The evolving effectiveness of UK’s monetary policy

Colin Ellis, Haroon Mumtaz, Pawel Zabczyk 06 August 2014

a

A

Over the past five decades, major industrialised economies underwent deep structural changes. These typically included dramatic shifts in macroeconomic policy and globalisation-induced changes in competition, technological advances, and financial innovation. This raises several concerns for policymakers, including whether the channels through which monetary policy affects the economy have changed over time, and what that might mean for how policy should be conducted.

Accounting for changes in the transmission mechanism

While a number of studies (Boivin and Giannoni 2002, 2006, Gali et al. 2003, Primicieri 2006, Canova and Gambetti 2009, Gambetti and Gali 2009) have analysed the stability of various facets of US monetary policy transmission, no clear consensus has been reached. The arguments in Bernanke et al. (2005) and Hansen and Sargent (1991) suggest that the conflicting evidence may be due to simple empirical models missing data necessary to correctly identify structural shocks. Boivin et al. (2010) additionally claim that the split-sample estimation strategy used in some of the US studies could be another factor responsible for the lack of consensus. This is because such a procedure may fail to allow for sufficiently rich dynamics of changes in the monetary policy transmission mechanism.

Ample evidence of widespread changes in the UK economy (Benati 2004, Mumtaz and Surico 2012) meant that in Ellis et al. (2014) – where we studied the UK transmission mechanism – we had to tackle both of these issues head on. To avoid relying on a limited set of data, we therefore followed Bernanke et al. (2005) and Boivin et al. (2010) and augmented our model with factors extracted from 350 data series. To further allow for the possibility of changes in the way various types of disturbances are transmitted, we additionally extended Bernanke et al. (2005) by allowing for time-varying coefficients in the vector autoregressive (VAR) transition equation and for heteroscedasticity in shocks, capable of accommodating variations in the volatility of the underlying series.

In our study, we identified unanticipated changes (so-called ‘shocks’) in UK’s demand, supply, and monetary policy, with Figure 1 plotting their respective volatilities.

Figure 1. Volatility of identified structural shocks

In terms of historical context, Thatcher’s disinflation – launched in 1979 – and the experiment with monetary targeting (ultimately abandoned in 1985), appear to have coincided with dramatic falls in the volatility of supply shocks. And although the volatility of monetary policy and demand shocks continued to decrease until the end of our sample period, supply shocks do seem to have become temporarily more volatile around the landslide Labour party election victory in 1997, which preceded the granting of operational independence to the Bank of England. Importantly, the fact that all three volatilities appear markedly lower in the 1990s than in the 1970s, serves as another illustration of what is often termed the ‘Great Moderation’.

Enhanced effectiveness of UK’s monetary policy: New evidence

The time-varying impulse responses to monetary policy shocks (normalised to increase the Bank Rate by 100 basis points on impact) reported in Figure 2 provide evidence that prior to 1992, there was hardly any impact on various measures of inflation from policy becoming unexpectedly tighter. Since that time, however, the response of inflation to the same policy tightening has become more negative and persistent. Importantly, the 1992 threshold, which broadly coincides with the introduction of inflation targeting, is not hard-wired into the model, but naturally emerges as one consequence of allowing for time-variation in a flexible fashion.

Figure 2. Impulse response of inflation to a monetary policy shock

Note: The left panels present the time-varying median cumulated impulse response. The three central panels show the average impulse response functions (IRFs) in the pre- and post-1992 periods as well as their difference, while the right panel shows the joint distribution of the cumulated response at the one-year horizon in the pre- and post-1992 periods.

The natural question then is: What do these impulse responses imply about changes in the transmission of monetary policy? To help identify the channels through which the increased efficacy of policy shocks on prices may have occurred, we inspected changes in the corresponding IRFs of asset prices (Figure 3). Using the response of yields on a 10-year government bond as a measure of the responsiveness of inflation expectations to policy, we found that long-term yields clearly responded by more post-1992 (fourth chart in the bottom row of Figure 3), which appears consistent with policy becoming better at exploiting the expectations channel.

Equity prices are important for two channels of monetary policy transmission.

  • First, as in the classic q theory, lower prices would imply that capital replacement costs are high relative to the market value of firms, with negative implications for investment. Furthermore, lower equity prices push down on the net worth of firms, potentially exacerbating adverse selection and moral hazard problems, and may lead to a decline in lending, spending, and aggregate demand (as discussed in Boivin et al. 2010).

While the FTSE does respond in line with the theory, the upper bound on the median difference is close to zero, suggesting little reason to expect considerably greater impact of policy on demand via either of these channels.

One of the more interesting pieces of evidence on changes in monetary policy transmission can be found in the first row of Figure 3.

  • On the one hand, expectations can have an important effect on the cost of capital and housing spending, suggesting that house prices should contract by more in response to negative policy shocks after 1992.
  • On the other hand, demutualisation of many building societies in the 1990s, coupled with new financial produces, may have helped to insulate consumers from the immediate impact of changes in prices.

Our results point to no change in the response of house prices to policy shocks, suggesting that these factors may have had offsetting effects.

