The mystery of quantitative easing (QE) is that no-one really knows how it works. Ben Bernanke, the former head of the Federal Reserve, once quipped:


The problem with QE is that it works in practice, but it doesn’t work in theory


In principle, the public sector is simply swapping one kind of government liability (bonds) for another (money). Under the expectations hypothesis of the determination of interest rates, this kind of financial engineering should have no impact on yields. So it’s a bit of a puzzle why it seems to have such powerful effects.


The chart below shows this transformation for the UK. In 2007, general government liabilities stood at just over 40% of GDP. Of that the vast majority was bond-financed. However, even before the days of QE, the Bank of England held some government securities so a tiny sliver was money-financed.


By 2017, total liabilities had grown to around 90% of GDP. The Bank of England’s asset purchases over the intervening decade have been considerable. Of total government liabilities, around a quarter are therefore now money-financed.


This is not a totally unprecedented restructuring of the public sector’s liabilities but we have to go back to World War II to find the last comparable episode in the UK.



Economic theory is pretty silent on the benefits of this kind of liability transformation exercise. Is it the flow of asset purchases that matters? Is it the stock of bonds held by the central bank that matters? Is it the act of creating additional money to finance those purchases? Or is it the signal about interest rate intentions?


This quasi-academic debate is crucial given that the Federal Reserve has now begun the multi-year process of winding down their balance sheet. After nearly a decade of quantitative easing, we are now entering the era of US quantitative tightening (QT). The September meeting delivered the plan laid out in detail in June. Starting in October, the Fed’s balance sheet will therefore shrink by up to $300bn over the next 12 months as maturing securities are not fully reinvested.


How much will this impact asset prices and the broader economy? It is very hard to know. But it is not entirely true to say that we’re entering uncharted territory.  A 2015 working paper from Niall Ferguson (among others) suggests we should be worried:

Historically balance sheet reduction episodes have gone hand-in-hand with lower growth rates, somewhat lower inflation rates and substantial slowdowns in financial sector lending activities  


If the US were equal to the world, then it would be reasonable to be pretty cautious about growth assets due to the onset of QT. However, there are a few caveats to this conclusion.


First, Lars has recently questioned the extent to which equity markets (in particular) have been puffed up by QE. Analysing the equity bull market, he finds that the majority can be explained by earnings growth and normal multiple expansion. If QE didn't have much of an impact on the way up, why should QT have a big impact on the way down?


Second, the Fed's balance has already been shrinking as a share of GDP for the last three years. From that perspective, we've been living with QT since late 2014 and it hasn't exactly been a difficult period for risky assets.


Third, at the global level, this is definitely not the end of the story. On our estimates, the European Central Bank and Bank of Japan will collectively purchase the equivalent of $1 trillion of assets over the next year. On top of that, we may have some reserve accumulation in emerging markets. If those purchases are funded by money printing (rather than sterilised by issuing domestic debt), they it will add further to the global QE flow.


From a global perspective, the QE train is starting to slow down. But, it will not come to a standstill until 2018 Q3 at the earliest. It will be over a year before it goes into reverse


So, for QT to have a big impact (at least in the next twelve months), it needs to be because of problems that are specifically associated with a shrinking in the nominal value of the US dollar narrow money supply. Those stresses will probably be seen first in short-dated dollar funding markets (eg. the spread between short-term borrowing rates for the US government and US banks, the spread between the cost of 'on-shore' US dollars and' off-shore' US dollars), and the price of the dollar itself.


Consistent with our 'policy divergence' investment theme (and despite little success on this front in 2017) we therefore think it is prudent to continue holding long US dollar positions as a hedge against unexpected strains due to the normalisation of US monetary policy.