“We don't understand fully how large-scale asset purchase programs work to ease financial market conditions.”
- New York Fed President Bill Dudley speaking as the Fed began its QE3 “taper” in 2014
Judging by conflicting opinions that we read and hear about the Fed’s quantitative easing (QE) program, Bill Dudley is far from alone in being confused about the program’s effects. Pundits disagree about the effects of QE on inflation, growth, and financial markets. And even worse, they disagree about the nuts and bolts of how the Fed implements QE.
Fortunately, though, cutting through the QE fog is easier than you might think. I’ll argue that one research result can take us a long way towards understanding what QE does and doesn’t do. (Spoiler: More “doesn’t” than “does.”) And in a note at the end of the article, I’ll point readers in the right direction as far as how QE is implemented, which is important because of all of the misinformation that’s been pushed into the public domain.
The research I’m describing is this: I netted all lending by banks and broker-dealers and compared it to the Fed’s net lending. And what else? Nothing, that’s all there is to it. But the chart that emerged from the research is, in my opinion, about as clear a look at QE as you’ll find. It shows that we should think of QE as little more than an “argyle effect”:
Apart from conjuring your favorite Burberry sweater, the key message is that QE was far less relevant than many people believe. Policymakers merely replaced growth in privately financed credit with growth financed by the Fed. They grabbed the credit growth baton for QE laps and returned it to the private sector for QE pauses, and whoever didn’t have the baton more or less stood still. As I concluded when I first produced the chart in 2014, QE is a substitution story, not an addition story.
Many pundits told the addition story as QE was underway. They expected banks to “multiply up” reserves by aggressively expanding their loan books. But reserves never significantly multiplied. I think there are five reasons why the “money multiplier theory” failed:
- High-quality borrowers don’t emerge mysteriously from cracks in the Eccles Building and parade zombie-like to bank loan desks. In other words, credit demand was probably about the same with or without QE.
- QE’s effects on bank balance sheets aren’t quite as distorting as they’re often depicted. Consider that new reserves are typically matched by new deposits, because dealers offering bonds to the Fed get paid for those bonds through their accounts at commercial banks. In other words, QE adds a similar item to both sides of bank balance sheets, which you might not appreciate if your information comes from those who call for banks to “lend out” reserves. That’s impossible - reserves can’t be “lent out” - and it often leads to exaggerated statements about the implications of excess reserves. (See my second author’s note below for further explanation.)
- To a significant degree, banks can neutralize excess reserves (and the corresponding “excess” deposits) with financial derivatives and other balance sheet adjustments. They can rearrange exposures to mimic a balance sheet of equal risk that’s not stuffed with reserves.
- Just as importantly, excess reserves flow naturally from banks that don’t want them to banks that don’t mind them nearly as much. Consider that Fed data shows a disproportionate amount of QE’s extra reserves landing at U.S. branches of foreign banks. Those foreign banks might have sound reasons for holding excess reserves.
- The money multiplier theory is inconsistent with real-world reserve management practices. The Bank of England has called it "reverse" to how bank lending and reserve management work in the real world. And the gap between theory and reality is so large that you don’t even need the four reasons above to reject the money multiplier - you just need a healthy skepticism about mainstream theory.
According to my chart, even QE’s wealth effects appear to be poorly understood. If credit growth is the same with or without QE, any effects on bond and stock prices might be more psychological than commonly believed. Or, those effects might transmit mainly through financial derivatives (see #3 above). Or, I hear at least a few readers asking, “What wealth effects?” We’ll never know for certain if QE boosted asset prices at all. Maybe the bull market only needed low interest rates, a slowly growing economy, the knowledge that our policy honchos wanted asset prices higher, and a soothing narrative that they have the tools to make that happen?
Think of it this way: New borrowers know approximately how many calories they can consume, and after the Fed starts delivering three meals a day, private banks find that their contributions are no longer needed. By necessity, private banks shut down their kitchens, and almost nothing changes economically. We get substitution, not addition.
To be sure, the argyle effect settles a few important QE debates but not every debate, and it doesn’t tell you exactly where to invest your money. Further, people may still believe that QE increased the overall lending trend (referring to the entire period’s lending growth), irrespective of the pattern from one QE to the next.