Figure 3. Impulse response of asset prices to a monetary policy shock

Note: The left panels present the time-varying median cumulated impulse response. The three central panels show the average impulse response functions in the pre- and post-1992 periods as well as their difference, while the right panel shows the joint distribution of the cumulated response at the one-year horizon in the pre- and post-1992 periods.

In Ellis et al. (2014) we also report the response of real activity to policy shocks. In line with the US-based evidence of Boivin et al. (2010), who report ‘modest’ short-run elasticities, we find that the response of UK’s investment to policy shocks is imprecisely estimated and displays no sign of changes over time. Additionally, and in contrast to the aggregate inflation series, the responses of consumption, industrial production, and real GDP appear relatively similar pre- and post-1992, with the intervals around the IRFs consistent with no time variation.

Implications for policy and conclusion

Both our main result and the way we interpret it are worth putting in context. First, unlike Boivin and Giannoni (2006), we do not augment the analysis by using a dynamic stochastic general equilibrium (DSGE) model built to match aspects of the VAR-based evidence. While this reduces the risk of contaminating the conclusions through misspecification of either of the two models, it also means that we can only focus on unanticipated changes in monetary policy. Alternatively, we have no direct evidence on potentially very important changes in the systematic component of policy, and we proceed by equating policy surprises with discretionary deviation from some underlying and possibly time-varying policy rule.

Interestingly, the findings we report in Ellis et al. (2014) suggest that the widespread practice of ignoring time-variation doesn’t necessarily lead to wrong conclusions. Specifically, there appears to be little evidence of the transmission of demand or supply shocks changing over time. Crucially, however, this does not appear to be the case for unanticipated changes in monetary policy, suggesting that central bankers do need to take account of how economic relationships change over time, or risk making significant policy errors.

On the whole, our results reinforce the view that credible monetary policy has a clear role to play in anchoring the economy. The fact that since the advent of inflation targeting, unanticipated changes in policy started having a stronger impact on underlying inflationary pressures is particularly encouraging as containing these pressures – rather than short-lived relative price shocks – should be the key focus for policymakers.

As discussed, our findings are consistent with the main impact of UK’s inflation targeting having been transmitted via the anchoring of inflation expectations, which reduced the costs associated with achieving low and stable inflation rates. This anchoring may have been particularly important in recent years, as households have endured the painful reduction in real wages associated with the financial crisis. These tangible benefits suggest that policy should respond very strongly if there are any signs that the credibility of the inflation target is starting to erode.

References

Benati, L (2004), “Evolving post-World War II U.K. economic performance”, Journal of Money Credit and Banking, vol. 36(4), pp. 691-717.

Bernanke, B, Boivin, J and Eliasz, P S (2005), “Measuring the effects of monetary policy: A factor-augmented vector autoregressive (FAVAR) approach”, The Quarterly Journal of Economics, vol. 120(1), pp. 387-422.

Boivin, J and Giannoni, M (2002), “Assessing changes in the monetary transmission mechanism: A VAR approach”, Economic Policy Review, (May), pp. 97-111.

Boivin, J and Giannoni, M P (2006), “Has monetary policy become more effective?”, The Review of Economics and Statistics, vol. 88(3), pp. 445-462.

Boivin, J, Kiley, M T and Mishkin, F S (2010), “How has the monetary transmission mechanism evolved over time?”, in (B.M. Friedman and M.Woodford, eds), Handbook of Monetary Economics, vol.3, pp.369-422, San Diego: Elsevier.

Canova, F and Gambetti, L (2009), “Structural changes in the US economy: Is there a role for monetary policy?”, Journal of Economic Dynamics and Control, vol. 33(2), pp. 477-490.

Ellis, C, Mumtaz, J and Zabczyk, P (2014), “What lies beneath? A time-varying FAVAR model for the UK transmission mechanism”, Economic Journal, vol.124(576), pp.668-699.

Gambetti, L and Gali, J (2009), “On the sources of the Great Moderation”, American Economic Journal: Macroeconomics, vol. 1(1), pp. 26-57.

Gali, J, Lopez-Salido, J D and Valles, J (2003), “Technology shocks and monetary policy: assessing the Fed's performance”, Journal of Monetary Economics, vol. 50(4), pp. 723-743.

Hansen, L P and Sargent, T J (1991), “Two problems in interpreting vector autoregressions”, in (R.E. Lucas and T.J. Sargent, eds.), Rational Expectations Econometrics, Boulder: Westview.

Mumtaz, H and Surico, P (2012), “Evolving international inflation dynamics: World and country-specific factors”, Journal of the European Economic Association, vol.10(4), pp. 716-734.

Primiceri, G (2006), “Why inflation rose and fell: Policymakers' beliefs and U.S. postwar stabilization policy”, The Quarterly Journal of Economics, vol. 121(3), pp. 867-901.

a

A

Topics:  Europe's nations and regions Monetary policy

Tags:  inflation targeting, UK, policy shocks

Head of Analytics for the Financial Institutions Group, Moody's Investors Service; Visiting Research Fellow, University of Birmingham

Professor of Economics, Queen Mary University of London

Adviser at the Centre for Central Banking Studies, Bank of England; Associate, Centre for Macroeconomics