Since the Global Crisis, many central banks have used balance sheet interventions – including quantitative easing, credit easing, and emergency lending assistance – to pursue their monetary policy objectives and address concerns with financial system stability. Changes in both the size and composition of central banks’ balance sheets have since become an indicator of the stance of monetary policy (e.g. Patel 2018).1 Yet, some challenges must be overcome before balance sheet data can be useful for making reliable inferences about the aim and effectiveness of central banks’ policy interventions.
Our recent study uses balance sheet data to analyse the stance of monetary policy, and asks whether balance sheet interventions create a trade-off between monetary policy objectives and financial system stability concerns (Ademuyiwa et al. 2018). We compile a mini-database of the major balance sheet items for over 30 central banks (20 advanced and 11 emerging), sourcing information directly from central bank websites, financial reports, and data archives. Using data from such individualised sources allows us to address analytical problems resulting from heterogeneity in reporting standards and differences in the level of transparency across countries.
This column provides some stylised facts on the evolution of balance sheet items over the past 18 years. The data exhibit a pattern of portfolio rebalancing – between private and public assets – during and after periods of financial stress. We then investigate the links between changes in balance sheet items and central banks’ ability to deliver desirable macroeconomic outcomes in terms of inflation-output variability. We find that balance sheet interventions are associated with higher output variability. However, considering the limitations of balance sheet data, our result should be viewed as suggestive at best.
Stylised facts of central banks’ balance sheets
A significant challenge for researchers when using balance sheet data to evaluate the conduct of monetary policy is the heterogeneity and transparency of published data. These issues stem from at least three sources:
- international differences in accounting standards;
- institutional arrangements defining the relationship between central banks and their government; and
- the neglect of central banks to use balance sheet accounting to help their stakeholders ascertain the changing stance of monetary policy (e.g. Archer and Moser-Boehm 2013).
We try to minimise the effect of these problems by sourcing data from individual central banks, and then re-categorising these data into a consistent format. The dataset spans from 1998 to 2016, allowing for an exploration of balance sheets pre- and post-crisis.
Changes in a central bank’s assets-to-GDP ratio is a rough indicator of the size of its financial market interventions overtime. The analysis reveals large and visible increases in the size of central banks’ balance sheets during the sample period, especially in economies most directly affected by the Global Crisis – such as Japan, the US, the UK, and the euro area countries. While the crisis has no obvious impact on the balances sheets of central banks in many emerging market economies, it is generally the case that their total assets-to-GDP ratios are higher post-crisis than at almost any time prior to 2007.
For a few central banks, there is a surge of private sector loans and asset holdings as a percent of GDP around 2008, which leaves little doubt that this is due, in one way or another, to fallout from the Global Crisis. These holdings appear to have disappeared or been phased out in the post-crisis period. In some euro area countries, the ratio of private sector holdings to GDP is higher than the average, reflecting the division of responsibility between the ECB and national central banks, which have more duty for the stability of the banking sector in their national jurisdictions.
The data reveal evidence of portfolio adjustments by major central banks during financial crises due to active balance sheet interventions – holdings of private sector assets rise in periods of financial stress, which is followed by an increase in the share of assets held in public debt. This trend is evident across most central banks in the sample. The lack of detailed data makes it impossible to further break down the government assets into short and long-term bond holdings across countries.
For almost all emerging market economies in our sample, we observe increasing holdings of gold and foreign currencies as a percent of total assets. However, this trend predates the Global Crisis and is likely related to the fallout from the Asian Financial Crisis of 1997–1998. Finally, there has been interest in the potential insolvency of central banks since the crisis. While there is considerable heterogeneity across countries, for 25 of the 31 central banks in our sample, their ratio of net worth to total assets is at least 20% by the end of the sample, but much higher in many of these cases.
Some empirical evidence on the monetary policy trade-off
Monetary policy effectiveness could be framed as a trade-off between inflation and output volatility (Taylor 1979,1994).2 The ‘optimal’ choice is for policymakers to set monetary policy such that both variabilities are minimised but positively related. We find that this trade-off can be observed in 21 of the 31 countries in our sample.
If balance sheet interventions are also used to address concerns over financial system stability, there is a risk that these interventions come at the cost of inflation-output variability. We investigate the potential connection between the composition of central banks’ balance sheets and the variability of inflation and output growth over time. The regression estimates include many indicators of balance sheet composition, described above, and control for proxies of global and domestic financial system stability.
The estimates show that while none of the balance sheet components has a statistically significant relationship with inflation variability, more than half of the balance sheet items appear to have a positive, albeit small, relationship with output variability. Other things equal, this implies that central banks’ intervention via balance sheet policies can potentially worsen the trade-off implied by standard macroeconomic models. However, considering the limitations of balance sheet data, our result should be viewed as suggestive at best. Furthermore, balance sheet interventions tend to increase during periods of economic stress, which are typically characterised by higher output variability. More informative data is indeed needed to better understand the connection between balance sheet interventions and their consequences for the inflation-output variability trade-off.
Monetary policy interest rates alone no longer serve as a sufficient representation of the stance of monetary policy, especially when they are at very low levels. The use of unconventional monetary policies, including balance sheet policies and forward guidance, will likely become more frequent in the future. Therefore, ongoing scrutiny of central banks’ balance sheets by its stakeholders is a welcome development towards improving central bank transparency and accountability. In addition, for small open economies and emerging market economies, global financial conditions can affect the composition of the central bank’s balance sheet. More detailed and consistent accounting in central bank’s balance sheets could be used, in cooperation with other indicators, to build a measure of the monetary policy stance, and to identify the use of policy interventions for the purpose of financial system stability.
Despite the potential usefulness of balance sheet data, several challenges must be overcome before we have adequate and reliable data to analyse the aims and effectiveness of central bank balance sheet interventions. A global effort to improve the availability, timeliness, and quality of central bank balance sheet data is needed in order to improve the accountability of central bank interventions and promote robust empirical research. These efforts should include establishing internationally consistent balance sheet accounting principles for central banks, and identifying separate components of the balance sheet targeting monetary policy and financial stability policy. Generating higher quality balance sheet data should go hand in hand with efforts to improve both policymakers’ and the public’s understanding of the assets and liabilities of monetary authorities, thus strengthening their transparency and accountability.
Ademuyiwa, I, P L Siklos and S St Amand (2018), “Central bank balance sheets and the interaction between monetary policy and financial stability,” CIGI Paper, forthcoming.
Archer, D and P Moser-Boehm (2013), “Central bank finances,” BIS Papers, No 71.
Patel, U (2018), “Emerging markets face a dollar double whammy,” Financial Times, 3 June.
Taylor, J B (1979), “Estimation and control of a macroeconomic model with rational expectations,” Econometrica 47(5): 1267–86.
Taylor, J B (1994). “The inflation/output variability tradeoff revisited,” In J C Fuhrer (ed), Goals, Guidelines, and Constraints Facing Monetary Policymakers, The Federal Reserve Bank of Boston Conference Series 38: 21–38.
 Balance sheet data is of course one of several important indicators of monetary policy, as many policy tools—such as interest rate setting and forward guidance—would not be fully captured by this data.
 In Taylor’s setup the policymaker is concerned with both inflation and output outcomes. Hence, it is easy to see how the trade-off referred to eventually produced the specification of what has come to be called the Taylor rule, where policymakers must decide how much weight to place on inflation versus output outcomes.