Demand for financial assets and monetary policy: a restatement of the liquidity preference theory and the speculative demand for money

Abstract:

Keynes implicitly used the concept of duration to analyze the impacts of expected changes in the price of a perpetual bond and coupon payments that led to his “square rule.” Keynes’s result (“square rule”), derived from the breakeven condition, was just a simplification to illustrate the importance of expected interest changes on the total return of an investor’s portfolio, constrained to money and consols. Keynes’s world is not one of yield, but one of total return, where price and yield form the returns. This paper aims to show that the same breakeven method can be used, but instead of a simple two-asset model (money and a specific financial asset such as consols) this framework can be generalized to take into account a broad spectrum of financial assets including perpetuals, zero-coupon, coupon bearing securities for different maturities and investment horizons, and risk preferences to analyze the impacts of expected changes in interest rates on total expected return. By using Keynes’s breakeven condition and total return analysis, it is possible to generalize this argument for different types of securities (and not just consols) and it follows the approach of chapter 17 of the General Theory. In this paper, I update and extend Keynes’s analysis to allow for different holding periods (or investment horizons) and different financial instruments. When we take into account duration and convexity effects, and given rising expected changes in interest rate, we get an upward sloping demand for securities with short duration relative to investment horizon rather than a downward sloping speculative money demand schedule. In this framework, duration and breakeven analysis play a crucial role in the demand for financial assets as bank’s expectation on rate changes will generate a demand for short (or long) duration financial assets impacting asset prices.

Key words::

• liquidity preference
• speculative demand for money
• user costs

JEL classifications::

• E41
• E43
• E52
• E58
• G11
• G12

Notes

¹However, the post Keynesian literature on banks’ liquidity preference (e.g., Carvalho, 1999; Wray 1990) pays little attention to the concept of duration as the basis for asset demand. Kregel (1998) first used the concept and, more recently, Tymoigne (2009) extended this analysis to incorporate Kahn’s distinction between income- and capital-sensitive investors.

²As post Keynesians have argued, Keynes challenged Fisher’s hypothesis that nominal interest rates would adjust to changes in inflation expectations and his definition of income (Kregel, 1998; Tymoigne, 2009). For the sake of exposition (and simplicity), Keynes used perpetual bonds to show the impacts of expected changes in interest rates on both coupon payments and bond market prices (see, e.g., Kregel 1998; Tymoigne, 2009). Keynes argued that Fisher’s effect generates capital losses on long-term bonds, offsetting income gains from changes in coupon payments and reinvestment income.

³Robinson (1951) stresses the importance of the structure of the market and Kahn (1954) draws on this approach to distinguish investors according to their sensitivity to income and capital risk, whereas Tobin (1958) relies on diversity of portfolio holdings instead of diversity of opinion to derive the demand for money schedule.

⁴As Minsky (1986), Wray (2009), and others have emphasized, the U.S. financial system is dominated by money managers. In this phase of capitalism, institutional investors dominate over the enterprise. The primary concern is to generate short-term returns using leveraged funds, which results in increasing speculation and volatility (declining holding period, high turnover rates, etc.).

⁵I draw heavily here from Toevs (1983).

⁶For a proof of the immunization theorem, see de La Grandville (2001, pp. 148–149).

⁷For the sake of exposition, this section assumes a parallel change in the term structure immediately after the bond is purchased and no additional changes for the rest of the investment horizon. These assumptions will be relaxed later.

⁸This analysis assumes a parallel change in the term structure of interest rates immediately after the bond is purchased and no other changes for the remainder of the investment horizon.

⁹If one adds Minsky’s instability theory then this uncertainty is augmented by the need to meet future financial obligations. Changes in market rates produce changes in cash flows and asset prices, thus affecting the financial position of an economic unit.

¹⁰The concept of the yield curve is not the same as the term structure of interest rates. Yield curve is not the appropriate measure because it combines securities with different coupon-payment characteristics for different maturities and different reactions to changes in yield in the same graph (see, e.g., Svensson 1994). Zero-coupon instruments and the theoretical spot-rate curve can then be used to price Treasury securities and to discount each cash flow of a given security to calculate the duration measure.

¹¹This concept goes back to Irving Fisher. As Kregel (1999) observed, “The modern financial analyst would be surprised to discover in The Nature of Capital and Income (1906) and The Theory of Interest, most of the basic tools of time value analysis of investment including spot and forward curves” (Kregel, 1999, p. 268).

¹²Much of the discussion of this subsection draws heavily from Ilmanen (1995a) and on Ilmanen (1995c).

¹³See, for example, Davis and Aliaga-Díaz (2007) and Ilmanen (1996). Note that if the underlying distribution is skewed, then forward prices can be regularly below the spot price, underestimating the spot price at expiry, and convergence does not occur. In the context of foreign exchange markets, the mainstream literature has shown that forward prices are biased predictors of the future change in the spot rate.

¹⁴Ilmanen (1995f) derived the equations that show the decomposition of bond returns. Expected returns can be decomposed into three main components: rate expectations, risk premium, and convexity bias.

¹⁵The increasing demand for short-term and safe securities, such as Treasuries, produced a shortage of supply driving their prices up in spite of rising outstanding government debt (see, e.g., Noeth and Sengupta, 2010).

Additional information

Notes on contributors

Felipe Rezende

Felipe Rezende is an assistant professor of economics at Hobart and William Smith Colleges and a research fellow at MINDS (Multidisciplinary Institute on Development and Strategies).