While one of the victims of the budget fiasco has been the tapering of QE, it is no secret that sooner or later the Fed (and some other central banks) will slow, then stop, and then perhaps even unwind some of the asset purchases they have embarked on under the heading of QE.
Central banks normally try to soften the business cycle by spurring or limiting credit creation. They do this by controlling the price and quantity of bank reserves, which makes the availability and price for funding bank lending cheaper or more expensive. Banks keep reserves to cover withdrawals and loans.
Banks have two sources for reserves, the Fed and other banks. They borrow reserves from other banks via the money market and the price they pay is the Federal Funds rate. The Fed normally has, by setting the discount rate and by purchasing or selling short-term assets, a firm grip on the Federal Funds rate.
If the economy threatens to overheat, the Fed tightens policy by raising the discount rate and/or selling short-term assets in the money market, which raises the Fed Funds rate. When the economy is in a slump, it does the opposite, in order to stimulate bank lending, and hence interest sensitive spending.
Since the present economic predicament isn't your garden variety business cycle downturn, but a crisis caused by private sector deleveraging (paying off debt) as a result of the financial and housing crisis, this route to stimulate demand has reached the end of the road, otherwise known as a liquidity trap.
But in order to get short-term interest rates low to stimulate lending, the Fed has flooded the banking system to such an extent that short-term interest rates are essentially zero. These excess reserves are just sitting in banks' balances at the Fed, earning 0.25% interest, instead of funding new bank lending.
So "normal" monetary policy doesn't really work; all those bank reserves aren't funding new bank lending. Why? Well, it's a bit of a collective action problem. Households are repairing their balance sheets, so they're inclined to pay off debt, rather than take on new debt.
As a result, they cut back spending, which greatly worsened the economic crisis post 2008. Firms are not very inclined to take on much new borrowing either, as they're sitting on stacks of cash and excess production capacity, the latter the result of households cutting back spending.
What is QE?
QE (quantitative easing) is basically a policy option that central banks use when their normal tools (rudimentarily described above) have reached a limit. Instead of buying short-term assets to get the Federal Funds rate down, the Fed buys long-term assets (government bonds, mortgage backed securities, etc.) to influence the long-term interest rate. The mechanics aren't that different.
Is QE effective?
If normal monetary policy isn't effective because of a deleveraging private sector, why would QE be any different? Indeed, most of the evidence (see here, here, here, here and especially here) show a rather limited impact of QE on long-term interest rates, although there are those, like Peter Schiff, who argue that QE is the only thing that keeps the economy (and the markets) from collapsing.
The predictions of Schiff and others haven't yet materialized, but people like Schiff argue that this is only postponement and the markets and the economy will inevitably collapse, and therefore that the Fed is basically a prisoner of its own policy.
However, it is true that talk of tapering might have been an important reason for rising bond yields. However, bond yields could also have risen in response to an improving outlook for the economy. In fact, it's extremely hard to distinguish these two, as they are heavily correlated. An improving economy is reason for the Fed to decrease asset purchases.
We have another gripe with QE. It either ends up as excess bank reserves or seeps through into the financial markets. Neither way is it likely to be very effective in providing the economy with escape velocity. Why use such ineffective and cumbersome policy instruments when there is an alternative?
It would have been much better to insert it directly into the veins of the economy, by directly financing public investments.
As long as there are plenty of unemployed resources and excess capacity, the risk of such direct monetary financing are rather remote and we wouldn't have "stuffed" the banking system with excess reserves to the hilt, something which, sooner or later, has to be unwound.
How to unwind?
When balance sheets are repaired and more normal conditions return, borrowing and spending will go up, bank lending will pick up and some of these excess reserves will be turned into loans, and have traction in the economy.
This is also a backdrop against which the Fed can start unwinding when certain capacity and unemployment metrics are hit and the output gap (the difference between actual and potential output) diminishes.
However, that doesn't guarantee unwinding will be problem free, but to discuss this, you have to realize that unwinding QE is simply tightening monetary policy. This has timing risk, the Fed can start too early, snuffing the recovery out before it has reached escape velocity, or start too late, when the economy is already too close to full capacity.
QE, because of the massive amount of assets on the Fed's balance sheet and, more importantly, the massive amounts of excess reserves in the banking system, will bring additional problems. How can the Fed deal with this? Well, the normal tools of monetary policy:
- Raise interest rates
- Sell assets on the balance sheet of the Fed
However, raising interest rates will not be very effective as the banking system is flushed with excess reserves. If the Fed becomes more expensive as a source for reserves, banks borrow from one another. As long as there is a large amount of these excess reserves, the Fed is fairly powerless to raise short-term interest rates, a sort of liquidity trap in reverse.
Selling assets (a literal unwind) would be the obvious solution to this, but the numbers are substantial. Selling assets would normally decrease bank reserves and rise rates.
This would slow economic growth and thereby reduce inflationary pressures, but for those that fear that the numbers are too big for the Fed to manage to navigate between keeping growth in tact whilst not letting inflation accelerate out of control, we have a surprise.
A surprising perspective
Here is Bill Mitchell:
Collectively, it turns out, official US government holdings (the Fed and other state bodies) have in percentage terms decreased since the 2008 crisis while private holdings, inclusive of foreign holdings, have gone up.
The Fed holdings as a percentage of outstanding US government debt is visible in the table below:
it looks like there are plenty of buyers both at home and abroad for US government debt and this is unlikely to change much even if the tapering program is initiated. Why? Because many investors aren't investing in the real economy right now and this goes a long way to explaining our present output and unemployment problems.
So it's fair to say not everybody is alarmed at the perspective of unwinding QE and we basically agree. There are two further issues to consider:
- There is no urgent reason for the Fed to get rid of the assets on its balance sheet. They can even let much of these assets mature.
- If credit creation, and thereby inflation threatens to accelerate out of control, there are other ways of throwing sand in the process.
One (called FARP) way is already in trial mode, and it's a little technical so we refer you to a useful source. But there is another, obvious way, a relatively outdated and rather blunt monetary policy instrument, setting reserve requirements.
In most countries, banks are usually left to set their own reserve levels, but setting reserve requirements as a policy tool does still happen in some countries (like China, or France before it joined the euro).
However, this instrument becomes surprisingly effective if the danger is that bank lending and inflation will accelerate beyond what policy makers would deem sustainable levels and are unable to sell assets in large amounts and or fast enough to mop up the excess reserves without causing interest rates to spiral upwards.
The Fed could simply oblige banks to keep bank reserves as reserves. It's crude, but it will do, as the above hypothetical situation is exactly the kind in which fine-tuning at the margins ("garden variety" monetary policy) is clearly inadequate.
Reserve requirements can be gradually returned to normal levels with the passing of time. So the people nervous about an acceleration of inflation as a result of the great unwind can sleep a little easier, we think. The Fed has the tools to deal with that.
Unwinding itself is contractionary monetary policy, and if that should run into problems, the Fed can directly intervene to put a break on any excessive credit creation.
Announcing the end of QE could in fact be something positive, as it is a sign that the Fed believes the economy is back on track. There is another possibility though, which is much darker. One in which the Fed will be near unable to quit QE because the economy doesn't recover anywhere near enough, but this we will deal with in a follow-up post.