Oil prices and economic activity.

Kirby, Simon ; Meaning, Jack

Introduction

Crude oil spot prices have fallen sharply since the summer of 2014. In the November 2014 Review, we had factored in a fall in the spot price of around US $10 per barrel to around $98 per barrel in 2015, following a temporary fall to close to $90 per barrel in the final quarter of 2014. (1) Outturns have been significantly lower than this: oil prices currently average below $50 per barrel. Figure 1 plots the assumed path for oil prices published in the November 2014 Review, together with the assumptions underpinning the forecasts presented in the World and UK sections of this Review.

There is uncertainty about the determinants of the sharp fall in oil prices, as well as how oil prices will evolve in the months and years to come. There has been a significant softening of demand, from emerging markets such as China and Brazil as well as advanced economies, in particular the Euro Area and Japan (see the World chapter in this Review for forecasts for these economies). At the same time, US oil supply, which is at its highest level since the 1970s, and the surprise increase in output from Libya last year have been identified as supply side factors.

The causes and implications of the recent oil price shock are discussed in Arezki and Blanchard (2014). In this note, we have two distinct aims. First, we decompose oil price developments into their demand and supply components via a Bayesian vector autoregression (VAR) with sign restrictions. Our analysis suggests that the recent oil price decline was set off by expanding supply but has been increasingly dominated by demand side factors, a result supported by the changes in supply and consumption estimates from the Energy Information Administration (EIA) in recent months. This insight informs our second aim, which is to illustrate the effects of recent oil price developments on global economic activity, and in particular economic activity in a number of OECD member states. In this analysis we stress the importance of temporary versus permanent oil price shocks.

[FIGURE 1 OMITTED]

A decomposition of oil price developments

We approach the question of whether supply or demand are driving oil price movements through a decomposition using sign restrictions on a standard Bayesian VAR, following Uhlig (2005). The VAR consists of two series, the 3-month moving averages of the log differences of world oil prices and world oil supply, with six lags. The restrictions applied are as general as possible--that prices and quantities move in the same direction following a demand shock and inversely following a supply shock. Once demand and supply shocks are identified, we follow Chadha, Corrado and Sun (2010) in creating a counterfactual historical series with the effects of either supply or demand shocks stripped out. This allows us to give an historical decomposition of the impact of each shock type on the raw series, reported in figure 2. (2)

Looking at what this decomposition implies about recent history is quite interesting. First, it confirms the result of Barsky and Killian (2004) who identify the fall in oil prices around the Asian crisis of 1997-8 as mostly emanating from an inward shift in the demand curve. Second, the drop in oil prices in the financial crisis of 2007-8 was also dominated by demand shocks as global activity slowed dramatically. Conversely, the price increases in 2010 and 2011 were driven in part by strong demand as the world economy began to recover from the Great Recession, reinforced by falling supply pushing up on prices. This ties in with the complications of oil supply in the Middle East.

[FIGURE 2 OMITTED]

Regarding the most recent fall, it appears to have been initiated by an expansion of supply, which weighs down on the price from around the second quarter of 2014. Commentators have highlighted the surge in supply from US shale and the recovery of Libyan supply as potential explanations. Although this supply effect persists into the more recent period, the negative price effect is now dominated by substantial negative demand shocks, none more so than in the final quarter of last year.

Our VAR analysis is supported by a comparison of the EIA's estimates and forecasts between October 2014 and January 2015. Over this period outturns for oil prices were substantially lower than the EIA forecast in October and looking at how they have revised their estimates and forecasts of supply and consumption in light of this reveals what they implicitly consider to be the driving factors. Viewed from January 2015, the first half of 2014 is now estimated to have experienced greater supply than was estimated back in October. Given the lag structure of our VAR analysis, this coincides well with the identified supply shocks weighing down on prices from the second quarter of the year. The EIA's estimates of consumption are relatively unaffected in the first six months of 2014, but from June have been revised significantly lower in a number of months, again echoing the findings of our VAR work that demand became an increasingly important part of the changing narrative on prices.

[FIGURE 3 OMITTED]

Looking at 2015, the EIA have revised their estimates for both supply and consumption by roughly equal amounts; the former up and the latter down. Both these shifts act to lower expected prices as they simultaneously act to increase the extent of excess oil supply. However, the extent to which it is supply or demand which most influences prices will depend on the elasticities of the price to each shock. Intuitively, the large contribution of demand shocks in our VAR analysis coupled with the roughly balanced revisions to supply and demand by the EIA implies that the oil price is more sensitive to shifts in demand, an assertion supported by the analysis of Killian (2010), but which needs further investigation before it can be claimed to hold true definitively in the current context.

