- The Fed will imminently begin to slim its balance sheets. Other central banks will follow
- Quantitative tightening may shrink money supply
- But the main impact is likely to be upward pressure on bond yields for a very long time
The Federal Reserve will soon begin to seriously slim its balance sheet. It is not quite the first time it has done so; it also let bonds run off in 2017-1019. But apart from that brief episode, this is really the first time in history that a central bank is actively shrinking its balance sheet. Nor will the Fed be the only one to do it. Other central banks will follow suit, some faster, others (the ECB springs to mind) more slowly and not for some time.
This is therefore a good time to revisit what Quantitative Tightening (QT) actually means. First, it should be noted that the Fed still only intends to stop rolling over maturing debt. It will not actively sell bonds that have not yet matured. This could be significant, although the FOMC could presumably always change its decision if it so desires.
Much has been talked about how Quantitative Easing (QE) involved ‘printing money’. The fear is, not unnaturally, that QT will therefore mean the opposite, namely destroying money. This is possibly justified, but not necessarily so. In order to understand why, I have to get technical, for which apologies.
‘Money’, in the sense of broad money, the monetary measure most relevant as an indicator of economic growth and inflation, consist of the liquid holdings – cash and bank accounts – of the non-bank private sector, that is to say households and non-financial companies. In some economies, other liquid investments may also be considered part of the stock of money. Crucially, however, deposits held by banks with the central bank (so-called reserves) are not included in broad money. Nor, equally crucially, are deposits held by the government.
This is important, because it tells us, first, that QE did not necessarily involve ‘printing money’.
Whether QE meant printing money depended on from who central banks bought the bonds they purchased. If they bought them from banks, then all they did was to create reserves that the selling bank held with the central bank. The impact on broad money was therefore zero.
By contrast, if the original sellers were households or companies, then money would be created, as the central bank paid the sellers and this money ended up in their bank accounts. This argument is supported by the continued weak growth of broad money in most economies during periods of QE.
Now let’s look at the reverse: If central banks were actually selling non-matured bonds, whether this involves destroying money or not, will depend on who the buyers are. If the buyers are banks, there is no money destruction, since the only thing that happens is that their (huge!) reserves diminish. If the buyers are households or companies, then, yes, money is destroyed.
However, as noted above, the Fed does not intend to sell any non-matured bonds. Instead, it is simply not rolling over maturing bonds. This is slightly different. What happens here is that the Fed presents the matured bond to the issuer, who pays up the nominal value of the bond.
If the bond was issued by a private entity (say, a mortgage-backed security), then their bank account will shrink and money is destroyed. But most of the Fed’s bond holdings (and those of other central banks) are government bonds. So much depends on what happens next. Here, there are three possibilities:
- The Fed presents a matured bond to the US Treasury. The Treasury pays the Fed and nothing else happens. Net impact on money supply: nothing.
- The Fed presents a matured bond to the US Treasury. The Treasury pays the Fed and then promptly issues a new bond to the same value, which is sold to a bank. Net impact on money supply: nothing.
- The Fed presents a matured bond to the US Treasury. The Treasury pays the Fed and then promptly issues a new bond to the same value, which is sold to a non-bank private entity (that is to say, a household or a company). Net impact on money supply: money supply shrinks.
However, neither the US Government, nor other governments borrow in order to build up a pile of money. They borrow because they intend to spend. And the moment they spend they funds they have just borrowed, they are transferred back to the private sector and so become money once again.
Hence, the monetary impact of QT is likely to be limited. That is actually good news, since broad money growth currently is slowing quite sharply in the leading western economies, raising the fear that central banks will over-react on the tightening.
Instead, the main impact of QT is likely to be felt on interest rates, specifically on the yield on long-term bonds. We are already seeing bond yields rising because of inflation. In addition, governments are still running substantial deficits, which provide further upward pressure on bond yields. But most importantly, with the end of QE and the beginning of QT, central banks disappear as the bond buyers of first resort. Not just that, they also become a further source of upward pressure on bond yields since the borrowers now will have to issue bonds with more attractive coupons to other buyers. In due course, this will once again make bonds an attractive investment for long-term investors, such as pension funds, which need to match income streams to expenditures. But the road from here to there is likely to prove not only painful, with steadily rising bond yields, but also long. The maturity profile of the Fed’s bond holdings means that a complete run-off (unlikely though that is) would drag on for decades. The likelihood is therefore (and for other reasons, see also this Comment from 2017) that we will see a secular (i.e., multi-decade) bond bear market.
How far will bonds ultimately rise? That is difficult to say. But over the very long term, real yields have averaged around 2%. At the moment, this would imply a nominal bond yield of around 10.5%, whereas the actual yield is about 2.9%. That is unlikely to happen, unless the Fed is perceived to have completely lost its grip on inflation. But it certainly points to yields well above where they are today, even if inflation eventually (perhaps in late 2023 or early 2024) falls back towards the 2% target. Even if higher inflation by itself did not push up bond yields (as it is already doing), the process of QT would certainly do so.