The global ‘splint’ of QE will continue to be used to hold up asset prices - and government bond prices in particular - as central banks dare not lift the tide of liquidity hiding the sharp rocks beneath.
While their policy rates lie close to, or through, the floor (negative), QE - even in the US - has yet to be turned off. Believing it’s the QE stock that matters, the US Fed and BoE have, by reinvesting maturing bonds, kept their sinks topped up. And the BoE (disrupted by Brexit risk), the ECB and BoJ are now running their taps faster, with little sign they can be shut down anytime soon. The risk for pension funds is that ‘looser for longer’ may have many years left to run.
If the US in the 1930s is a guide, we are today only half way through our QE... the chart [below] shows that, as a result of their asset purchases, the world’s big four central banks‘ balance sheets have in total ballooned to over $13trn. This liquidity injection to the private sector is equivalent to about three quarters of US GDP, or 1¼ times China’s. With total world central bank assets (including foreign exchange reserves etc) of $25trn, it suggests about one half has been amassed in just seven years.
Size of central banks’ balance sheets into and since QE (all $bn). Grey is US recession
Source: Thomson Reuters DataStream
That is, after the official US recession ended in mid-2009. There are no counter-factuals, but early QE can probably be credited with unclogging the financial system in 2009, providing liquidity, keeping bond yields down, and yield curves steep. It loosened the monetary reins further at a time when rates were already on the floor. In fact, the extent of the overall loosening has probably been underestimated, given the ‘Taylor Rule’ the Fed uses for setting rates does not explicitly track QE.
When we factor it in, the US and UK are still running de facto negative policy rates. But, it’s been a less than perfect remedy. First, its transmission mechanism has been woefully slow. The responsiveness of GDP to money growth has been far less than the one-to-one assumed, as consumers/producers wary of unemployment and deflation became interest-rate insensitive. No matter how much liquidity, its reflationary impact rests on how quickly we pick it up and push it around the system.
Economics students will remember Fisher’s ‘quantity of money’, where ‘MV=PT’. The impact of a money stimulus (M) on GDP (PT) will be quantified by the speed (V) at which agents push money around the economy. In reflation terms, it’s no good throwing money out of a helicopter if no-one spends it. In practice, with consumers and firms in a liquidity trap, velocity has been slow to recover. Only in the UK, helped by housing’s sensitivity to low short rates, is velocity rising.
Second, the fact that GDP and money growth were lacking till about 2012 suggests the inflation QE spawned came more by inflating asset prices, than the direct, consumer route hoped for in 2009. By running predominantly cost, rather than demand-inflation, QE has come to act more like a tax than a subsidy. Cash stayed cheap, but with confidence fragile and banks repairing their balance sheets, further bouts of QE did little to boost disposable incomes. And as demand stuttered, central banks turned on more QE, creating a vicious circle.
The irony is that QE thus undermined a major justification for using it: to assuage the ”squeeze on households’ real incomes weigh on ...spending” (BoE, October 2011). Added to that have been obstacles from: (i) the unwillingness of institutions to sell back their bonds when riskier assets are unattractive; (ii) regulatory pressure for them to hold onto bonds; and (iii) the reluctance anyway of banks to raise lending.
Third, QE offers little directly to push productivity and wages: still the missing piece from most recoveries. US real GDP is up 10 per cent from its pre-crisis peak in Q4 2007, where rising productivity (+9 per cent) has helped average wages (+19 per cent) beat the CPI (+14 per cent). Yet, the UK’s flat productivity has impeded wage growth (17 per cent) relative to the RPI (+26 per cent). Consumers will have benefitted of course from higher asset income under QE. Though, if aiding disproportionately the higher-end earners, QE could be accused of helping those that needed it ‘least’.
For the stronger-growth US and UK, QE may now have had its day. As we know from Japan, the main benefit is to keep bond yields down for even longer, as others now attest. There may still be benefit to the euro-zone which looks halfway down the Japan route. There, two-thirds of private borrowing is long-rate driven, so low/negative yields should in time lift activity and inflation.
Yet, by distorting financial markets, suppressing saving, and increasing the funding strains on many pension schemes (DB scheme deficits now total about £1trn), QE may be fast becoming a problem not the solution. Under QE, markets know that central banks remain the biggest sponsors of bonds. But with the US Fed now holding one quarter of US Treasuries outstanding, the BoJ over 40 per cent of JGBs, and the BoE one third of gilts, private institutions may increasingly struggle to find the bonds they need, despite heavy government debt ratios. China (page 7) could yet also turn to QE.
And, in the absence of fast growth, rising inflation and low debt, it’s doubtful central banks can intimate to markets they intend to step away without triggering an aggressive rise in yields. By risking asset-price deflation (falling equities, property etc), this would be counterproductive if it hurts the consumers/firms they’re trying to support. Another reason not to risk it is central banks’ own interests, given the ‘skin in the game’ from managing their own balance sheets.
So, central banks seem hell-bent on QE. With Japan still accelerating it after 17 years, most have never experienced a central bank turn it off. The last time the US Fed did QE proper was to pull its economy out of the 1930s depression. Then, it ran QE unbroken for 14 years up to 1951 - despite double-digit inflation touching 20 per cent in 1947.
