Federico Di Pace and Christoph Görtz
There is ample evidence that a monetary policy tightening triggers a decline in consumer price inflation and a simultaneous contraction in investment and consumption (eg Erceg and Levin (2006) and Monacelli (2009)). However, in a standard two-sector New Keynesian model, consumption falls while investment increases in response to a monetary policy tightening. In a new paper, we propose a solution to this problem, known as the ‘comovement puzzle’. Guided by new empirical evidence on the relevance of frictions in credit provision, we show that adding these frictions to the standard model resolves the comovement puzzle. This has important policy implications because the degree of comovement between consumption and investment matters for the effectiveness of monetary policy.
Empirical evidence on the comovement puzzle
We first conduct an empirical investigation into the source of comovement and find this to be the credit channel, ie the provision of funds by the financial sector to finance production and investment activities of non-financial firms. To do so, we employ state-of-the-art methodology to identify (conventional) US monetary policy shocks in a structural vector autoregression (SVAR) by using an external instrument recently proposed by Miranda-Agrippino and Ricco (2021). This instrument utilises high-frequency data to ensure robustness to the presence of information frictions in the economy. Chart 1 shows the impulse responses to an unexpected rise in the policy rate. We find that a contractionary monetary policy shock triggers a decline in both, investment and consumption goods production. We find this goes hand in hand with a tightening in financial conditions, evident from the rise in the excess bond premium. Chart 2 shows that the shock also triggers a substantial and persistent contraction in private banks’ equity capital. This tightening in financial conditions is evidence that the transmission mechanism of monetary policy operates via financial markets and its link with investment. The credit channel has become increasingly popular after the global financial crisis and its empirical relevance has recently been stressed particularly for the transmission of monetary policy shocks (Caldara and Herbst (2019) and Miranda-Agrippino and Ricco (2021)).
Chart 1: Empirical responses to a contractionary monetary policy shock
Chart 2: The response of bank equity to a contractionary monetary policy shock
The importance of supply-side financial frictions
No existing studies that address the comovement puzzle account for the empirical movements in bank equity and credit spreads (see eg DiCecio (2009); Sterk (2010); Carlstrom and Fuerst (2009); Katayama and Kim (2013); Di Pace and Hertweck (2019) among others). They either assume frictionless financial markets, or consider frictions in firms’ demand for investment funding. Neither can give rise to positive credit spreads, nor account for the documented supply-side frictions in financial markets. As such, they miss out an important channel that has been widely acknowledged in the literature as important for the transmission and amplification of various economic shocks. Further, the empirical evidence on the relevance of the credit channel begs for an extension of the two-sector New Keynesian model with a mechanism that accounts for these supply-side financial frictions. This mechanism can potentially be helpful also in aligning the model predictions with the empirical evidence and thereby offering an appealing way to resolve the comovement puzzle. This is exactly what we propose in our paper.
The role of financial frictions to resolve the comovement puzzle
We resolve the comovement puzzle by building on the New Keynesian model of Görtz and Tsoukalas (2017), which has two distinct sectors that produce consumption and investment goods, respectively. We account for credit frictions, which our empirical analysis found to be the source of the comovement, by adding financial frictions as in Gertler and Karadi (2011). These frictions arise because financial intermediaries are subject to an endogenous leverage constraint since the value of collateral limits banks’ ability to fund the real economy. As such, the two-sector New Keynesian model consists of standard components and nests a number of widely used frameworks (eg Justiniano et al (2011) or DiCecio (2009)). The model also includes the usual set of nominal and real rigidities that the literature has found important to match the hump-shaped responses in the data.
Chart 3: Model responses to a contractionary monetary policy shock
Chart 3 shows the model’s impulse responses to a contractionary monetary policy shock. We consider our baseline model with financial frictions (blue solid lines) and a model version without frictions in financial markets (red dashed lines). The chart illustrates that a rise in the policy rate in our baseline model triggers a decline in total output, as well as in investment and consumption. This model can solve the comovement puzzle and match the empirically observed rise in credit spreads alongside a sharp decline in bank’s equity capital. In contrast, the responses of the standard two-sector New Keynesian model without financial frictions display a comovement problem: the red dashed lines in Chart 3 show that investment does not decline alongside consumption in response to the tightening in monetary policy. In addition, the model without financial frictions falls short in matching the empirically observed rise in credit spreads and the decline in bank equity capital.
Why does the standard New Keynesian two-sector model fail to deliver comovement between consumption and investment? In the model without financial frictions, a rise in the policy rate causes a decline in inflation and a contraction in demand for consumer goods and services. This causes a decline in production in the consumption sector as firms that cannot adjust their prices reduce output in line with the reduction in demand. The fall in output in the consumption sector reduces the demand for labour in that sector, which in turn puts downward pressure on real wages and real marginal cost. Since labour is perfectly mobile across sectors, real wages also decline in the investment goods sector making it cheaper to produce these goods. For this reason, the relative price of investment goods falls, causing the demand for investment goods (ie, capital accumulation) to rise, contrary to what is seen in the data.
What is the mechanism that allows for comovement of investment and consumption in the baseline model? The baseline model exhibits an additional channel that dampens capital accumulation by weakening the financial position of banks. Chart 3 shows that the contractionary monetary policy shock reduces the price of capital (Tobin’s marginal Q). Since banks collateralise debt against the value of firms’ physical capital stock, a fall in the price of capital reduces the collateral value of capital claims, resulting in in a deterioration of bank equity capital. The severe contraction in equity capital, shown in Chart 3, is consistent with our empirical evidence. This is important as the dynamic response of equity capital is behind a sharp contraction in credit supply. Since banks are highly leveraged, the decline in bank equity exacerbates the reduction in lending to the real economy. The reduced funding for investment projects results in a contraction in investment. To rebuild their balance sheets, banks must charge a higher interest rate over the base rate, thereby increasing credit spreads.
The tightening of credit conditions in the baseline model is the channel, in comparison to the frictionless model, that limits the financing of investment projects and thereby induces a contraction in investment. Alongside the reduction in consumption, the fall in investment is consistent with the empirical evidence on comovement of sectoral outputs. Also consistent with their empirical counterparts are the dynamics of credit spreads and bank equity, which are crucial for facilitating comovement across expenditure categories.
Developing structural models which resolve the comovement puzzle has important policy implications since a lack of comovement can result in near monetary neutrality. We show that supply-side financial frictions are an important mechanism to resolve the comovement puzzle. This channel can be complemented by other mechanisms suggested in the literature – we show for example that the financial channel is strengthened by the introduction of nominal wage rigidities. However, our work accounts for an important dimension that has been ignored in the existing literature on the comovement puzzle. As a distinguishing feature to previous work, we highlight the importance of the financial channel. It helps not only matching the empirical comovement between expenditure categories, but accounting for financial frictions is also crucial to resemble the empirical responses of the excess bond premium and bank equity to an unexpected monetary contraction.
Federico Di Pace works in the Bank’s Monetary Policy Outlook Division and Christoph Görtz is a senior lecturer in macroeconomics at the University of Birmingham.
If you want to get in touch, please email us at email@example.com or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.