reliant on monetary policy for short-run stabilization, then we should want full employment to occur in the vicinity of nominal rates around 10 percent, versus five percent. (One intuitive way of seeing this: Higher interest rates are equivalent to attaching a low value to events in the future, while low interest rates are equivalent to a high value on those events. Given the fundamental uncertainty about the far future, choices in the present will be more stable if they don’t depend much on far-off outcomes.) In particular — I think it is a special case of the logic I’ve been outlining here, though one would have to think it through — very low interest rates are likely to be associated with asset bubbles. But the conclusion, then, is not to accept a depressed real economy as the price of stable interest rates and asset prices, but rather to “tune” aggregate demand to a higher level of nominal interest rates. One way to do this, of course, is higher inflation; the other is a higher level of autonomous demand, either for business investment (the actual difference between the pre-1980 period and today, I think), or government spending.

[1] The most invigorating economics book I’ve read in years. It’ll be the subject of many posts here in the future, probably.

[2] Why there should be a pure term premium is seldom discussed but actually not straightforward. It’s usually explained in terms of liquidity preference of lenders, but this invites the questions of (1) why liquidity preference outweighs “solidity preference”; and (2) why lenders’ preferences should outweigh borrowers’. Leijonhufvud’s answer, closely related to the argument of this post, is that the “excessively long” lifespan of physical capital creates chronic excess supply at the long end of the asset market. In any case, for the purpose of this post, we will ignore the pure premium and assume that long rates are simply the average of expected short rates.

[3] Keynes did not, as is sometimes suggested by MMTers and other left Keynesians, reject the effectiveness of monetary policy in general. But he did believe that it was much more effective at stabilizing full employment than at restoring full employment from a depressed state.

[4] I will do up these equations properly once the post is done.

[5] I anticipate an objection to reasoning on the basis of an equilibrium condition in asset markets. I could just say, Keynes does it. But I do think it’s legitimate, despite my rejection of the equilibrium methodology more generally. I don’t think there’s any sense that human behavior can be described as maximizing some quantity called utility,” not even as a rough approximation; but I do think that capitalist enterprises can be usefully described as maximizing profit. I don’t think that expectations in financial markets are “rational” in the usual economists’ sense, but I do think that one should be able to describe asset prices in terms of some set of expectations.

[6] We were talking a little while ago with Roger Farmer, Rajiv Sethi, and others about the desirability of limiting economic analysis to equilibria, i.e. states where all expectations are fulfilled. This implies, among other things, that all expectations must be identical. Keynes’ argument for why long rates are more responsive to short rates within some “normal” range of variation is — whether you think it’s right or not — an example of something you just can’t say within Farmer’s preferred framework.

[7] Despite this consensus, this may not be entirely the case; and in fact to the extent that monetary policy is effective in the real world, other channels, like income distribution, may be important. But let’s assume for now that demand for long-lived assets is what matters.

[8] Hicks had an interesting take on this, according to Leijonhufvud. Since the production process is an integrated whole, “capital” does not consist of particular goods but of a claim on the output of the process as a whole. Since this process can be expected to continue indefinitely, capital should be generally assumed to be infinitely-lived. When you consider how much of business investment is motivated by maintaining the firm’s competitive position — market share, up to date technology, etc. — it does seem reasonable to see investment as buying not a particular capital good but more of the firm as a whole.

[9] There’s an obvious parallel with the permanent inflation-temporary employment tradeoff of mainstream theory. Except, I think mine is correct!