RDP 2023-09: Does Monetary Policy Affect Non-mining Business Investment in Australia? Evidence from BLADE 1. Introduction

Non-mining business investment in Australia was fairly weak over much of the 2010s, despite declines in interest rates and moderate economic growth (e.g. Debelle 2017; Heads of Treasuries 2017; van der Merwe et al 2018; Hambur and Jenner 2019). While several explanations have been put forward, such as risk aversion and uncertainty, or pessimism about future outcomes (‘animal spirits’), one potential explanation is that monetary policy is not very effective at stimulating business investment, or has become less effective over time. For example, survey evidence (generally for larger firms) suggests that the hurdle rates that businesses require investments to meet can be ‘sticky’ and not responsive to monetary policy (Lane and Rosewall 2015; Edwards and Lane 2021; Sharpe and Suarez 2013).

As business investment is a key driver of economic growth – both in a structural sense through its contribution to productivity, and in a cyclical sense – it is important to understand how and whether monetary policy affects business investment. The key theoretical channels are well documented. A reduction of the policy interest rate stimulates investment by: increasing aggregate demand; lowering the cost of investment (e.g. borrowing rates) – the ‘user cost of capital’; and loosening credit and financing constraints by freeing up cash flow for indebted firms or raising the value of collateral that firms can pledge to obtain loans (Bernanke and Gertler 1995). However, in practice several factors make it difficult to quantify the effect of monetary policy on investment, and to understand the transmission channels.

A key factor that makes it hard to assess the effect of monetary policy on investment is endogeneity. Softness in (current or expected) economic conditions is likely to lead to expansionary monetary policy, but also to weaker investment, biasing downwards any estimated effect of an interest rate change on investment. Indeed, a negative relationship between investment and interest rates is hard to find in aggregate time series data (e.g. Chirinko 1993; Caballero 1999; Hassett and Hubbard 2002; Cockerell and Pennings 2007). Another factor is that investment is very heterogenous. At any point in time, some businesses are investing a lot while others are not investing, reflecting business-specific conditions and past investment. This can make it hard to identify the effect of monetary policy on investment using aggregate data.

To improve our understanding of the effects of monetary policy on investment we combine the exogenous monetary policy shock measures of Beckers (2020) with quarterly firm-level tax data on the near universe of firms from ABS Business Longitudinal Analysis Data Environment (BLADE). In using firm-level data, we follow earlier work by Ottonello and Winberry (2020), Durante, Ferrando and Vermeulen (2022), Jeenas (2023) and Cloyne et al (forthcoming) in estimating the dynamic effect on investment for a period of up to 12 quarters after a monetary policy shock using a local projections model.

By using firm-level data we can also explore monetary policy transmission channels and whether monetary policy affects different types of firms differently. As well as providing additional insights into how monetary policy works, our work could yield insights into the reasons behind the low levels of non-mining investment evident throughout the 2010s. For example, Cloyne et al (forthcoming) shows that young firms might be more sensitive to monetary policy if they are more credit constrained and expansionary monetary policy alleviates these constraints. If this is the case, declining firm entry rates over the past two decades, and an associated decline in the share of young firms in the economy, could potentially have made monetary policy less effective in stimulating investment. Alternatively, some have argued that the investment of very large ‘leader’ firms is more responsive to monetary policy due to their established relationships with financial institutions, access to various fundings sources and ability to navigate economic fluctuations. If so, then expansionary monetary policy could entrench existing industry leaders and weaken competitive pressures, causing aggregate investment to be weaker than otherwise (Kroen et al 2021; Liu et al 2022). This could potentially help to explain the apparent decline in competition documented in Hambur (2023), which Hambur and Andrews (2023) link to slower investment, particularly among more productive firms.

Our key findings are that contractionary monetary policy decreases both the likelihood that firms invest (extensive margin), and the extent of investment (intensive margin). The effects are economically significant, though it is difficult to compare them to estimates from macroeconomic models given the implied persistence of the shock differs.

We find that firms of different sizes and ages respond similarly to monetary policy shocks, though the likelihood of investing is slightly more sensitive to monetary policy for smaller firms. This has several implications for policymakers:

• The decrease in the share of young firms in the economy, due to slower firm entry, is unlikely to have lowered the effectiveness of monetary policy (at least not directly).
• Recent US evidence that expansionary monetary policy disproportionately supports leader firms, and so reduces competition, does not appear to apply to Australia, at least in terms of tangible investment.
• Survey evidence that large (and some other) firms have ‘sticky hurdle rates’ should not be taken as evidence that they do not respond to monetary policy. This could reflect the fact that monetary policy can affect investment through channels other than by lowering the marginal cost of capital or potential differences between surveyed and actual behaviour.

We also find some evidence that sectors that are more dependent on external finance, and firms that are financially constrained, are more responsive to monetary policy. This highlights the important role that cash flow and financing constraints play in the transmission of monetary policy.

As well as providing Australia-specific results, our work contributes to the broader literature in several ways. First, we provide results for an advanced small open economy, whereas most of the literature has considered larger economies such as the United States (Jeenas 2023; Cloyne et al forthcoming) and the euro area (Durante et al 2022). This is relevant given in a small open economy like Australia, firms are potentially exposed to a greater range of international shocks and not needed given the global cost of capital may play an important role in their financing decisions, rather than only the domestic costs.

Second, we provide some of the first evidence on the effects of monetary policy on expected investment by combining our shock measures with survey measures on expected investment. We find that monetary policy shocks tend to affect both actual and expected investment with a similar lag, rather than influencing expectations earlier. This suggests that firms' expectations are based on current conditions, rather than being based on a clear view of what monetary policy will do to future outcomes.

The economic implications of this finding depend on whether lower expectations of investment translate into lower actual investment in the future. If so, this could be a mechanism through which the effects of economic shocks are perpetuated. If not, there may be limited real economic implications, but this mechanism could lead to cyclical variation in realisation ratios (the ratio of actual to expected investment), which would have implications for practitioners' ability to use these data for economic forecasting.

Third, our finding that monetary policy affects investment on both the intensive and extensive margin speaks to the ongoing discussion on how endogenous capital accumulation should be modelled in macroeconomic models. For example, many models, including Woodford (2005), model investment using convex adjustment costs, leading to smooth investment. This contrasts with the existing evidence (including in this paper) that investment tends to be lumpy. Sveen and Weinke (2007) model lumpy investment; however, they do so using an exogenous Bernouilli process that implies that monetary policy cannot affect the extensive margin of investment, in contrast to our findings. Rather, our results support the use of fixed adjustment costs in investment, as proposed by Reiter, Sveen and Weinke (2013).