In summary, our empirical analysis presented here suggests that there has been a significant role for both supply and demand shocks in moving the oil price in recent months, with tentative signs that the contribution from demand shocks has been growing.

Illustrating the output effects of the recent oil price shock

As noted above, the recent shock to oil prices appears to be a combination of demand and supply shocks. In this section we depart from the concepts of demand- or supply-led oil price shocks and focus on the important distinction between temporary and permanent shocks. Separating the two is key, as temporary and permanent shocks have different implications for the impact on the real economies of oil importing nations, as well as their respective monetary policy decision-makers. We use our global econometric model, NiGEM, to illustrate the effects on advanced economies (and also on net oil importing nations) from both of these types of oil price shock. If we take our November 2014 baseline as the counterfactual, then our current baseline implies a $40-$45 fall, $20 of which is temporary and fades out after a year, and $20-$25 of which is more permanent (figure 1).

In these simulations we assume that financial markets, including the monetary policy reaction function, are forward looking. Expectations are formed consistent with the outturns from the respective scenario. This allows changes to financial instruments that are consistent with the shock. Exchange rates are determined by a forward-looking uncovered interest rate parity condition where risk premia are assumed constant unless directly 'shocked'. As such, exchange rates can 'jump' in response to changes in expectations about the future path of interest rates. We use our default monetary policy rule whereby interest rates adjust to deviations of the inflation rate and a nominal magnitude from their targets to ensure price stability. (3) In determining the monetary response, policymakers are assumed to look forward by eight quarters. Crucially, we assume that monetary policy authorities retain credibility through inflation expectations remaining anchored. The wage-price system is also assumed to be forward looking, where the anchored inflation expectations play an important role.

Across many advanced economies, policy rates were reduced to close to the zero nominal lower bound as monetary policymakers fought the deflationary environment of the Great Recession. These extraordinarily low interest rates have been broadly maintained through to the current period. Policy rates close to the zero lower bound constrain the ability of the monetary policy authorities to respond when faced with a disinflationary shock, such as the ones described here. (4) For the UK, we have assumed that the floor for policy rates is 50 basis points. (5) While monetary authorities do not have the ability to lower interest rates further in the near term in these scenarios, they do have the ability to keep interest rates at their extraordinarily low levels for longer. Financial market expectations of future interest rate paths have indeed adjusted to assume this type of monetary policy response over the coming months and years.

Table 1 presents the results from both permanent and temporary $20 falls in the price of oil. Table 2 presents the results from permanent and temporary $40 oil price scenarios. The two sets of scenarios provide us with some useful rules-of-thumb as we try to gauge the impact of recent oil price developments on the economic outlook.

Temporary shock

The impact of a temporary oil price shock affects oil importing economies predominantly via a positive terms of trade effect. With a two-year horizon for monetary policymakers it takes a couple of quarters for them to realise the shock is temporary and start to reverse any initial interest rate cuts. The developments are factored into financial market decision making and, as a consequence, real long-term rates barely change, leaving investment decisions by firms little different from the baseline. The short-term boost to GDP growth stems from the terms of trade effect on consumer spending. The sharp reduction in import prices passes through to consumer prices, to an extent which is country-specific and dependent on oil shares of imports as well as the composition of the consumption basket.

US consumption is more oil intensive than in the other countries listed here. In the face of a $20 temporary oil price decline, US GDP growth increases by 0.2 percentage point in the first year of the shock, followed by 0.1 percentage point in the second year. At its peak, the level of US GDP would be 0.3 per cent higher than the baseline (see table 1). A $40 shock would push the level of GDP 0.5 per cent above the baseline by year 2 of the shock. In both instances, the sharp rebound in oil prices is enough to reduce demand for a few years, as the rate of inflation shifts to positive and the central bank raises rates in advance of this, following its prescribed reaction function. In the $20 scenario these negative effects are modest, as are the interest rate movements. In the UK and Euro Area the temporary effects are very modest, barely increasing and then lowering GDP growth. In the $40 temporary scenario UK GDP growth expands by 0.2 percentage point, and then falls back as monetary policymakers adjust to offset the period of above-target inflation as the oil price snaps back by $40 towards the start of 2017. In both scenarios, the level of GDP in all these economies remains unchanged in the long run.

