Keynes's theory of potentially destabilizing bond markets, part II: uncertainty and interest rate instability

Keynes's main goal in writing chapters 13 and 15 was to create a theory founded on the assumption of fundamental uncertainty about future eco­nomic states that, unlike classical theory, could explain why long-term interest rates in the USA declined in the late 1920s but then leapt up in the early 1930s, exacerbating the economic and financial crises of the period.

To accomplish this goal, he focused on the choice investors have to make between holding their financial wealth in the form of risky long-term bonds that normally have a substantial interest rate but are vulnerable to capital loss when interest rates rise or holding it instead in the safe "liquid" form of "money" that has zero-interest yield but whose value cannot suffer a capital loss.^6,7 When investors begin to fear that interest rates will rise and bond prices will fall, they may sell bonds and hold on to the money they get from the sale. This will cause interest rates to rise, reinforcing the pessimistic expectations that induced the bond sales. If pessimism about future bond prices becomes strong enough in a period in which financial markets have become financially fragile, a panic may ensue.

In The General Theory, Keynes observed that in the USA in 1932, "there was a financial crisis or crises of liquidation, when scarcely anyone could be induced to part with holdings of money on any reasonable terms" (CW 7, pp. 207-208).

Disequilibrium processes in bond markets 227 that Keynes tried to explain in chapters 13 and 15 through his innovative theory of liquidity preference as a behavior toward uncertainty in bond markets.

Keynes began the second prong of his attack on the classical theory of interest rate determination and his presentation of his own theory of the bond market with an attack on the Quantity Theory of the demand to hold "money" because it does not incorporate what has become known in economics as the "speculative" motive for so doing - to hold money rather than bonds in order to avoid an expected future capital loss on bonds and to be in a position to make capital gains by buying bonds whenever expectations of rising bond prices return. Once expectations of future bond prices are incorporated into the demand for bonds function and, with Keynes, we assume that the future is unknowable in the present, out-of-equilibrium processes in the bond market can become destructive or destabilizing. In particular, a negative "shock" to AD may cause interest rates to rise, accelerating the rate of decline of income and employment.

Keynes argued that households and businesses will not automatically restrict themselves to holding the minimum amount of money required for transactions and invest the rest of their savings in long-term bonds as classical theory assumes. After the decision about what percentage of income to save has been made by a wealth holder:

there is a further decision which awaits him, namely in what form will he hold the command over future consumption which he has reserved, whether out of his current income or from previous savings [or wealth]. Does he want to hold it in the form of immediate, liquid command (i.e. in money or its equivalent)? Or is he prepared to part with immediate command for a specified or indefinite period, leaving it to [unknow­able] future market conditions to determine on what terms he can, if necessary, convert deferred command over specific goods into imme­diate command over goods in general? In other words, what is his liquidity-preference?... [The] definition of the rate of interest is the reward for parting with liquidity ["money"] for a specified period. For the rate of interest is, in itself, nothing more than the inverse propor­tion between a sum of money and what can be obtained for parting with control over the money in exchange for a debt for a stated period of time.

(CW 7, pp. 166-167, emphasis in original)

Once we acknowledge with Keynes that the future is uncertain or unknowable and therefore that there is, at times, a serious possibility of substantial capital loss in the holding of long-term securities like

bonds, the classical theory of the determination of the long-term interest rate becomes fatally flawed. Saving is no longer identically equal to the demand for bonds. There is a separate decision about how much saving should be used to purchase bonds and how much to hold in the liquid form of "money." Much more importantly, the agent has to ask: what per­centage of his or her accumulated wealth or "previous savings" does he or she want to hold in the risky long-term asset bonds rather than in the financial asset "money" that is immune from nominal capital loss? What is his or her "liquidity preference?"

Keynes's theory of the determination of the interest rate is a stock­equilibrium theory.⁸ He refers to "the amounts of his resources" that an investor "will wish to retain in the form of money" (CW 7, p.

The classical theory of interest rate determination assumes that investors know the correct value of the future interest rate and bond price as a point estimate or as a known and stationary probability distribution. There is no Keynesian uncertainty. But if the future interest rate and bond price are known with certainty, the demand to hold bonds would equal total financial wealth minus the amount of money needed for transactions and precautionary purposes. The demand to hold money as an investment asset would be zero in equilibrium at every positive rate of interest, no matter how low, because the possibility of capital loss has been ruled out by assumption.⁹ The demand to hold money would thus be a function of Y alone, with zero elasticity with respect to changes in the rate of interest. Keynes's argument here applies as well to the LM model; if there is no uncertainty about future interest rates, the LM curve would be a vertical line and all money would be held for transactions purposes.¹⁰

In a static society or in a society in which for any reason no one feels any uncertainty about the future rates of interest, the liquidity [or demand to hold money] function L, or the propensity to hoard [money] (as we might term it), will always be zero in equilibrium.

