Fernando Eguren-Martin and Andrej Sokol
What’s the relationship between financial conditions and risks to growth in an economy? And, in a world of highly integrated financial markets, to what extent are these “local” risks rather than reflections of global developments? In this post we offer some tentative answers. Financial conditions, measured across a broad range of asset classes and countries, display an important common component reflecting global developments. Loose financial conditions today increase the likelihood of a growth boom over the following few quarters, but when global financial conditions are loose, they increase the chances of a sharp contraction further ahead, highlighting some of the challenges of managing risks to growth across time from a policy maker’s perspective.
Financial conditions and growth prospects
The notion that financial conditions today can help predict future growth is both plausible and well-established. Financial markets process and price an enormous amount of information about economic conditions and prospects, and this is reflected in the ability of asset prices to help forecast future growth. Until a few years ago, as elsewhere in economics, the focus was mainly on the predictability of average future growth (see here, here or here). However, recent advances in the literature have highlighted the ability and necessity to go beyond average growth forecast to focus on the whole distribution of future growth, or, in other words, on risks.
One popular concept in this area is that of GDP-at risk (or growth-at-risk), inspired by the notion of value-at-risk (VaR) in the finance literature. Intuitively, GDP-at-risk marks particularly rare and bad growth outcomes in a forecast (say, growth outcomes that should occur once every 25 years or less). In technical terms, it is nothing but a low percentile of a GDP growth forecast distribution (usually somewhere between the 1st and 10th percentile). Two key recent insights are that GDP-at-risk can depend on current financial conditions (see also here), and that this relationship can change significantly depending on the forecast horizon, in analogy with the older idea (and historical record) that credit-induced booms can sow the seeds of future busts and growth slowdowns.
Our contribution is to explicitly link the notion of GDP-at-risk to another important strand of economic literature, that on the so-called “global financial cycle” (but see here for a dissenting view). This is the idea that financial prices and quantities share an important common component across asset classes and countries, and that therefore financial conditions in a country are always at least partly a reflection of global developments. We find, in the language of GDP-at-risk, that when global financial conditions are loose, GDP-at-risk improves in the near term (i.e. the corresponding percentile of the growth distribution becomes less negative), but worsens two to three years ahead (indicating a greater cost of a tail realisation, or equivalently, a greater chance of a “bad” growth outcome). This goes in line with earlier findings that global credit booms can help predict domestic busts and recessions. In contrast to our global results, we find that the effect of country-specific financial conditions on GDP-at-risk is both smaller and limited to the near-term. Our results also chime with a recent paper suggesting that global financial shocks are an important driver of local GDP growth, whereas local financial conditions play only a small role.
Measuring financial conditions
The first challenge when studying the relationship between financial conditions and risks to economic growth has to do with measurement of financial conditions. One could think of good reasons to focus on particular variables, such as sovereign interest rates or corporate spreads. However, in this post we avoid choosing a particular variable, and instead construct country-by-country financial condition indices (FCIs), a summary measure that extracts common variation in eight market-based variables at monthly frequency. The index aims at measuring the ease with which finance can be accessed. We compute FCIs consistently for 43 advanced and emerging economies from 1995 to 2018.
Figure 1: Financial Condition Indices
Figure 2: Global FCI
Figure 1 shows our FCIs for a selection of countries; the indices are standardised, and a higher value means tighter financial conditions. It is evident from the plot that the degree of co-movement across countries is quite high (as highlighted here). Motivated by this high correlation, we compute a global FCI as the simple average of our 43 country-specific series (Figure 2). An alternative would be to use PPP weights for aggregation, but results do not change substantially. The share of variation in individual countries’ series explained by the global component varies across countries, but it is typically high: it averages approximately 30%, and goes up to around 65% for countries like the US or the UK. As noted above, these stylised facts are consistent with the notion of a “global financial cycle”. In a second stage, we then use this global FCI to decompose individual countries’ FCIs into a global component and country-specific one that is uncorrelated with the global component. These two components will be the driving variables of our analysis.
Financial conditions and GDP growth
Armed with a global FCI and a series of country-specific FCIs, we turn to analysing their influence on risks to domestic economic growth by using the idea of “GDP-at-Risk” described above. In order to do so we follow a growing number of academic papers and policy contributions that rely on quantile regressions.
The quantile regression framework allows to model in a tractable way the whole forecast distribution of a variable, in our case GDP growth, given a set of explanatory variables. In particular, we regress a range of percentiles of the distribution of GDP growth at various horizons on global and domestic FCIs. The resulting coefficients tell us by how much do the different percentiles of the distribution of GDP growth change in response to a one standard-deviation change in the FCIs. Given our interest in downward risks to growth, that is, the “left” tail of the distribution, we will focus on the lower percentiles. For example, Figures 3 and 4 show the effects of global and domestic FCIs on the 5th percentile of the distribution of domestic GDP growth forecasts at various horizons for the average country (the swathe shows the interquartile range of individual country estimates). This will be our measure of “GDP-at-Risk” (GaR).
Figure 3: Effect of Global FCI on GaR
Figure 4: Effect of Domestic FCI on GaR
These results yield a number of insights. The most direct conclusion is that a tightening in financial conditions leads to a deterioration in short-term GDP-at-Risk; that is, it pushes the left tail of the distribution of GDP growth down. However, the source of this tightening seems to matter: a one standard-deviation change in global financial conditions has a much larger impact on GDP than a comparable change in domestic financial conditions. One explanation is that this could be the result of spillovers from effects in the rest of the world, as a change in global financial conditions would affect all countries, albeit to different degrees. Another distinct feature of changes in global financial conditions is that the impact is horizon-specific: while the short-run impact is negative as described above, the sign flips at longer horizons. This means that tighter global conditions mitigate GDP-at-Risk at longer horizons, or, alternatively given the symmetry of our exercise, loose global financial conditions spur short term growth at the cost of increasing longer-term risks. This effect is not apparent in the case of domestic financial conditions.
Some implications for policy
Financial conditions contain useful information about risks to the outlook for GDP growth. Tighter financial conditions increase downside risks in the short run, and vice-versa for loose conditions. Additionally, global financial conditions have very different effects in the short and long run, with upside risks from loose financial conditions turning into increased chances of a growth slowdown further down the line. There are at least two lessons for the conduct of any policy that has amongst its objectives to reduce risks to growth: first, the source of the shock matters. Changes in global financial conditions have a much more sizeable impact on risks to domestic growth for the average country, plausibly because the shock affects all countries at once. Second, as already pointed out by earlier analyses, there’s a trade-off in managing risks to domestic growth stemming from global financial conditions, with a price to pay in terms of near-term growth in order to reduce risks of bad outcomes further down the line.
Fernando Eguren-Martin works in the Bank’s Global Analysis Division and Andrej Sokol contributed to this post while working in the Bank’s International Directorate, and is currently on secondment to the European Central Bank.
If you want to get in touch, please email us at firstname.lastname@example.org 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, or the European Central Bank .