By Aidan Levi-Minzi

In most economics courses, students are taught models that are built in dynamic stochastic form. These models attempt to explain certain economic phenomena based on general equilibria, such as the famous supply and demand curve. However, a new method of modeling has become increasingly popular after its successful prediction of both the 2000 dot-com bubble and the 2008 Global Financial Crisis. The Stock-Flow Consistent approach was introduced nearly five decades ago by John Maynard Keynes and James Tobin. By integrating a strict accounting framework, Stock-Flow Consistent (SFC) allows for the real side of the economy to be determined by the financial side (as is usually the case in practice), rather than simple measures of supply and demand as seen in Dynamic Stochastic modeling. It also provides for some interesting conclusions in its own right because it imposes extreme constraints and reduces the model’s degree of freedom. 

SFC models are built around three main pillars, the first being flow consistency. Every monetary flow that comes out of somewhere goes into someone else’s pocket. All inputs in the model must be balanced out, so there are no exogenous flows of capital in the system. A household’s income is a firm’s cost. Exports of one country are imports of another. This creates the second pillar, which is stock consistency. In introducing a Treasury bond, there is now a liability for the government and an asset for its holder. And finally, the third pillar is stock-flow consistency. Every flow involves the change in at least one stock. The idea is to account not just for movement in assets, but the change in total wealth that comes with it. 

Given that an SFC model accounts for both the real side of the economy and its financial markets in an intuitive manner, it would be beneficial to take a look at how quantitative easing would play a role as a macroeconomic policy, specifically during periods of financial hyper valuation and asset price bubbles. The Federal Reserve started to purchase securities as a response to the Covid-19 Pandemic, which some economists claim has created a bubble. The current Buffett Indicator, which measures the ratio of total market cap to total GDP, has recently capped out at 224%: nearly two standard deviations above the mean, and a measure not seen since right before the 2000 dot-com bubble. By using a stock-flow consistent model, we can study the effects of quantitative easing on these asset bubbles, as well as other parameters that may act as catalysts. 

Tulipmania was the first recorded asset bubble in history. While price data from the 17th century is difficult to come by, it was recorded in a few shopkeepers’ records that a single bulb could be exchanged for two barrels of wheat, four barrels of rye, four fat oxen, eight fat pigs, twelve fat sheep, two barrels of wine, four barrels of beer, two barrels of butter, a thousand pounds of cheese, a complete bed, a suit of clothes, and a silver mug. In modern day practice, central banks have turned to quantitative easing in response to these dilemmas, especially when they have exhausted other weapons in their arsenal, such as interest rate reduction. The purpose of this was not to affect long-term equilibrium, but rather to ease the recession’s short-term pains. Economists came to two different conclusions. Some argued that the conversion of the larger monetary base to actual M1 would be gradual in a contracting economy. Those more skeptical were worried that with central banks buying long-term government bonds, investors would have no other choice but to continue speculating into riskier and riskier assets, eventually leading to a bubble. In any case, it is important to look at the macroeconomic effects of these policies. 

The literature on Dynamic Stochastic General Equilibrium (DSGE) modelling makes the case for these asset bubbles generally being the result of some exogenous shock to the supply-side of the economy, rather than the endogenous catalysts as described by Hyman Minsky. Minsky’s bubble is based around profligate investing: a behavioral variable with endogenous traits. In other words, bubbles aren’t caused by current events or political interference, but rather by investor sentiment from within the system. Hence, this SFC model is built with exogenous quantitative easing and endogenous bubble behavior. 

While there are too many equations in the model to be listed here (47 of them), they can all be seen here. For our purposes, we only really need to look at households’ propensity to consume, the size of central bank buyouts, the duration of the assets they are acquiring, and (surprisingly) a firm reactivity parameter, which is the rate at which firms alter their inventories in order to meet next period’s expected demand. The three dependent variables being measured are overall consumption based on disposable income, GDP (based on the typical Y = C + G + I), and economic panic. Economic panic is based on Tobin’s q, which is a stock market valuation metric that divides the value of the stock market by all corporate net worth (similar to the Buffett Indicator). In theory, this ratio is at equilibrium when it reaches 1. For every period that q is less than 1 (and stocks are still seen as undervalued), economic panic is multiplied by some panic rate > 1, giving a discrete version of exponential panic. 

In the first scenario, we assume a low propensity to consume out of income: about 0.3. While quantitative easing gave a boost in long term consumption, GDP saw little change, with nearly identical peaks and troughs before smoothing out around the same time. The impact of quantitative easing does, however, abate some of the economic panic, which would allow bubbles to continue to grow. 

