Federal Reserve Economic Data

The FRED® Blog

Posts tagged with: "QUSPAM770A"

View this series on FRED

What’s a countercyclical capital buffer?

A new monetary policy tool, that's what

The Federal Reserve has tools to achieve its monetary policy goals: the discount rate, reserve requirements, open market operations, the interest rate on reserves… and now also the countercyclical capital buffer (CCyB). The CCyB is intended to avoid the banking failures of the Great Recession by ensuring individual banks and the banking sector as a whole have enough capital on hand to provide a flow of credit to the economy without jeopardizing the solvency of this sector. To achieve this goal, a monetary authority (such as the Fed) would require banks to hold a percentage of their capital in a reserve or buffer account. The exact percentage can be changed depending on the state of the economy. If the authority believes too much credit exists, then the percentage of banks’ capital held in reserve must be increased to reduce banks’ ability to lend or provide credit. On the other hand, if the authority believes the economy needs to be stimulated, the percentage of banks’ capital held in reserves can be reduced to encourage more lending.

In practice, this is accomplished by setting a minimum value for the ratio of banks’ Tier 1 capital relative to its risk-weighted assets. Tier 1 capital can be thought of as stockholder equity and retained earnings, while risk-weighted assets are mostly the risk-weighted value of outstanding loans. The more risky the outstanding loans relative to held capital, the lower this ratio will go. To raise this ratio, banks can hold more capital and/or issue fewer risker loans.

The first graph shows the ratio of Tier 1 capital to risk-weighted assets since 2010 (blue-line) for the banking sector. Even without the CCyB, the minimum capital requirement and capital conservation buffer ensures that this ratio should not decline below 8.5% for any bank (the lower green line). The Fed could use the CCyB to raise this minimum value by up to 2.5 percentage points, to 11% (the higher green line). Currently, however, the Federal Reserve hasn’t used its authority to set a CCyB and so the lower bound for this ratio remains at 8.5%. As the graph shows, the aggregate banking sector ratio (blue line) hasn’t fallen below the 11% maximum ratio that the Federal Reserve can put in place.

In some sense, the aggregate ratio of Tier 1 capital to risk-weighted assets could be misleading: Even though the overall average is relatively high, the individual ratio of any given bank may very well be below the 11% or even the 8.5% cutoff. By looking at public call report data, we can get some idea of the position of individual banks. From our calculations, assuming the data are normally distributed, about 16% of banks will have ratios below the dotted red line. The line for this subset of banks is above the 8% line and crossed the 11% line beginning in the first quarter of 2018. This suggests that raising the minimum ratio level would have a significant effect on many banks. (Of course, this is only an approximation: The data aren’t exactly normally distributed. Instead, there are a substantial amount of firms clumped around the 8.5% lower limit, because dipping below that limit will introduce substantial regulatory burdens. In either case, however, the conclusion that a significant portion of banks would be effected by a rising CCyB holds true.)

So how do the monetary authorities decide when to use a CCyB in this way to limit the amount of credit in the economy? In other words, how much credit in the economy is “too much?” It’s not enough to simply look at the total amount of credit because any given amount of credit may be too much or too little depending upon the size of the economy. To resolve this quandary, the Federal Reserve monitors the ratio of total credit to GDP and detrends the series to ignore any long-run movements in the ratio unlikely to be associated with any given financial crisis. Perhaps the simplest method of detrending is to create a line of best fit for the data, as we have done in the second graph. Whenever the credit-to-GDP ratio gets high enough above the trend line, say, by 2.5% of GDP, a monetary authority may want to consider increasing the capital buffers to slow the growth of credit.

How these graphs were created: First graph: Search for and select “Financial Soundness Indicator; Regulatory Tier 1 Capital as a Percent of Risk-weighted Assets, Level” and click “Add to Graph. From the “Edit Graph” panel, use the “Add line” feature to add the other data series (red) and the two (green) user-defined lines: For the red line, add the original series again and use the “Formula” bar to specify either plus three or minus three depending on what you’d like to show. For the green lines, use the “Create user-defined line” option and click on “Create line”: Change the “Value start/end” fields to reflect the constant value you want to display (we graphed 8.5 and 11). All colors and line styles can be changes in the “Format” panel.

Second graph: Search for and select “Total Credit to Private Non-Financial Sector, Adjusted for Breaks, for United States.” Then, below the series, select “Percentage of GDP” option to see the ratio of interest. Detrending this data series is the tricky part: FRED doesn’t yet have a built-in function for detrending; but you can add a line to the graph as we did for the previous graph and set it so that seems to come as close to as much of the graph as possible. We downloaded the data as a CSV file using the download button and then performed a quick regression. Next we used the fitted value for 01/01/1952 (60.763) and the fitted value for 10/1/2018 (163.7093) to draw the computer-generated line of best fit in the FRED graph. If you’re more experienced in statistics, read on: The Basel Committee, an international body that set initial international guidelines on the use of CCyBs, recommends detrending with a Hodrick-Prescott filter with a lambda value of 400,000. We did this, too, and our results aren’t substantially different.

Suggested by Ryan Mather and Don Schlagenhauf.

View on FRED, series used in this post: BOGZ1FL010000016Q, QUSPAM770A


Subscribe to the FRED newsletter


Follow us

Back to Top