(CW 7, pp. 208-209, emphasis added)

It is impossible to overstate the importance of the assumption of funda­mental uncertainty to Keynes's attack on classical bond market theory and

Disequilibrium processes in bond markets 229 to the construction of his own theory of potentially unstable bond markets centered on the concept of liquidity preference. Keynes put this point as follows in his 1937 QJE article:

Partly on reasonable and partly on instinctive grounds, our desire to hold money as a store of wealth is a barometer of the degree of our dis­trust [or lack of confidence] of our own calculations [expectations] and conventions concerning the future. Even though this feeling about money is itself conventional or instinctive, it operates, so to speak, at a deeper level of our motivations. It takes charge at the moments when the higher, more precarious conventions have weakened. The possession of actual money [rather than expectations of money generated by interest payments or bond sales in the future] lulls our disquietude; and the premium which we require to make us part with money [the long-term interest rate] is the measure or our disquietude. (Keynes 1937, p. 116)

Or, as he put it in 1945 in a statement that assumes that investors can choose between risky long-term bonds and relatively safe short-term debt to include in their wealth holdings:

What determines the reward the individual requires to surrender his liquidity for a long or short period? In practice, of course, what some stockbroker who knows nothing about it advises him, or convention based on old dead ideas of past irrelevant experience. But assuming enlightened self-interest (which probably influences convention) it is your expectation of or a lack of expectation and temporary uncer­tainty about future changes in the r. of i. [rate of interest].

(CW 27, p. 391)

Keynes said that "the existence of uncertainty as to the future of the rate of interest" is "a necessary condition [for] the existence of a liquidity preference for money as a means of holding wealth" (p.

Keynes argued that, in a world of uncertainty, the bond-money choice will depend on two things. One is the investor's expectations of the movement of bond prices in the near to intermediate future. If bond prices are expected to either increase, remain the same, or at least fall by less than the interest payment on the bond, bonds will, on this criterion alone, be a better investment than zero-interest money. If they are expected to experi­ence a significant decline or a capital loss larger than interest payment, investors will sell bonds.

But there is a second crucial determinant of the desired bond-money split: the "confidence" investors place in the validity or truth content of their expectation-formation process - "on how likely we rate the likeli­hood of our best forecast turning out quite wrong" (CW 7, p. 148). Even if an investor's best guess is that bonds are not likely to suffer a significant future capital loss, they may want to sell some of their bonds and hold on to the money they receive if they have little confidence in the validity or accuracy of their forecast. Keynes said, as quoted above, that when we "dis­trust our own calculations [or expectations] and conventions concerning the future" and when "our precarious conventions have weakened," we shift from bonds to money. What he means is that when investors lose "confidence" in their ability to form reliable expectations of future interest rates, they will sell some bonds and hold on to the money they receive. This will cause bond prices to fall and interest rates to rise. This process, if strong enough, can trigger a financial panic. Keynes's crucial theory of "conventional" expectation and confidence formation is discussed in chapter12 of The General Theory and in Chapter 17 of this book.

Because expectation and confidence formation are discussed in chapter 12, Keynes assumed that the reader of chapters 13-15 would be familiar with the material in that chapter. In chapter 12, he argued that in forming expectations, "our usual practice [is] to take the existing situation and to project it into the future, modified only to the extent that we have more or less definite reasons for expecting a change" (CW 7, p. 148). In other words, in normal times, agents formulate expectations by extrapo­lating the recent trajectory of the economy into the intermediate future, and they presumably have more or less confidence in the reliability of expectations thus formed depending on how accurate such forecasts have been in the recent past. Because both expectations and confidence are endogenously determined, the economy does not need to be exogenously shocked to move - it moves on its own, endogenously.