In most oil importing economies the exchange rates would be expected to jump down, given the expectation of a reduction in interest rates for a period, offsetting some disinflationary pressures (see Kirby and Meaning, 2014 for an illustration of exchange rate pass-through effects from NiGEM scenarios). But, as noted previously, at the current juncture oil consuming economies have little scope to cut interest rates further. Consequently, 'jumps' in exchange rates are constrained, inhibiting one of the natural mechanisms which would offset some of the impact of oil price changes on the price level. Indeed, given that oil exporting economies have more monetary space to operate in, it is not clear cut that oil importing countries would necessarily experience a depreciation in their effective exchange rates. This is the situation for the Euro Area, where the effective exchange rate increases, exacerbating the disinflationary impact of oil price developments.

Permanent shock

The initial effects in the permanent scenarios are similar. The positive effect on GDP is enhanced by the shift in relative prices of a key factor input (oil). The broad downward trend in the oil intensity of output for advanced economies suggests that the boost to the potential output as a consequence of permanent oil price falls will be of lesser magnitude than in preceding decades. Nevertheless, we would expect it to have a permanent positive effect on GDP. In the scenarios presented it takes a few years for inflation rates to return to baseline. Over this period, real wages are allowed to rise above baseline, given that the cost of the average consumption basket will have fallen relative to nominal incomes. Given the zero lower bound, nominal interest rates are not able to drop as much as the reaction function would suggest. (6) However, they do remain at their current low levels for a sustained period of time. This is exactly how financial markets' expectations have adjusted in recent months. The lowering of the profile for interest rates results in a jump down in the exchange rate, of a greater magnitude than in the temporary scenarios. This induces a rise in exports, relative to baseline. However, this effect is somewhat dampened in economies that export more to oil exporting economies. It should be noted that the positive effect on GDP growth lasts, on average, only three years, before the new equilibrium level of output is reached.

The effects on the US and Japanese economies are far greater than in the Euro Area or UK, due mainly, but not exclusively, to differences in the relative oil intensity of output. For example, our estimates suggest that US output is more than twice as oil intensive as UK output. A permanent $20 decline in oil prices would be expected to increase US GDP by almost 1 1/2 per cent in the medium to longer term. In the first year GDP growth would be expected to be as much as 0.6 percentage point higher in the first year. In the $40 scenario GDP growth would be expected to be around 1.2 percentage points higher. In reality, we have revised up our forecast for US GDP growth in 2015 by 0.3 percentage point, the offset being mainly the sharp appreciation of the dollar against a trade-weighted basket of currencies. These results suggest a medium to long-run effect on UK GDP of around Vi per cent in a $20 shock and just over 1 per cent in a $40 shock. The impacts on the Euro Area are significantly larger, at around 0.8 and 1.6 per cent respectively. As we note below, there are important factors likely to inhibit significantly the effectiveness of the stimulus from oil prices in the Euro Area.

A decline in oil prices is typically viewed as a positive stimulus to global economic growth because it redistributes incomes from oil producers to oil consumers, the latter typically economies with a greater propensity to spend than the former. But we are not in 'normal times' at present.

The results presented here probably represent upper bounds to the effects of the current oil price shock. Much of the Euro Area suffers from a deficiency of demand. While oil price developments should stimulate demand across the Euro Area, heightened uncertainty within the household and corporate sectors may limit the degree to which normal transmission mechanisms function. Increases in private demand are the mechanism by which economies adjust to a disinflationary oil price shock. If this mechanism is significantly impaired, we should not expect to see the positive effects on GDP reported in tables 1 and 2.

The disinflationary pressures from this oil price shock also have the potential to exacerbate deflationary conditions in oil consuming economies, in particular within the Euro Area. Of acute concern is whether the pass-through from the decline in oil prices will cause the anchor on inflation expectations to drift, exacerbating the current travails of the Euro Area.

REFERENCES

Arezki, R. and Blanchard, O. (2014), 'Seven questions about the recent oil price slump', IMF direct--The IMF Blog, 22 December, available at: http://blog-imfdirect.imf.org/2014/12/22/seven-questions-about-the-recent-oil-price-slump/ [verified 1st February 2015].

Barrell, R. and Kirby, S. (2008), 'Oil prices and growth', National Institute Economic Review, 204, pp. 39-42.

Barrell, R. and Pomerantz, O. (2004), 'Oil prices and the world economy', Focus on European Economic Integration, Oesterreichische Nationalbank, 1/2004.

--(2008), 'Oil prices and world inflation', National Institute Economic Review, 203, pp. 31-4.

Barsky, R. and Kilian, L. (2004), 'Oil and the macroeconomy since the 1970s', NBER Working Paper No. 10855.