For now, though, I’ll refer readers to my past and future articles for discussion of investment implications and related topics. It seems better to stick to a single purpose here, which was to share the clearest result I can produce when it comes to understanding QE’s consequences over the past eight years. That, to me, is the first step to battling through a sea of misinformation about how QE affects the economy and financial markets. Understand the Burberry backlash, and you’ll be well equipped to evaluate what might or might not happen as QE unwinds.
Author’s note: We separated QE and non-QE periods according to the credit the Fed added to the financial system through all of its activities, not just open market operations. Because we included loans, repos, and various emergency facilities enacted during the financial crisis, our time periods are slightly different than the announced start and end dates for QE alone. (For more details, see this note from 2014, although replicators should be aware that the Fed recently replaced its “credit market instruments” category with a few subcategories.) In other words, the line showing the Fed’s net lending is jagged by design. By separating periods of high versus low Fed-sourced credit, we can test whether the private sector’s net lending would show a reverse correlation to the Fed’s activities. As you can see, it did.
Second author’s note: In a series of articles published recently on a heavily trafficked financial site, we can find several of the common misconceptions about banking practices and QE. The arguments in those articles relied on the assertion that “the Federal Reserve conducts open market operations or quantitative easing (QE) by crediting the reserves of primary dealers (banks) in exchange for securities, namely treasuries and mortgage-backed securities.” The articles further held that QE has no effect on M1 and M2, and the author also relied on the notion that banks “lend out reserves,” as though reserves deplete when a bank extends a loan. Let’s compare each of those claims to the real world.
First, primary dealers don’t hold reserve balances at the Fed, because they’re not deposit-taking institutions. In other words, dealers and banks aren’t one and the same as in the excerpt above - they’re two completely different entities. Even when a dealer and a bank are part of the same holding company, they have different business models and perform different functions. And there’s no such thing as a primary dealer “reserve balance,” which you can confirm in many places, but the Fed’s Z.1 report that was used to create our chart is as good a place as any. You won’t find “reserves” in any of the Z.1 data on broker-dealers.
Second, when the Fed “quantitatively eases” by buying bonds from a primary dealer, the dealer receives a credit on its commercial bank account, and the Fed simultaneously credits the commercial bank with reserves. Therefore, it’s wrong to say that QE doesn’t affect M1 and M2 - those money supply measures rise with the primary dealers’ bank account balances. That’s why commentators say that QE is equivalent to “printing money,” and according to their definition of printing money, they’re exactly right. At the same time, the argyle effect I discussed above shows that the activities of private banks can offset the Fed’s money printing, such that overall money supply growth is modest or nonexistent. But that’s a different issue - QE feeds directly into M1 and M2 as commonly and correctly understood.
Third, only the Fed can change the amount of reserves in the system. Banks can’t “loan out reserves” - that’s impossible. When banks make loans, they deliver the proceeds by crediting the borrower’s account or producing a cashier’s check or creating some other form of money purchasing power. Their reserve balances don’t change. Of course, banks can ask the Fed for more reserves in return for assets, but that would be a completely different transaction.
There are plenty of places to confirm this information, but also much misinformation, which is why fallacies persist. For example, a recent, widely read article linked to a page on the Fed’s website claiming that it doesn’t “print money.” Readers of that webpage should recognize that the Fed defined “printing money” as permanent debt monetization by a profligate government, not as increasing M1 and M2. Also, they should understand that the webpage exists to counter critics who argue that the Fed prints too much money. A brief PR exercise by the Fed isn’t the best source of information if you’re trying to confirm how open market operations work.
So what is the best source of information? When it comes to the mechanics, one source that’s very good is this old document published by the Chicago Fed, although you have to be careful about the theory sections, which reflect a “reverse” view to reality as discussed above. You can also confirm my points above on this New York Fed web page. Also, my mention of the Bank of England referred to a report that the “Old Lady of Threadneedle Street” published to clear up misconceptions found in textbooks, commentaries, and blogs. You can find that report at the link in #5 above. Finally, my book Economics for Independent Thinkers covers the economic implications of different types of money creation in more detail than the other sources I’ve mentioned.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.