Have FSRs got news for you?

Rhiannon Sowerbutts, Vesko Karadotchev, Richard Harris and Evarist Stoja

While communication has been recognised as an important aspect of monetary policy for over three decades and received an enormous amount of attention in the academic literature, there has been almost no attention paid to the importance and effects of financial stability communication. In a new working paper we examine financial markets’ reaction to the Financial Stability Report.

Since the global financial crisis, over 40 nations have set up financial stability committees with powers to intervene in financial markets by, for example, changing banks’ capital requirements or placing loan-to-value limits on new mortgage lending. But policy intervention is not the only way to affect financial stability. Communication is another important tool, and over 60 countries now publish a financial stability report at least once a year.

We examine the effects of financial stability communication on financial markets. Even though the aim of FSRs is not to affect market prices but to inform the public about potential financial stability risks they could have an impact on financial markets if they reveals hitherto private information about the health of the financial system and regulatory policies. On the other hand, if the content is predictable then there should be no impact on market prices. This is considerably more challenging than for monetary policy because (i) there’s no obvious financial instrument tied to policy levers for markets to trade (you can’t buy a future on the FPC’s instruments like you can for Bank rate) and (ii) there are fewer financial instruments relating to forecasts/realisations for the relevant outcome variables (there are official figures for inflation, and the Bank publishes a forecast for future values, but there are no official figures or forecasts for, say, the probability of a crisis).

Have FSRs got news for you?

We look at the market reaction to the Bank of England’s Financial Stability Report. As in many other countries, this is a flagship publication of policymakers and arguably their primary communication channel on matters of financial stability. It appears after two of the Financial Policy Committee’s (FPC) four meetings per year. Alongside this, the FPC also communicates by issuing (much shorter) policy statements and FSR records. In addition, FPC members can communicate individually through their speeches.

We use an event study to examine whether FSRs represent “news” to market participants. In the econometric literature, a piece of information such as a formal report or policy announcement is “news” if – and to the extent that – it says something is different to what market participants were expecting. Since markets can process information fairly quickly, empirical researchers often test for “news” by looking at whether financial markets react around some narrow window after the announcement. If market participants fully expect a given policy intervention, its actual announcement should have no “news” value and should not move prices. If markets do move in response, then there was some news contained in the announcement.

For each FSR between December 2010 and June 2018 (16 in total), we examine whether there is a significant change in banks’ and other financial firms’ (financials hereafter) equity prices in the days after an FSR is published (including the publication day).

The FPC’s actions can affect banks by making them safer, and/or affecting their profitability, so equity price is a reasonable indicator to look at. Bank equities are very liquid, so the price should reflect news quickly. But we can’t just look at financial sector (banks, insurers and other financial institutions traded on the stock market) share prices in isolation. Why? Suppose financials’ prices went up by 5% but all other equity prices went up by 10%: that would mean financials underperformed by 5%, but we wouldn’t know it by looking at their share prices alone; we would need to know what happened to the rest of the market. And we can’t just subtract one change from the other, because it might be that financials’ share prices are generally more volatile than other equities, so can’t be compared directly on any particular day.

To tackle these issues we use a fairly standard technique, which has three steps:

First, we build a model of the relationship between financials’ share prices and the FTSE 250 index in “normal times”, when no FSRs are published (grey bit in figure below). We do this by running a standard (OLS) regression. We can use this model to predict financials’ share prices on any given day. The difference between what our model predicts and the actual share price is called “excess return”.

Second, we compute the excess returns for what we call our “placebo sample”: randomly selected days when no FSRs are published (the blue bit on the figure below). Full geeky details of how and why we construct the placebo sample are in the working paper. We also compute excess returns for days when FSRs are actually published (red bit in the figure below).

Third, we consider whether the excess returns on random (placebo) days are significantly different to the returns on publication days.

Excess returns will on average be zero, but on any given day they can fluctuate around zero. Our model can’t explain why returns turned out as they did on any particular day, but crucially, we can test if there are any systematic differences between days just after FSRs and days in our placebo sample.

Figure 1: The timeline of our estimation

In theory, if there are no other large new stories in the two days following the FSR publication, then any abnormal returns can be attributable to the news contained in the FSR. But we aren’t always lucky in that respect: for example, the results of the referendum on leaving the EU in June 2016 shortly before the FSR was published changed the relationship between banks and the ‘market’ (i.e. the other companies in the FTSE 250) – largely due to concerns about the impact of the results of the referendum on banks’ business models. Likewise on the day the FSR was released in H2 2011, the Federal Reserve, ECB and BoE announced a coordinated dollar swap line, which relieved dollar funding pressures on banks and therefore affected their share prices positively. We treat these events carefully when making inferences because the results might be contaminated by big non-FSR news.

Our analysis is summed up in Figure 2 below. The blue bars show the distribution of excess returns from the placebo period – lighter blue represents  the middle 90% of the distribution, while darker blue correspond to the upper and lower extremes (top and bottom 5%). The red dots show the actual excess returns on FSR publication days.

Figure 2: Abnormal returns around FSR dates are not statistically significant

With one exception, the dots always lie within the central 90% of placebo distribution. So FSR dates don’t appear to be associated with systematically bigger, or more extreme, excess returns. One data point does lie outside the central 90%, but of course we’d expect for this to happen 10% of the time on the basis of chance alone, so one observation in 16 in the extreme 10% is not unusual. And the single observation outside the 90% is the one published in July 2016 shortly after the referendum and as we noted earlier we can’t make inference from that date. Because the relationship had already been changed by the referendum news we would probably expect an abnormal return even if there was no news in the FSR. So UK bank equity prices don’t appear to behave systematically differently around FSR report publications compared to other times.

There are a several different potential explanations for this lack of statistically significant impact. First, the FSR and any Financial Policy Committee (FPC) actions contained within it may be anticipated by market participants, perhaps because the FPC has signalled the likelihood of policy action in advance, perhaps in discussions in previous FSRs. Alternatively, the lack of impact could be because the policies announced are complex and the implications for bank equities take time to be understood, meaning any price impact occurs slowly and not in the days immediately following publication. A third explanation is that financial markets are not able to trade the news of the FSR because its ramifications are too diffuse to show up in a simple movement of share prices. It’s difficult to adjudicate between these different explanations, as they are hard to test by formal statistical means.

To investigate the “fully anticipated” hypothesis we did explore whether FPC policies appear to be signalled in advance and find some evidence that this is the case. For example when the FPC announced the increase of the UK countercyclical capital buffer rate to 0.5% in the July 2017 FSR, it also signalled that it expected to increase the rate to 1% in November 2017 – which it then did. We also used the sentiment analysis developed by Federal Reserve Board Researchers to see whether the sentiment of the text in the FSR – i.e. how optimistic or pessimistic it is – is predictable based on the sentiment in the previous FSR and changes in economic variables such as GDP: we find some tentative evidence that it is.

The key takeaway from our analysis is that financial markets do not appear to be surprised by the Bank of England’s Financial Stability Reports and the Financial Policy Committee actions contained within them. One reason seems to be that market participants are able to largely predict the content of announcements in advance based on their familiarity with the way policymakers usually react. So the actual publication of reports is not actual news and doesn’t move markets.

Rhiannon Sowerbutts works in the Bank’s Macroprudential Strategy and Support Division and Vesko Karadotchev works in the Bank’s Stress Testing Strategy Division. Richard Harris works at the University of Exeter and Evarist Stoja works at the University of Bristol.

If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.

Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.

Share this: