Some Charts that Explain Doubts about Quantitative Easing and What they Mean for the FOMC
In a recent speech, Atlanta Fed President Dennis P. Lockhart took note of divisions within the Federal Open Market Committee (FOMC) as to whether further quantitative easing (QE) would be effective. When I cited Lockhart’s remarks in a post, a commenter expressed surprise. Yes, said the comment, it is possible to see how some people might doubt that QE would have a beneficial effect on the real economy, but how could anyone, especially inside the Fed, think that large-scale expansionary policy would not speed the growth of nominal GDP?
The continued doubts can be expained, at least in part, with three charts that show the limited effects of past QE. Before showing them, though, it will be worth reviewing how expansionary monetary policy is supposed to work.
The first important point is that the Fed has no direct control over market interest rates. Except for a few administrative rates (the discount rate and the deposit rate on reserves), the Fed can influence interest rates only indirectly through open market operations, that is, purchases and sales of items from its portfolio of assets. It carries out day-to-day open market operations mostly with repurchase agreements on short-term instruments like T-bills. QE involves larger, longer-term purchases of a greater variety of long- and short-term assets.
When the Fed buys assets from the public, the funds it uses to pay for them appear on the right-hand side of its balance sheet as an increase in the monetary base—the sum of bank reserve deposits and currency. An increase in the reserves component of the monetary base is supposed to stimulate lending by banks, which in turn, causes the money stock (M2) to expand. Just how much it expands depends on the money multiplier—the ratio of M2 to the base.
Finally, an increase in the money stock is supposed to stimulate spending, so that nominal GDP increases. How much it increases depends on velocity, that is, the ratio of nominal GDP to M2.
Your Econ 101 textbook calls all of this the transmission mechanism for monetary policy. The question is, then, how well has the transmission mechanism actually worked for the recent episodes of quantitative easing, popularly known as QE1 and QE2? If we look at the following chart, we see that it has not worked very well.
The chart shows that over the past four years, the Fed has done a fine job of expanding the monetary base. QE1 and QE2 stand out as the steep segments of the top line. The transmission mechanism is another story, however.
Until 2008, the base, M2, and GDP (in the chart, all of them are normalized so that 2005 = 100) move closely together. The close relationship of these variables would continue back as far as one cared to extend the chart; this one starts in 2005 only in order to show the years since 2008 in greater detail. Since 2008, though, the monetary base has gone its own way. The quantity of money in circulation has increased only modestly. The effect on nominal GDP was even smaller. Nominal GDP actually fell in the immediate aftermath of QE1. Since then, it has struggled even to reach its historical trend, despite the modest expansion of M2 achieved by QE2.
The widening gap between the base and M2 indicates that the money multiplier, one key element of the transmission mechanism, has fallen. The next chart, which shows the money multiplier directly, indicates just how much. Most of the decrease in the money multiplier is attributable to growth of bank reserves.
The gap between M2 and nominal GDP shows a decrease in velocity, another key element of the transmission mechanism. Velocity, roughly speaking, measures the number of times per year people spend each dollar of the money stock to purchase the final goods and services that go into GDP. When interest rates are high, firms and households sharpen their cash management practices so that they can do their business with the smallest money balances possible. When interest rates are low, as they have been since 2008, there is less incentive to economize on cash, so firms and households allow idle liquid balances to accumulate.
The charts clearly show why there are doubts about the effectiveness of quantitative easing. However, they do not conclusively show that further QE would be a bad idea.
Those who favor further expansionary monetary policy argue that past efforts have simply been too timid. There is no practical limit to how large the Fed’s balance sheet can grow. Double it again, quadruple it, keep going long enough and the nominal GDP line surely will begin to budge. In the unlikely event that it suddenly grows too rapidly, say QE advocates, the Fed has plenty of tools available to shrink its balance sheet rapidly if it needs to.
The bottom line: Yes, there are reasons to think the transmission mechanism is weak and to be disappointed with the effects of past quantitative easing. Looking ahead to next week’s FOMC meeting, that is a strong argument against announcing any further QE program that is limited either in time or in size. In view of recent experience, as reflected in the above charts, limited QE would not be enough to change expectations and, therefore, might accomplish nothing at all.