Now let’s raise the propensity to consume to 0.4. During the current economic shutdowns, Americans have found themselves with more liquidity and higher personal income, which has led to an increase in spending and market participation. In this model, the immediate difference is that, with quantitative easing, the cyclicality of GDP is much more volatile. While both scenarios see GDP smoothing out around the same time, it is the quantitative easing approach that shows much more dramatic peaks and troughs. With households consuming more, they are more likely to be investing, including stocks and bonds. This is observed in the near zero economic panic indicators from both simulations. Hence, the model concludes that quantitative easing only creates a bubble when coupled with yield-seeking households looking to invest in riskier and riskier assets. This creates a sort of chicken or the egg dilemma: is it the lack of risk-averse investors, or does quantitative easing give those investors no option but to speculate? 

Let’s turn to other parameters to see their effect on economic panic. One such variable is the rate at which firms are able to bridge the gap between their current inventories and market demand for their goods. Another term for this phenomenon is the Bullwhip Effect, which seeks to describe the output inefficiencies yielded from incorrect demand forecasts. If this reactivity variable is a bit lower, GDP takes massive peaks and dips without ever reaching a steady state. With firms never really reaching their expected market demand for the next time period, there is a vicious cycle of exogenous demand shocks followed by supply shocks in perpetuity. This creates a cycle of economic panic that increases over time, and these panics are what ‘pop’ the bubble. Interestingly, during February and March of 2021, many manufacturers are finding themselves vastly under-staffed and under-producing as frustrated Americans are buying cars or furniture to make up for their lack of spending on restaurants and hotels during the Coronavirus pandemic. It is important to note that the bullwhip effect does not persist and firms will eventually alter their production schedule to match expected demand more accurately. However, the model does point to this parameter as the culprit of large differences in cyclical behavior. 

According to the SFC model, asset bubbles are a result of risk-tolerant investors being given the opportunity to speculate with quantitative easing. Quantitative easing lowers overall economic panic, which allows bubbles to get larger and larger. It also has little effect on overall GDP in the long-run. Perhaps central banks could use their tools of forward guidance to better assist firms in future expected demand and therefore raise their reactivity rate of current inventories, thereby lowering the likelihood of an exogenous supply shock to an already endogenous panic shock (something that is much easier said than done, as proven in the past 12 months). They could also raise interest rates when implementing quantitative easing, with the hopes of decreasing households’s propensity to consume, which would allow Americans to increase their wealth in the short run to better prepare them for the long run. Unfortunately, central banks would most likely never implement this strategy as the entire idea of quantitative easing is to stimulate the economy; a tool that is used only when interest rates have already reached the zero-lower bound. 

With this in mind, the model shows that quantitative easing only prolongs the inevitable. Fed Chair Jerome Powell has made it clear that the Fed is focused only on its dual mandate: controlling inflation and reducing the unemployment rate. While the models show that quantitative easing does smoothen cyclicalities in GDP, it also proves that it only takes a few other indicator shifts to alter that outcome. □

---

Work Cited

1. Image source
2. Current Market Valuation. (2021, February 18). “Buffett Indicator”. Current Market Valuation. Retrieved from http://www.currentmarketvaluation.com/models/buffett-indicator.php
3. Tita, Bob and Hufford, Austen. (2021, February 22). “Consumer Demand Snaps Back. Factories Can’t Keep Up”. The Wall Street Journal. Retrieved from https://www.wsj.com/articles/consumer-demand-snaps-back-factories-cant-keep-up-11614019305?mod=hp_lead_pos5
4. Godley, Wynne and Lavoie, Mark. (2012, 27 March). “Monetary Economics”. Palgrave MacMillan. 
5. Hass, Cameron. (2017, September). “Quantitative Easing and Asset Bubbles in a Stock-flow Consistent Framework”. Levy Economics Institute. Retrieved from http://www.levyinstitute.org/publications/quantitative-easing-and-asset-bubbles-in-a-stock-flow-consistent-framework. 
6. Mackay, Charles. (1841). “Memoirs of Extraordinary Popular Delusions and the Madness of Crowds”. London: Richard Bentley. 
7. Nikiforos, Michalis. (2017, May). “Stock-Flow Consistent Macroeconomic Models”. Levy Economics Institute. Retrieved from http://www.levyinstitute.org/publications/stock-flow-consistent-macroeconomic-models-a-survey. 
8. Palley, Thomas. (2011). “Quantitative Easing: A Keynesian Critique”. PERI Working Paper. Retrieved from https://econpapers.repec.org/paper/umaperiwp/wp252.htm#:~:text=Abstract%3A%20Keynesian%20economists%20have%20generally,that%20response%20may%20be%20misplaced.
9. Solow, Robert. (1956, February). “A Contribution to the Theory of Economic Growth”. Quarterly Journal of Economics. Retrieved from http://rcin.org.pl/Content/39010/WA51_55789_r2013-t85-z4_Przeg-Geogr-Judge.pdf