The following, written exclusively for The Cobden Centre, is by Dean Buckner (Life Insurance and Pensions Department, Bank of England)
Should regulators call the market? It has been a matter of vigorous debate in the regulatory community since the crisis. ‘Cleaners’ think you shouldn’t mess around with the market, although be prepared to clean up if markets collapse. ‘Leaners’ take the opposite view, divided between those who want to lean on the asset side of overheated balance sheets, principally by increasing capital requirements, and those who want to lean on overheated liabilities, using an ingenious accounting method called ‘matching adjustment’, which allows firms to reduce the present value of future liabilities by increasing the rate at which they are discounted. If spreads blow out on your assets you are allowed to add some of that spread to the discount rate for your liabilities, which cushions the impact in the short term.
This article focuses on whether economists can identify trends or bubbles in the residential property market. The question is important for insurers, whose balance sheets have been damaged by the protracted fall in long gilt yields, and who are hunting for yield in alternative assets such as commercial and residential property. I asked three economists for their views.
John Cochrane of Stanford’s Hoover Institution, author of the magisterial Asset Pricing, complains about a common misunderstanding: ‘people say the market can’t be efficient, because it didn’t predict the 2008 crash. That’s exactly backward. Efficient markets theory says that the market aggregates all information that people have – and no more. The market is not clairvoyant. Consequently, the central empirical prediction of the efficient markets hypothesis is precisely that nobody can reliably tell where markets are going – neither benevolent government bureaucrats, nor crafty hedge-fund managers, nor ivory-tower academics. This is probably the best-tested proposition in all the social sciences.’
He is unimpressed with leaning: ‘regulators are just as human and irrational as individual market participants. Regulators act alone or in committees, without the discipline of competition, where behavioural biases are much better documented than in market settings. They are easily captured by industries, and face politically distorted incentives. Crying ‘bubble’ is empty unless you have an operational procedure for identifying bubbles, in real time and not just after the fact, distinguishing them from rationally low-risk premiums, telling a ‘bubble’ from a justified ‘boom,’ and crying wolf too many years in a row. ‘Bubble’ has some vague extra normative connotation involving looking down on those poor ordinary people paying too much for stuff, and we the benevolent bureaucracy must save them from their silliness. If anyone could tell what the price of tomatoes should be, let alone the price of Microsoft stock, then central planning would have worked.’
What about leaning on liabilities, I ask? If our assets have a 4% average return over risk free because they contain a lot of risk, why not discount liabilities at the same rate? Cochrane disagrees. He is a vocal advocate of more capital, which can flexibly bear losses, but is not impressed by manufactured discount rates. ‘Risk premia come with risk,’ he says. ‘What do leaners do if housing crashes? Using a higher discount rate and then pretending there is no risk is like the most classic error (or skulduggery) of pension accounting’. And in any case, he thinks housing is a rotten long-term investment. ‘[Robert] Shiller has this one exactly right. In the long run, the value of a house is construction cost of new housing. If a city gets too expensive, sooner or later people start new cities’.
I turn to those who think that the housing market is overheated, prominent among whom is Australian economist Steve Keen, author of the iconoclastic Debunking Economics. Capitalising the future by incorporating the predicted future value of an asset into its current value is one of the worst regulatory ideas he has ever heard of. ‘Given that firms taking this position expect house prices to rise, this would increase their capital position, which would enable them to lend more’. It would thus establish ‘a regulatory positive feedback between expected price growth and the level of mortgage lending’. This might not be a problem if there were no relationship between lending and house price appreciation. ‘But there clearly is one – US housing data debunks the assumption that ‘house prices always rise’ – between 1955 and 1975, real house prices fell by over 10 per cent (chart). Even since then, their rise in real terms has hardly been smooth: there were two notable bubbles in real house prices in 1980 and 1990, followed by the daddy of them all in the Subprime boom and bust. Accelerating mortgage demand will lead to rising house prices, while decelerating mortgage debt will lead to falling house prices’.