Aggressive, open-ended QE could be another story. As Sebastian Mallaby argues in yesterday’s Financial Times, such a program would work best if it were anchored in an explicit target. Chicago Fed president Charles Evans suggests a 3 percent inflation target until unemployment reaches 7 percent. An explicit nominal GDP target would be even better. If the FOMC has no stomach for such a bold program, it might be better to do nothing rather than further tarnish the reputation of QE with a program that is too small to work.
11 Responses to “Some Charts that Explain Doubts about Quantitative Easing and What they Mean for the FOMC”
Sad how long is taking these economists to connect the dots. My question is how much longer the shadow banking system will be allowed to keep making outrageous profits at the expense of the real economy. I suppose they tell the politicians 'we need the money so when you have to write down your sovereign debt, we can survive the losses", or some such rot.
Just try to get rid of interest on reserves. What can possibly go wrong?
first off there is no doubt who is in charge of the Fed…and given the nasty Philly print Chairman Bernanke's hand is only strengthened…if it could be anymore actually. Second "he will do what he does." Interestingly he has a habit of intimating actions rather forcefully…then executing when all others have poo-poohed the intimations. Successfully i might add. Operation Twist comes to mind. My guess is on reverse repo's coming. Just a guess of course. Don't have a clue what it means should he do that of course. Should I?
The predictable money multiplier started to break down decades ago as reserves requirements were loosened and complicated. It disappeared altogether with the introduction of interest on reserves and large-scale asset purchases that expanded base money far faster than it could possibly be processed into multiples of broad money. That's all old hat. The question is can QE in our present situation expand nominal GDP? That is, can it spur growth, inflation, or some combination of the two?
I'd say that first we should recognize that QE definitely won't spur inflation if it can't spur growth. The notion that a central bank could plausibly target a positive NGDP growth rate during a recession is crazy. So the question boils down to whether QE can spur growth and inflation.
The argument against is that QE only increases the base money supply, and we have so much excess of that, more couldn't make any difference. There's a lot of truth to this argument.
But QE in our current situation does also tend to increase the broad money supply, some – roughly one-for-one with the base money increase (eg $600b of QE adds about $600b of M2, which isn't a huge relative gain). That's because the QE asset purchases tend to be ultimately sourced from the non-financial sector. Only if banks used QE to unload assets they didn't want to the Fed would QE not create M2. That hasn't happened – banks have tended to buy Treasurys and agencies alongside the Fed.
So, then, does this addition of M2 boost activity? It's hard to say, since we don't know what would have happened without it, and how important the growth in M2 was relative to the early growth in M0 and all else that was going on.
My opinion is that the early growth in base money was the most powerful thing that QE could do, and further QE wouldn't replicate that. The M2 growth that results from QE seems too small to matter much, especially since it seems to stay largely within the financial sector.
In short, I don't believe monetary policy alone can generate meaningfully more growth or inflation in this situation. My opinion would only change if monetary policy became a quasi-fiscal or quasi-industrial policy, ie direct central bank lending to the real economy.
As Wayne Angell from the Kansas City Fed, elevated to Vice-Chairman of the FOMC, told Paul Volcker in that fabled summer of 1982, the KC Fed research of Hoenig, Sellon, and Hakkio demonstrated clearly that the monetary policy transmission mechanism had been broken by Reagan's banking deregulation that allowed savings and checking to commingle.
It's all in the literature. The clowns at the St. Louis Fed, drooling with monetarist zeal, just did not understand, but Volcker was converted which led to the MASSIVE easing in that summer to address, inter alia, the BAM (Brazil Argentina Mexico) crisis.
Mallaby still has not addressed Paul Volcker's reaction to a very similar suggestion last year.
What happens if your plan doesn't work? Are you going to raise the inflation target to 5%?
Until Volcker's concerns are addressed, I don't think this is going to fly.
The Fed does NOT want to let the value of its two trillion dollar portfolio to decline, so it's DEFLATION slowly, within the framework of the Humphrey Hawkins Federal Reserve dual mandates.