Any change in expectations or in confidence will cause a shift in the liquidity preference (or demand to hold money rather than bonds as an asset) function as it is typically specified in Modern Keynesian textbooks and by Keynes on page 199 of The General Theory: L = L₁ (Y) + L₂ (r). L₁ is the transaction demand plus the precautionary demand to hold money, while L₂ represents the cru­cial "speculative" demand to hold financial wealth in the form of money rather than bonds. "L₁ mainly depends on the level of income, whilst L₂ mainly depends on the relation between the current interest rate and the state of expectation" (CW 7, p. 199, emphasis added).¹¹ Keynes's liquidity prefer­ence function should therefore have been written as L₂ (r_t, r^e_t) or perhaps L₂ (r_t - r^e_t), where r_t is the current interest rate and r^e_t is the expected future interest rate in period t.¹²

Since the expected future value of the interest rate is included in Keynes's L₂ function in The General Theory as an exogenous shift parameter,

Disequilibrium processes in bond markets 231 changes in the degree of liquidity preference "are primarily due to changes in the expectation affecting the liquidity preference function itself" (CW 7, p. 197). In other words, every time expectations change - and they change much of the time - the L₂ function shifts. Keynes explained that in order to generate the standard classical result that a fall in r will cause an increase in the demand to hold money, the L₂ function must assume a constant or "given rate of expectation" (CW 7, p. 202). Given that Keynes assumed in chapter 12 that expectations are typically formed by extrapolation from the recent past, every time the interest rate changes in an unexpected way the expected interest rate also changes, causing the L₂ (r) function to shift and the actual interest rate to change again in an ongoing endogenous process.

Chapter 12 also points to the importance of confidence in the reli­ability of expectation formation to agent choice under uncertainty. The complete specification of Keynes's liquidity preference function there­fore should have been written as L₂ (r_t, r^e_t, C_t), where C_t is an index of confidence in the reliability of expectations. If the interest rate moves through time as expected, confidence in the reliability of expectations will increase, causing a shift in the standard L₂ = L (r) function that will cause the interest rate to change. Since both expectations and con­fidence change endogenously, the standard L₂ function is always shifting, constantly changing the temporary “equilibrium” value of the interest rate.¹³ The interest rate is thus what Shackle called "an inherently restless vari­able" (Shackle 1972, pp. 163-164). Minsky argued that the longer the price of a financial asset remained stable, the more confident investors would become that the asset is not very risky, which would lead them to buy more of that asset, raising its price. In his words: "stability is destabilizing."

Keynes pointed out that shifts in the L² function could, at times, be sharp or "discontinuous," creating or exacerbating interest rate volatility. "Changes in the liquidity function itself, due to a change in the news which causes revision of expectations, will often be discontinuous, and will, therefore, give rise to a corresponding discontinuity of change in the rate of interest" (CW 7, p. 198). This discontinuity will be especially pronounced if a large majority of investors come to believe with convic­tion or confidence that bond prices are likely to suffer a significant fall. To state the case in extreme form: "If the change in the news affects the judgment and requirements of everyone in precisely the same way, the rate of interest... will be adjusted forthwith to the new situation without any transactions" (CW 7, p. 198). This is obviously not a vision of a bond market that always has deeply rooted equilibriums.

Keynes also argued that the forces that made stock prices so volatile in the "insane" US gambling casino also operated in the US bond market in this era. In chapter12, he wrote:

A conventional valuation [of stock prices] which is established as the outcome of the mass psychology of a large number of ignorant individuals is liable to change violently as the result of sudden fluctuations of opinion due to factors that do not really matter much to the prospective yield; since there will be no strong roots to hold it steady.

(CW 7, p. 154)

In chapter13, he wrote:

Just as we found the [price of a stock] is fixed, not by the "best" opinion, but by the market valuation as determined by mass psych­ology, so also expectations as to the future of the rate of interest as fixed by mass psychology have their reactions on liquidity preference.

(CW 7, p. 170)

There is a paradox of sorts in the role played by money, the safe asset, in Keynes's theory of bond market instability. Where markets them­selves are liquid, so that securities can be traded quickly for money at low transaction costs, the ability to rapidly shift from risky long-term securities to money in the event of market trouble will calm the nerves of investors. This will induce them to take what are objectively (though not subjectively) greater risks in a financial boom because they believe they can quickly shift to money and out of risky securities if trouble develops.¹⁴ If markets were not liquid and there was no safe asset, investors would have to be much more cautious about holding risky securities, which would, ceteris paribus, raise interest rates and lower income and employment. But at the same time, the existence of an asset without risk of nominal capital loss can worsen the degree of panic in all long-term financial markets because, by providing a safe haven for frightened investors, it enables them to flee all long-term risky assets, potentially creating or exacerbating a financial panic. Keynes said that the high degree of liquidity in modern markets:

presents us with a dilemma. For, in the absence of an organised market, liquidity-preference due to the precautionary motive would be greatly increased [raising interest rates]; whereas the existence of an organised market gives an opportunity for wide fluctuations in liquidity-preference due to the speculative motive.

(CW 7, pp. 170-171)

The highly liquid bond market is therefore subject to points of "wide fluctuations" in the value of the long-term interest rate - a constant theme in these chapters.