Those of you who follow me on Twitter know that my economic views do not fall neatly within any of the established schools of thought. Over the last several years, I’ve worked on a new framework for understanding the economy (summarized in a series of posts here: Alternative View of the Business Cycle).
I would characterize my views as post-Keynesian with neoclassical flavors.
- Post-Keynesian because I point to insufficient aggregate demand as the reason for unemployment, plus I incorporate banks and endogenous money.
- Neoclassical because micro-foundations and capital markets feature very prominently in the proposed framework.
For those interested, here is my SSRN paper with the formal model:
Gain-Loss Utility, Equilibrium and Interest Rates by Valeri Popov :: SSRN
Below I’ve copied the Introduction. I think it provides a good and brief overview. I would love to hear your thoughts and feedback.
Introduction
The author uses a new utility framework to construct a micro-founded model of the economy that reveals the monetary base as the source of economic fluctuations.
The benefits of the new utility framework, referred to as gain-loss utility, are three-fold (Popov June 2017 and Popov Dec 2017). The framework uses reference-dependent preferences to explain observed violations of standard consumer and decision theory, including experimental evidence on utility discounting. Time preference and risk tolerance, the two aspects of human behavior that govern borrowing and saving decisions, arise endogenously within the utility framework itself. In contrast, mainstream micro-foundations rely on exogenous parameters that give rise to a number of puzzles, most notably the Equity Premium Puzzle (Kocherlakota 1996) and the empirical failure of the Euler Equation (Canzoneri 2007). Last but not least, utility is defined as the agent’s maximum willingness to work to obtain the goods and services she wants to consume. Since willingness to work is measured in time and time is a universal cardinal measure, preferences under gain-loss utility are cardinal in nature. This allows for agent heterogeneity and macro aggregation.
According to gain-loss utility, the desire to borrow or save is determined by the agent’s income-growth expectation relative to the geometric mean of the risk-free interest rate and the expected change in the agent’s needs over the respective term. Using this basic insight, an intertemporal-choice model can be constructed that reveals the necessary and sufficient condition for equilibrium between aggregate desired borrowings and savings. Aggregate desired borrowings and savings are in equilibrium when risk-free interest rates equal the aggregated income-growth expectations in the economy (Popov June 2017).
In this paper, the author assesses the factors that determine market interest rates and the conditions under which market interest rates converge with the equilibrium level defined above. The author starts by describing the fundamental building blocks of a monetary economy — agents endowed with labor-time, an omnipresent bank and, of course, money. Agents use their labor-time endowment to produce goods and services that can be exchanged for money. The omnipresent bank is an institution that represents the central taxing authority and the entire banking system. As the omnipresent bank lends to agents, it effectively buys claims against their future income in exchange for claims against itself in the form of demand deposits. Such bank-issued claims are universally accepted as money because they are guaranteed by fiat to redeem debt and tax obligations.
Next, the author puts forth a new general-equilibrium framework. General equilibrium in the Walrasian tradition collapses all markets in space and time into a single market where an omni-prescient auctioneer finds the price vector that zeros-out excess demands (Arrow 1954). Money is assumed to exist ex-ante as an object with no intrinsic value in and of itself, simply a veil that facilitates exchange (Hahn 1965). Instead, the author collapses the economy into two markets — the spot market for labor-time where agents trade goods and services[1], and the futures market for labor-time where agents issue and redeem claims against their future income, with the omnipresent bank creating and destroying money in the process.
Effectively, the bank acts as the market-maker in the futures market for labor-time. Agents borrow from the bank in anticipation of future incomes. The bank funds such borrowings with newly-created money. Agents use the money to trade in the spot market for labor-time, buying and selling goods and services. This generates nominal incomes that agents use to pay-off their debts to the bank, resulting in the redemption of the corresponding money balances. It is this cycle of money creation and destruction that enables trade in a monetary economy. The omnipresent bank plays a pivotal role — the bank’s willingness and capacity to extend credit and enforce claims against borrowers’ incomes provide the commitment and enforcement mechanisms required for a medium of exchange to exist.