Yes, we agree that the money multiplier does not hold constant during times of stress. It does have periods of remarkable stability, to be sure. For example, in 2001 through 2005, it was never below 7.99 and never above 8.09. But lately, we all agree, it is pushing on a string.
However, if a growing base has little or no effect on M2, and M2 little effect on NGDP, then there is a definite possibility of fiscal gain from switching bonds around between the Fed and the Treasury. The Fed should buy up federal debt, especially higher interest long-term bonds, and recycle the interest back to the Treasury. The profit would be even greater if the Fed stopped paying interest on reserves. (I am assuming you do not think that would produce a big boost to NGDP–Pietro, commenting above, thinks it would but I know lots of folks do not.)
So as I see it, the Fed is in a win-win situation: It buys Treasury debt like gangbusters. If NGDP starts to move, then QE does what we all hope it would. If NGDP remains insensitive to who holds the federal debt, then at least we make a significant dent in the budget deficit by reducing interest expense. What is the downside?
Re: the money multiplier, that's interesting about 2001-05 and I didn't know that it was so stable then. I assume that's the M2/M0 ratio? That is very stable and suggests that Greenspan's reforms did less to make the money multiplier unpredictable than other experts have claimed. (There's a paper on this topic by the Fed, I think Sack lead author. The different conclusions may have to do with varying definitions of the multiplier.) But I still maintain that the radical changes in the money multiplier's behavior since 2008 have been the result of policy changes. Granted those policy changes (IOR, LSAP) were responses to stress, and in theory the stress alone would have had important effects.
A few points on your proposal for the Fed to buy Treasurys "like gangbusters":
– QE doesn't cut interest costs any more than Twist. If you look at the Fed and Treasury as a pair and consolidate their financial transactions, with QE they issue reserves paying 0.25%, with Twist they issue short bonds paying about the same. The main difference is that QE increases M2 while Twist doesn't.
– If QE "like gangbusters" means up to the level of current Treasury net issuance, that still wouldn't save Treasury much in interest costs, and even in the unlikely situation that the savings were spent, I don't think it would spur significant inflation. Although I admit I can't be certain because we don't really have any comparable situation from history to guide us. M2 would grow by $1.2-$1.3 trillion/year.
– If QE "like gangbusters" means a coordinated program with Treasury to monetarily fund a fiscal expansion, that's politically impossible, but yes if it could somehow happen it would spur spending and increased NGDP.
– A mostly inflationary boost to NGDP would in my opinion not be beneficial. The bailout of fixed-rate, long-term debtors from higher NGDP would be outweighed by the penalty to variable-rate debtors and the discouragement to would-be debtors from higher interest rates to the non-financial sector. In general I think this kind of approach, to attempt to create inflation in order to bail out a class of debtors, is nutty. It would make much more sense to legislate a fiscal bail out or better yet in my opinion restructuring with an element of forgiveness. Unfortunately even Obama is very afraid of taking on the banks and other financial interests that would have to take losses.
"A mostly inflationary boost to NGDP would not be beneficial."
I would be less enthusiastic about if it I thought it would be mostly inflationary. But why would it be mostly inflationary? Are you among those who think there is currently zero output gap? Even if there is no output gap, an aggregate-demand driven expansion above potential GDP is normally split between real output and inflation in the short run.
PS On the idea of cutting the IOR rate to zero, what we've seen from Europe so far suggests that this puts strong downward pressure on the FX rate, but also threatens the viability of money market funds. It's too early to judge how it is effecting lending but my expectation is almost zilch. Europe is a very different situation, where liquidity measures tend to merely facilitate capital flight by letting people cash out of the periphery without losses. There is also an argument that the money saved by the Fed/Treasury from not paying IOR would come out of the pockets of demand depositors who would have to pay higher banking fees.
QE will continue until it is obvious that it is not having an effect on the general economy and we are close. It will always have an effect on equity markets and banking systems, but at what costs? Scare tactics by the Fed's PR machine will keep the presses running as long as possible.