Financial cycles from technology shocks
Phelan et al., in this paper, explore the dynamics of macro-financial boom-bust cycles driven by technological shocks. The authors develop a theoretical model to explain how these cycles can be anticipated and rationalized within a macro-finance framework. More specifically, Phelan et al. employ a macroeconomic model based on the work of Kiyotaki and Moore (1997), which incorporates credit markets and collateral constraints. The key innovation in their model is the introduction of an anticipated technology shock to a secondary, innovative sector. This shock does not initially impact the primary leveraged sector but instead affects a new sector that may take on leverage to invest. The model assumes perfect foresight and rational expectations, meaning that economic agents can anticipate the consequences of the productivity shock without any behavioral biases or sluggishness in beliefs.
Key findings:
- Boom-bust dynamics: The model demonstrates that news of future technological improvements in a secondary sector can lead to predictable boom-bust cycles. Initially, asset prices rise due to optimistic expectations about future productivity, allowing for increased leverage and credit expansion across the economy.
- Capital reallocation: As the shock impacts the secondary sector, capital is reallocated from primary users (who are leveraged) to secondary users. This reallocation leads to a bust when primary users need to repay their debts but lack sufficient resources due to the earlier credit expansion.
- Predictability of cycles: Unlike typical financial crises narratives that focus on unexpected shocks, this model shows that such cycles can be perfectly foreseen as part of normal economic dynamics. The cycles align with Minsky’s narrative where financial instability arises naturally within capitalist economies.
- Implications for policy: Understanding these predictable cycles can help in formulating policies that mitigate the size of booms and busts by managing credit expansions more effectively.
- Relevance to historical events: The model’s dynamics reflect historical episodes like the Great Recession and sudden stops in emerging markets, where technological innovations led to initial booms followed by financial contractions.
The paper’s findings underscore the importance of considering reallocative technology shocks as drivers of financial cycles and suggest that these shocks provide a better explanation for Minsky cycles than direct financial condition changes.
Technology Shocks and Predictable Minsky Cycles
Authors: Jean-Paul L’Huillier, Gregory Phelan, Hunter Wieman
From: Office of Financial Research, Federal Reserve Bank of Cleveland, Princeton University
The evolution of technology shocks: evidence from U.S. data
L’Huillier et al. in this paper examine changes in the nature of technology shocks and their propagation in post-war U.S. data. The authors use a Vector Autoregression (VAR) model with an agnostic identification procedure to identify technology shocks. They employ a “max share” approach, in order to identify the shock that best explains the variance in Total Factor Productivity (TFP) at a long but finite horizon. The analysis is conducted over two subsamples: 1954Q2-1983Q4 and 1984Q1-2019Q4, split at the onset of the Great Moderation. Finally, the baseline VAR includes 5 variables: TFP, GDP, consumption, hours worked, and the S&P 500 stock market index.
Key findings:
- The nature of technology shocks has changed over time: The transmission of technology shocks differs markedly between the two subsamples. The importance of technology shocks as drivers of aggregate fluctuations has increased over time.
- In each subsample TFP only rises after several periods, and then grows gradually beyond that.
- There are striking differences in the response of key variables across subsamples:
– In the first subsample, hours worked fall in response to the technology shock, while GDP and consumption rise modestly.
– In the second subsample, there is a broad-based expansion with hours worked, GDP, and consumption all rising.
- The identified shock explains a similar share of TFP variation in both subsamples, but accounts for a much larger share of GDP fluctuations in the second subsample.
- Additional variables show consistent patterns:
– Inventories, investment, and real wages fall in the first subsample but rise in the second.
– The BAA spread rises in the first subsample but falls in the second.
- The authors interpret these findings as evidence of changes in the structure of the U.S. economy, affecting shock propagation rather than changes in the shock processes themselves.
The paper goes on to develop a structural model to explain these empirical findings, focusing on changes in monetary policy, the nature of intangible capital accumulation, and capital utilization costs as potential drivers of the observed changes.
The Changing Nature of Technology Shocks
Authors: Christoph Görtz, Christopher Gunn, Thomas A. Lubik
From: University of Augsburg, Carleton University, Federal Reserve Bank of Richmond