This setup reveals another notable difference from a Walrasian economy. In a Walrasian economy, supply creates its own demand because the omni-prescient auctioneer changes prices until excess demands across all markets, including the labor market, are zeroed-out. In other words, the Walrasian auctioneer ensures that the economy is always in equilibrium consistent with full employment — if you bring your goods to market (or labor-time), there is a price at which you will find willing buyers.
In a monetary economy, agents spend in anticipation of future incomes, using newly-created money borrowed from the bank. In the aggregate, such spending generates nominal incomes, thereby fulfilling agent income expectations. In other words, a monetary economy bootstraps — in the aggregate, nominal incomes must be anticipated ex-ante in order to be generated ex-post. This means that bank-financed demand creates its own nominal supply with prices and employment being the unknown residuals. Prices determine only relative demands, but aggregate demand is determined ex-ante pursuant to income expectations plus desired borrowings less desired savings. Unlike a Walrasian economy, there is no guarantee that a seller of goods and services (or labor-time) will find willing buyers unless demand has been bootstrapped in the aggregate.
With respect to equilibrium, a monetary economy is always in equilibrium because demand, by identity, equals nominal supply. However, because of bootstrapping, such equilibrium may not be consistent with full employment. To illustrate, unemployed agents will not borrow to finance their demand (nor is the bank going to extend them credit) because such agents do not have anticipated incomes. As a result, aggregate demand will be insufficient to provide them with jobs (unless someone else bootstraps demand for the potential incomes of unemployed agents).
In Section 1, the author starts with a model of a single-period monetary economy with period-stamped money that expires at the end of the period. The purpose is to illustrate that the persistence of money is not essential for exchange. In other words, the value of money is unrelated to the medium-of-exchange motive. Rather, the value of money stems from the fact that money represents a claim on future incomes. Another take-away from Section 1 is that an employer-of-last-resort can bootstrap a monetary economy to full employment. Since the omnipresent bank in this model allows agents to redeem debt with either money or labor-time, the single-period economy enjoys full employment and price stability.
In Section 2, the author expands the time horizon to two periods. Money is no longer period-stamped, and agents have the option to borrow and save beyond the current period. On the flip side, agents are no longer able to redeem debts to the omnipresent bank with labor-time. This ensures that debits and credits always balance (for every dollar of savings, there must be a dollar of borrowings). Without an employer-of-last-resort, the economy is subject to uncertainty with respect to nominal incomes, employment and prices.
If the bank sets interest rates below the equilibrium level, desired borrowings will exceed desired savings, boosting ex-post incomes above expectations and causing prices and employment to rise. If the bank sets interest rates above the equilibrium level, desired savings will exceed desired borrowings, depressing ex-post incomes below expectations and causing prices and employment to fall. If the bank sets interest rates at the equilibrium level, desired borrowing will equal desired savings and incomes will meet expectations with no change to employment and prices. However, it’s notable that such steady state may not be consistent with full employment. Because of the bootstrapping problem, a two-period economy, where money persists from period to period, may suffer from unemployment.
Section 3 expands the time horizon beyond two periods and introduces risk-free financial assets that have duration. Duration measures the change in the value of an asset or liability with a change in interest rates. Duration exposes the bank to interest rate risk because money, which is a liability to the bank, has zero duration. Long-term loans at a fixed rate of interest create a duration mismatch on the bank’s balance sheet because such loans have negative duration. This can lead to losses if future interest rates rise. Accordingly, the bank duration-matches its balance sheet by selling assets with duration directly to savers[2].
Borrowers and savers also engage in duration-matching. Long-term borrowers want to borrow at a fixed rate of interest that does not exceed their income-growth expectations. Long-term savers with positive income-growth expectations want to buy assets with duration in order to lock-in expected returns in excess of their required returns. The supply of duration by borrowers, that’s intermediated by the bank, and the demand for duration by savers give rise to capital markets and place restrictions on the bank’s capacity to set interest rates. The bank sets interest rates such that assets with duration, issued in the current period, trade at par[3].