Rather than amplifying the positive feedback loop between mortgage debt and house prices, he thinks the PRA should be looking at a negative feedback loop, such as his proposal that the maximum amount lent would be some multiple of the expected rental income. ‘This would end the motivation for borrowers to leverage. Currently, when two people of the same income level compete for a property, the winner is the one who takes on more debt’. It is ‘almost universally realized’, he claims, that house prices in London (and to some extent the rest of England) are already too high. ‘The last thing regulations should do is allow lending to be based on manifestly false assumptions about the future. The proposal to let firms factor in expected price appreciation into current valuations is the exact opposite of the kind of regulatory reform we actually need’. Cochrane agrees: ‘not all house prices have gone up. Ask the residents of Detroit about the great returns they got on house prices. Actually, ask just about anyone in the UK outside of London, no?’
Unlike Keen, David Miles (professor of financial economics at Imperial College, member of the Bank’s Monetary Policy Committee from 2009 to 2015) believes that the long run tendency of residential housing is likely to be upwards. ‘A reasonable guide is that house prices in nominal terms will tend to rise in line with nominal incomes, maybe by a bit more. Relative to consumer prices I would expect them to rise pretty consistently’. Nonetheless, he is sceptical about anyone’s ability to call markets, and feels that the system should avoid any ‘chain reaction’ set off by large falls. Like Cochrane, Miles feels that equity is the answer, pointing out that the dotcom collapse did not have systemic consequences because losses were borne by investors who were consciously bearing the risk. The problem is when equity losses spill over into debt, particularly when the debtors have reasonable expectations that their exposure (a pension plan, a no-profits policy) had little risk. This is unacceptable, he says.
Like the two other economists, Miles is wary of leaning on discount rates. ‘If a liability is issued on the expectation or promise that it is risk free, then it must be discounted at the risk free rate, otherwise we descend into nonsense. Suppose we raise £1bn from policyholders with the expectation that their exposure is absolutely safe, and we back this liability with £1bn of assets with the promise of a higher than risk free return. Are we to discount liabilities so that their present value is £800m or even £600m? Can we make £200m or £400m out of nothing? That’s nonsense and a dangerous road to go down’.
What about relaxing capital requirements? I tell Miles that the fuel warning light in my car makes my wife shriek even when there is clearly lots of gas in the tank. Sometimes she wants to get out of the car. Couldn’t I install a button that turns off the light? He thinks that there are two really stupid things you could do. ‘Stopping the car in the middle of nowhere, getting out and looking at it, hoping something will happen is not a good idea. But it’s equally bad to switch the light off, saying “that’s solved that”. You clearly need warning lights plus a plan. For example, drive on, stop at the next fuel station, get more fuel’.
In summary, given the striking differences of opinion among the three economists, the profession seems as divided as ever on the question of market efficiency and the ability of policymakers to act pre-emptively. Keen thought housing markets were headed downwards, Miles thought the long run tendency was upwards, Cochrane thought that nobody (not even regulators) can reliably tell where markets are going. Only one of them supported ‘leaning’, and none supported the concept of valuation adjustment. Of course, this is just a sample of three, but it suggests that the profession has not reached a consensus on prudential leaning, nor is it not clear how regulators could define an operational procedure for identifying a bubble in the housing market that would be universally acceptable within the profession.
Any views expressed are solely those of the author and so cannot be taken to represent those of the Bank of England or to state Bank of England policy. This paper should therefore not be reported as representing the views of the Bank of England or members of the Monetary Policy Committee, Financial Policy Committee or Prudential Regulation Committee.
See e.g. Bank of England Working Paper No. 664 (http://www.bankofengland.co.uk/research/Documents/workingpapers/2017/swp664.pdf)July 2017. Footnote 5 (p15) explains how falls in asset prices are cushioned by a fall in the value of liabilities. For example, a fall of 10% in asset values is matched by a (presumed) fall of 9.7% in liability values, meaning that the presumed impact on firms’ equity is small. See also Financial Stability Paper No. 42 (http://www.bankofengland.co.uk/financialstability/Pages/fpc/fspapers/fs_paper42.aspx), which places work such as this in the context of ‘system-wide stress simulation’.