Erik Thedéen: Higher countercyclical capital buffer?

A speech given by Erik Thedéen, FI's Director General, at the UBS Annual Nordic Financial Services Conference in Stockholm today.

  • Date: 2018-09-06
  • Speaker: Erik Thedéen
  • Meeting: UBS Annual Nordic Financial Services Conference i Stockholm. 

Both the global and the Swedish economies continue to demonstrate strong growth. This growth, combined with low interest rates and high risk-taking, has resulted in an upward trend for asset prices that has persisted for a long time - both in Sweden and globally. Swedish house prices, however, fell by 9 per cent between August and December 2017. Since then, house prices have stabilised and are currently about 3 per cent higher than in December. Despite the slow-down, house prices remain elevated, and there is still a risk that prices will fall further. Prices of commercial real estate have also increased sharply since 2015.

The growth rate of total debt has been exceeding sustainable levels for a long time. With the exception of a short period during the 2008 financial crisis, total lending to households and non-financial corporations has increased faster than nominal GDP since 2004. This growth has resulted in an increase in total debt in relation to GDP of approximately 8 percentage points since the beginning of 2016, and it now amounts to almost 160 percent. Overall, the combination of rapid lending growth and high asset prices indicates that systemic risks have increased.

One of the drivers behind the high growth rate of total debt is non-financial firms' market funding. Because market funding is not supplied by banks, the argument is sometimes raised that it should not influence how much capital banks should hold. However, Finansinspektionen (FI) sees two main reasons for why an increase in market funding can still contribute to the build-up of systemic risks.

First, higher leverage in non-financial firms makes them less resilient to shocks irrespective of whether the firm borrows from a bank or the financial markets. Less resilient firms increase the risk that banks supplying loans to non-financial firms will face credit losses should a crisis occur. Second, market funding could dry up in times of financial distress. Firms can then use more of the credit lines they already have with the banks. Banks can also substitute market funding for bank loans to maintain their customer relationships. It is therefore FI's view that an increase in market funding contributes to the build-up of cyclical systemic risks and cannot be excluded when gauging how much capital a bank should hold.

FI's assignments are to promote financial stability and take measures to counteract financial imbalances. In the presence of rising systemic risks, FI has been active in using both bank- and borrower-based measures. The bank-based measures have included high capital requirements, with a large share of usable capital buffers, and liquidity requirements to make the banking system more resilient. The borrower-based measures have included the LTV cap and the amortisation requirements to reduce the risks stemming from household debt.

The capital requirements include a countercyclical capital buffer, which is a macroprudential tool provided by the EU's Capital Requirements Directive (CRD). This tool addresses cyclical systemic risks, i.e. risks that vary over time due to, for example, excessive credit growth. By increasing the countercyclical capital buffer when risks are building up, this provides banks with an additional cushion of capital. If a crisis were to occur, the countercyclical capital buffer could be reduced or set to zero. Lowering or removing the countercyclical capital buffer in a severe downturn releases capital that can be used to absorb losses. It is also a signal to the banks that the supply of credit to households and businesses should not be contracted too much.

An increasing number of countries have positive countercyclical capital buffers in place or plan to have them in place in the near future. In Europe, seven countries have announced positive buffer rates. A Swedish countercyclical capital buffer of 2.5 per cent corresponds to about SEK 35 bn of usable capital in the four large Swedish banks by. This corresponds to about 5 per cent of their total capital requirements and about 10 per cent of their usable capital.

However, the countercyclical capital buffer only has marginal effects on lending rates. Our estimates suggest that a 2.5 per cent countercyclical capital buffer increases mortgage rates by about 5 basis points. The corresponding effect on lending rates for Swedish corporates is about 10 basis points.¹ Consequently, the dampening effects of the countercyclical capital buffer on house prices and the growth rate of credit are small.

Our view is therefore that the countercyclical capital buffer is an important tool when building resilience in the banking system since it has non-negligible effects on the amount of usable capital. While the countercyclical capital buffer also has effects on lending rates, these effects are too small to make it an efficient tool to dampen the build-up of cyclical systemic risks from excessive credit growth.

The buffer requirement generally enters into force one year after it has been decided. This makes it easier for banks and other credit institutions to adapt to a higher capital requirement by retaining a portion of their profits rather than raising new equity from the financial markets. Banks can also meet the higher capital requirements by reducing their assets, for instance by supplying less credit to households and firms. Raising the buffer too fast or too much could therefore potentially create a credit crunch. Because of this risk, it is important to start the build-up at an early stage, increase the buffer requirement gradually, and give banks sufficient time to build up the buffer.

When analysing the build-up of cyclical systemic risks, we look at a variety of indicators. The credit gap and the corresponding benchmark buffer rate are only a starting point in the determination of the buffer value, and they are not a determining factor in the size of the buffer. The credit gap is particularly troublesome at this stage. Except for a short period during the financial crisis 2007-08, total credit has been growing faster than GDP since 2004, but the trend growth is catching up and closing the gap. We think that it is too early to say that the credit gap is signalling diminishing cyclical systemic risks. Many of my colleagues in Europe agree with me that the credit gap is currently a poor indicator for gauging systemic risks.

The idea behind the countercyclical capital buffer is to gradually increase it while risks are building up. But when should we reduce or release it? There is little empirical evidence to rely on here since only one country has reduced its countercyclical capital buffer: the UK following the Brexit referendum. In general, we will have to look at other indicators in a downturn. One obvious indicator is when banks are experiencing credit losses and reducing the supply of credit, but we could also study indices of financial stress, non-performing loans, etc.

So, let me conclude. The countercyclical capital buffer is one of the tools at FI's disposal. It increases resilience by making banks hold more capital when risks are building up. When those risks materialise, the countercyclical capital buffer can be reduced. Should a crisis occur, this releases capital that banks can use to mitigate a credit crunch. In Sweden, cyclical systemic risks are continuing to build up. Credit is growing fast, risk-taking is elevated, and asset price valuations are high. When cyclical systemic risks are building up, we should increase the resilience of the Swedish financial system. This is why I have proposed to FI's Board of Directors, for its' decision at its' next meeting, to raise the countercyclical capital buffer to 2.5 percent.


1) These estimates are based on the assumptions that equity is more expensive than debt and that the increase in costs from having to hold more capital is passed on in full to the banks' customers.