Crucially, the author finds that long-term savers, who have negative income-growth expectations, will not engage in duration-matching but will demand to hold zero-duration money, instead. The existence of cash means that the interest rate on demand deposits cannot fall below zero. Accordingly, savers with negative income-growth expectations can meet or exceed their negative minimum required returns simply by holding money in the form of interest-bearing demand deposits. They keep all the upside if interest rates on demand deposits were to rise in the future, but none of the downside if rates were to fall to zero since zero is still higher than their minimum required returns. This demand for money, referred to in the paper as asset money demand, can be satisfied only if the omnipresent bank goes long on duration and takes on interest-rate risk. In the real world, the central bank is the institution that meets this demand by supplying the monetary base.
In this model, the omnipresent bank has only partial control over interest rates. The bank cannot set the absolute level of long-term interest rates, which is determined in capital markets, but the bank does determine the relative level in relation to the equilibrium level. If the monetary base is greater than asset money demand, observed interest rates will be lower than the equilibrium level. If the monetary base is less than asset money demand, observed interest rates will be higher than the equilibrium level. Only when the monetary base equals asset money demand, would interest rates converge with the equilibrium level. The implications for the economy in terms of bootstrapping, employment and prices are the same as in the two-period economy in Section 2, the only difference being that the omnipresent bank controls the supply of base money as opposed to the absolute level of interest rates.
The author also examines the market for zero-duration borrowings and savings (aka, money markets) and finds that there is an identity between the supply and demand for money that is enforced through fluctuations in nominal incomes. Under income uncertainty, not all agents bootstrap their demand. Accordingly, the author splits agents into two groups — bootstrapping entrepreneurs, who spend in anticipation of future revenue, and non-bootstrapping workers, who spend only after incomes have been earned. To illustrate, entrepreneurs borrow in the current period to hire workers and produce an inventory of goods and services. The newly-created money, borrowed from the bank, flows to workers in the form of wages, giving rise to what the author refers to as transaction money demand. Workers use such zero-duration savings to fund spending in future periods, buying the very same inventory they produced in the current period.
Because of transaction money demand, entrepreneurs are not in position to sell all the inventory produced in the current period. Accordingly, they must carry the corresponding zero-duration borrowings into the future, giving rise to the supply of endogenous money. This component of money supply is referred to as endogenous because commercial banks can facilitate the underlying borrowings without creating a duration mismatch. Endogenous money supply and transaction money demand are residuals of ex-post incomes that enforce the ex-post identity between total money supply and demand.
Section 4 introduces risky assets such as corporate bonds and equity. The author finds that risk, measured by the expected variance of asset returns, does not materially change the model in Section 3. According to gain-loss utility, minimum required returns are determined by the risk-free rate, the riskiness of the asset and the risk tolerance of the respective investors (Popov Dec 2017). Risk tolerance is given by the asset ratio, which measures the investment in the risky asset as percent of the agent’s expected future income that’s unrelated to the asset. The author finds that savers allocate their savings into risky assets until expected returns on a risk-adjusted basis equal the risk-free rate. By the same token, borrowers issue risky assets until they are indifferent between borrowing at the risk-free rate from the omnipresent bank or issuing risky assets. In other words, for the purpose of this model, risk can be ignored because agents perceive all saving and borrowing opportunities on a risk-adjusted basis.
In Section 5, the author compares the proposed framework to the neoclassical and Keynesian views of unemployment and interest rates. The paper concludes with a brief discussion of the business cycle implications under three monetary regimes: fixed reserves such as the gold standard, fiat money in advanced economies and fiat money in developing economies.
[1] Goods and services can be thought of as encapsulations of labor-time
[2] Alternatively, the bank can issue liabilities with duration such as CDs.
[3] The author assumes that the omnipresent bank duration-matches its balance sheet and seeks to break-even. Both rules are a realistic. Commercial banks are required to duration-match by their regulators. Also, banks seek to make a profit, but competition among banks eliminates excess profits.
