Why governments should not be in hock to the bond market

 

My starting point is not what Andy Burnham said but points recently made by Chris Dillow. Here he argues that we shouldn’t worry about the bond market per se, because that market is just an early warning system for future problems that any government should be worring about anyway. In that sense, the bond market is no different from, and less important than, an OBR forecast.

For example, if the government were to start borrowing more to increase public spending, interest rates in the bond market might well rise, but this is only because additional public spending would generally increase the demand for resources, putting upward pressure on inflation. The government should worry about raising inflation, not the bond market anticipating this.

I think this argument is basically correct, but it needs a bit of elaboration. In particular, while arbitrage means the bond market is generally a predictor of future inflationary pressure, there are circumstances in which other factors might influence interest rates because arbitrage breaks down, as it threatened to do when the pandemic hit, or after the Truss fiscal event. In this post I will explore those possibilities, and suggest once again the government has little to worry about as long as the central bank is doing its job. In addition I want to explain why these basic truths are hardly ever acknowledged in the media, because acting as if the bond market is a vengeful god is just too attractive for most journalists.

Bond rates are interest rates on long term assets, so they depend on what investors think future interest rates on shorter term assets will be. If the two get out of line then there is an expected profit opportunity, and arbitrage is the process which drives those opportunities to virtually zero. As short term interest rates are set by the central bank, then the main influence on bond market rates are expectations about what the central bank will do in the future, which in turn primarily depends on what inflation will do in the future.

Any price that depends so heavily on expectations is likely to be quite volatile. This volatility is a gift to political journalists. If some political event happens, then there is a good chance that bond market rates will have subsequently risen, so they can write a story about that political event spooking the markets. That may be fair enough if that political event increases the chance that future inflationary pressure, but equally the movement in rates could be about something completely different. As Dominic Caddick notes, there were plenty of headlines about a ‘Burnham premium’ when rates rose in the early days of the Makerfield election, but no headlines when rates subsequently fell even though expectations he would win increased over time. Remember that most of the time no one actually knows why market rates move on a day to day basis.

Politicians and the public can safely ignore modest short term volatility in bond market interest rates. If more sustained increases in rates are due to growing expected future inflationary pressure then they should worry, but not because of what the markets are doing. They should worry because inflationary pressure is a problem in itself. Even if the central bank manages to suppress that pressure by raising interest rates, in normal circumstances higher interest rates discourage private investment and therefore longer term economic growth.

Doesn’t the link between bond market rates and the cost of borrowing mean that governments should worry about the markets whatever the reason that rates are moving? No, because the interest rate the government pays on much of its debt is fixed at the time the debt is issued, so rising rates only affect the much smaller amount of new borrowing. As a result the only reason governments should take note of the bond markets is if the markets are making intelligent guesses about future inflationary pressure.

A consequence of arbitrage is that the supply and demand for government debt is irrelevant to what the current market interest rate on government debt is, or more accurately demand for debt will adjust to ensure arbitrage holds. That in turn means that the budget deficit, which is a key determinant of the supply of new government bonds, is also irrelevant. Government deficits were highest after the Global Financial Crisis and interest rates steadily fell. Now of course current and future deficits, if they reflect additional government spending or cuts in tax rates, may indicate higher future inflationary pressure, so in those circumstances journalists are only half wrong if they write that higher deficits are increasing bond rates. But if deficits are rising because the economy is weakening then higher deficits will not raise bond market rates.

What is never the case for an economy like the UK is that interest rate movements reflect worries about government default. A country with its own central bank that borrows in its own currency cannot be forced to default by the markets, and debt would have to be orders of magnitude above current levels before any rational advanced economy government actually chose to default.

Arbitrage works most of the time, but sometimes it doesn’t, and those sometimes are the times we may well remember. Two related things can interfere with arbitrage: uncertainty and thin markets. If uncertainty about future short term interest rates increase, then the risk of holding long term government debt increases relative to holding shorter term assets, which means that investors will require a risk premium to hold longer term assets like government debt, and in extremis may decide to withdraw from the market altogether.

This is what happened during the onset of the pandemic in Spring 2020, when the markets stopped buying the debt of most governments completely. In its place central banks bought government debt instead (Quantitative Easing). With a recession likely, the last thing inflation targeting central banks wanted was surging longer term interest rates. Sometimes, however, uncertainty can be increased by government actions or inaction, and the central bank may be understandably reluctant to shield the government from the consequences of its own actions. That was a key factor in what happened after the Truss fiscal event, but before looking at that we also need to talk about thin markets.

Arbitrage works when there are lots and lots of potential buyers of government debt, so demand for debt can rise and fall depending on what the profit opportunities are. But occasionally that isn’t the case for UK government debt. These occasions are most likely to occur at times of high uncertainty, when lots of potential buyers stay out of the market because they are risk averse. The clearest case of that was in the final stages of the crisis following the Truss fiscal event, when for reasons I won’t go into rising rates meant pension funds had to sell off a large quantity of government debt very quickly. Supply increased, but because of heightened uncertainty caused by the fiscal event the demand to meet at rates implied by arbitrage wasn’t there, so rates rose. (My discussion of the market reaction to that event is here.)

On that occasion too the Bank of England stepped in and bought government debt. But its actions were limited, because it didn’t want to shield the government from the consequences of its own irresponsibility in announcing tax cuts whilst giving no clear indications of what spending cuts would be made, if any. This government irresponsibility meant it was quite right to call part of the rising level of bond market interest rates a 'moron premium'. 

These events, and what happened at the start of the pandemic, tells us that the Bank is prepared to intervene to restore arbitrage when the market is thin or external events create intense uncertainty, but not if greater uncertainty is created by the government itself fuelling inflationary pressure. Whether that warrants claims that the Bank ‘brought down the Truss government’ I will leave you to decide.

So do governments need to worry about the bond markets because sometimes arbitrage fails? Again the answer is no, because the central bank will intervene on those occasions. The exception is where greater uncertainty is caused by the government itself, but governments should worry about that in any case. Once again the bond market is just an indicator of underlying problems, and not in itself a constraint on what the government can do.

Why is it so rare for the media to acknowledge any of this, and act as if the bond market is some vengeful god that could strike at any point? First, when a Labour government is in power it suits journalists on the right to do this. Second, most people have no idea what arbitrage is and so it is much easier to pretend the actions of the bond market are mysterious and, for political journalists, highly political. As a consequence, journalists hardly ever acknowledge that the ‘market experts’ they like to quote have no secret knowledge about why rates move, and work for firms that make more money when governments tighten fiscal policy.

All this means policy that is based only on fears about the bond market, or worse still about what journalists are writing about the bond market, will be bad policy. Andy Burnham was right that governments should not be in hock to the bond markets. Ironically this will also become one of the first tests of Burnham's leadership. Will he appoint the obviously best candidate as Chancellor, or will he succumb to pressure from the media and market pundits to appoint someone more to their liking.