Fergus Cumming and Paul Hubert
How does the transmission of monetary policy depend on the distribution of debt in the economy? In this blog post we argue that interest rate changes are most powerful when a large share of households are financially constrained. That is, when a higher proportion of all borrowers are close to their borrowing limits. Our findings also suggest that the overall impact of monetary policy partly depends on the behaviour of house prices, and might not be symmetric for interest rate rises and falls.
From Micro to Macro
In a recent paper we use the universe of UK loan-level mortgage data to construct an accurate measure of the proportion of households that are close to the limits of what banks will lend them. Our rich mortgage data allow us a rare glimpse into the various factors that drove individual debt decisions between 2005 and 2017. After stripping out the effects of regulation and other developments in the macroeconomy, we then estimate the proportion of highly indebted households to construct a measure that is comparable across time (Figure 1). In doing so, we collapse the information contained within 11 million mortgages into a single time series, which allows us to explore macro questions in a tractable way.
Figure 1 – The share of conditional-LTI> 4 and LTI>5 mortgages
We use this state variable as a proxy for the share of people that are financially constrained, and use its variation over time to explore how it affects the economy’s responsiveness to monetary policy. In particular, we focus on the response of aggregate consumption when we interact it with a set of standard monetary shocks using local projection methods. This involves running regressions of consumption on monetary shocks at different horizons. We interact them with a variable that captures the proportion of highly indebted households, to see if the strength of response varies with the proportion of indebted highly households, as well as other controls.
These yield a set of impulse responses that tell us how the path of consumption might respond to an unexpected increase in the monetary policy rate. We can then take two different values of our state variable (1 standard deviation above and below the mean) to compare the path of consumption following a change in monetary policy when the share of highly indebted households is above, or below, its historic average. Intuitively, we know that a contractionary monetary shock leads to a near-term dip in consumption, which is why monetary policymakers tend to increase interest rates when they want to apply the brakes. We explore whether this pattern changes according to what else is going on, including whether more-people-than-normal are financially stretched.
State-Contingent Monetary Policy
We find that monetary policy is more powerful when a large proportion of households have taken on relatively high debt burdens. In that sense, the transmission of monetary policy depends on the state of the economy. In Figure 2 below, we look at the response of non-durable, durable and total consumption in response to a 100bp increase in interest rates. The grey (blue) swathes plot the path of consumption when there is a large (small) share of people who might be more constrained by higher debt holdings. The gaps between the blue and grey swathes suggest monetary policy is more potent when more people have higher debt levels.
Figure 2 – The response of consumption to 100bp contractionary monetary policy shock
This differential effect likely works through at least two mechanisms: first, when households increase their borrowing relative to their income, the mechanical effect of monetary policy on disposable income is amplified. Those with a large borrowings suffer the most from a proportional increase in monthly repayments! Second, households close to their borrowing limits adjust spending more in response to income changes (they have a higher so-called marginal propensity to consume). If there are more of these people, overall consumption will respond more to monetary policy. Interestingly, we find that our results are driven more by the distribution of highly-indebted households rather than a general rise in borrowing.
We also find evidence of some asymmetry in the transmission of monetary policy. When the share of constrained households is large, interest rate rises have a larger absolute impact than interest rate cuts. To some extent this is unsurprising. When your income is very close to your outgoings, the effect of a small income squeeze is very different to receiving a small windfall. So it follows that interest rate increases are likely to affect behaviour more than interest rate falls.
Our results also suggest that the behaviour of house prices affects how monetary policy feeds through. When house prices rise, homeowners feel wealthier and are more able to refinance their mortgages and release housing equity in order to spend money on other things. This can offset some of the dampening effects of an increase in interest rates. In contrast, when house prices fall, this channel means an increase in interest rates has a bigger contractionary effect on the economy, making monetary policy more potent.
Finally, we find that the household debt distribution also affects not just consumption responses, but also how monetary policy feeds through to the economy more broadly. Running the same exercise to gauge the effects on wages, employment, real output growth and industrial production, we find a similar result: responses are stronger when more households are highly indebted. This is likely driven by the direct effect of consumption, but also an amplification from spending through to changes in firm and labour market behaviour – or what economists call general-equilibrium effects.
Policy Implications
Our results suggest that the potency of monetary policy is not always the same. It can vary with respect to economic conditions. In particular, we show that the distribution of household indebtedness might explain some of the variation in the effects of monetary policy. It’s not so much the average level of indebtedness that matters, but rather its distribution. The greater the proportion of indebted households, the bigger the effect it has.
Economists, central bankers and other policymakers often look at the role of debt from a financial stability or regulatory perspective — how it affects the stability of individual institutions and the financial system as a whole. Our results suggest another reason why debt matters — it affects how monetary policy changes are transmitted through the economy. This is an emerging area of research, facilitated by new detailed micro-level datasets, and there’s much more for the profession to do on this topic.
Fergus Cumming works in the Monetary Policy Outlook Division and Paul Hubert works at Sciences Po – OFCE.
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk 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.
The allocation of bank credit to the real economy, one of the most important transmission channels of central banks’ monetary policy, is totally distorted by risk weighted bank capital requirements, based on expected perceived or decreed credit risk
http://subprimeregulations.blogspot.com/2019/07/risk-weights-are-to-access-to-credit.html
“When house prices rise, homeowners feel wealthier and are more able to refinance their mortgages and release housing equity in order to spend money on other things.”
Isn’t the wealth gain here is just someone else’s wealth loss ? Given there’s no productivity gain and therefore no net new wealth. It’s just a wealth transfer (usually the wrong way; from poorer young people to richer old people).
The debt levels are too high, which creates a systemic risk. We should not forget that consumer debt is mainly floating, which increases the risk further. If we compare contraction in spending in Denmark with the UK, contraction in Denmark is less in times of stress. The effective 4.5 LTI cap on mortgage lending is a cap that helps to protect, but to goal has to be for consumers to fixed their debt costs for the term of the exposure. Unfortunately the good of the consumer is not aligned with the good of the large mortgage lenders.
Re Per, Dan and Arjan
You are all right. Unconventional monetary policies led to nothing but bubbles in asset prices (houses, shares, etc.) and a squeeze in yields. This meant terrible redistribution of wealth from the poor to the rich (when house prices are higher, the poor would be less able to own a home). Also, to your point Arjan, most of the increase in debt was channelled to non-productive uses (massive increases in mortgages fuelling house prices, and massive bond issuances to fund share buy-backs and fuel bubbling share prices. I think QE was the worst thing central banks have ever done, and they are stuck with it forever.
Hi John, I totally agree. In my view the mistake made was not to control that asset inflation, and have in place affordability rules that restrict lending pre the financial crisis. Bank lending increases the money supply, and allowed the asset bubble to happen. While QE was bad, was there another choice? Personally I am a supporter of the approach to crystallise the QE on the central banks’ balance sheets and focus on getting a normal economy back, this low rate environment is dangerous, it makes debt free and the temptation to take on too much debt for the sake of shareholder return (or consumer consumption to have that bigger house) is great. Debt ruins companies, and lives, it should be controlled, and definitely should not be free like it is now.