Chadha, J., Corrado, L. and Sun, Q. (2010), 'Money, prices and liquidity effects: separating demand from supply', Journal of Economic Dynamics and Control, 34, 9, pp. 1732-47.

Hamilton, J.D. (2011), 'Historical oil shocks', NBER Working Paper No. 16790.

Kirby, S. and Meaning, J. (2014), 'Exchange rate pass-through: a view from a global structural model', National Institute Economic Review, 230, pp. F59-64.

Uhlig, H. (2005), 'What are the effects of monetary policy on output? Results from an agnostic identification procedure', Journal of Monetary Economics, 52, 2, March, pp. 381-419.

NOTES

(1) The average of Brent and Dubai spot prices.

(2) One caveat of this technique is that it requires the assumption that supply and demand shocks are uncorrelated.

(3) NiGEM has considerable flexibility in the choice of the monetary policy rule to be used for each economy in any shock. See Barrell and Pomerantz (2004, 2008) for a discussion of the effect of different reaction functions during an oil price shock.

(4) The Swiss and Danish central banks are currently testing the validity of this assumption.

(5) See the MPC minutes from the June 2012 meeting for a discussion of issues related to reducing Bank Rate below 0.5 per cent, available at http://www.bankofengland.co.uk/publications/ minutes/Documents/mpc/pdf/2012/mpc1206.pdf.

(6) We assume the lower bound acts as an effective constraint on policy and further asset purchases either are not carried out or are ineffectual at significantly loosening policy.

Simon Kirby * and Jack Meaning **

* NIESR and Centre for Macroeconomics. E-mail: s.kirby@niesr.ac.uk. ** NIESR. E-mail: j.meaning@niesr.ac.uk. Thanks to Graham Hacche, lana Liadze, James Warren, Oriol Carreras and Rebecca Piggott for helpful comments and discussions.

Table 1. Effects of a $20 decline in oil prices on the level of GDP
(per cent difference from baseline)

         Euro Area        France          Germany         Japan

       Perm.   Temp.   Perm.   Temp.   Perm.   Temp.   Perm.   Temp.

2015    0.3     0.1     0.4     0.2     0.4     0.2     0.6     0.3
2016    0.7     0.2     0.7     0.3     0.7     0.3     0.9     0.5
2017    0.8     0.0     0.8     0.0     0.7     0.0     1.0     0.2
2018    0.8    -0.1     0.7    -0.2     0.6    -0.2     1.0     0.0
2019    0.8    -0.1     0.7    -0.2     0.6    -0.2     1.0     0.0

            UK              USA            OECD

       Perm.   Temp.   Perm.   Temp.   Perm.   Temp.

2015    0.2     0.1     0.6     0.2     0.4     0.1
2016    0.4     0.1     1.3     0.3     0.9     0.2
2017    0.5    -0.2     1.4     0.1     1.0     0.1
2018    0.5    -0.2     1.4    -0.1     1.1    -0.1
2019    0.5    -0.1     1.4    -0.2     1.1    -0.1

Source: NiGEM simulations.

Note: Temporary shock lasts for eight quarters; table shows the
cumulative impacts on the level of the volume of GDP.

Table 2. Effects of a $40 decline in oil prices on the level of GDP
(per cent difference from baseline)

         Euro Area        France          Germany         Japan

       Perm.   Temp.   Perm.   Temp.   Perm.   Temp.   Perm.   Temp.

2015    0.7     0.3     0.8     0.4     0.8     0.4     1.1     0.5
2016    1.4     0.4     1.5     0.5     1.4     0.6     1.9     1.1
2017    1.7     0.0     1.6    -0.1     1.5     0.0     2.0     0.4
2018    1.7    -0.3     1.5    -0.5     1.4    -0.4     2.0    -0.2
2019    1.6    -0.2     1.4    -0.4     1.3    -0.4     2.1    -0.1

            UK              USA            OECD

       Perm.   Temp.   Perm.   Temp.   Perm.   Temp.

2015    0.4     0.2     1.2     0.2     0.9     0.2
2016    0.8     0.1     2.6     0.5     1.9     0.4
2017    1.1    -0.4     3.0     0.1     2.2     0.0
2018    1.1    -0.5     3.0    -0.4     2.3    -0.3
2019    1.1    -0.3     2.9    -0.5     2.3    -0.3

Source: NiGEM simulations.

Note: Temporary shock lasts for eight quarters; table shows the
cumulative impacts on the level of the volume of GDP.