Tilting At Windmills: The Faustian Folly Of Quantitative Easing

As I explained in my last post, banks can’t “lend out reserves” under any circumstances, which undermines a major rationale that Central Bank economists gave for undertaking Quantitative Easing in the first place. Consequently, the hope that Bernanke expressed in 2009 is “To Dream The Impossible Dream”:

To dream the impossible dream
To fight the unbeatable foe

To bear with unbearable sorrow
To run where the brave dare not go

But without the poetry:

Large increases in bank reserves brought about through central bank loans or purchases of securities are a characteristic feature of the unconventional policy approach known as quantitative easing. The idea behind quantitative easing is to provide banks with substantial excess liquidity in the hope that they will choose to use some part of that liquidity to make loans or buy other assets. (Bernanke 2009, “The Federal Reserve's Balance Sheet: An Update”

What a folly this was—almost. The one out that Bernanke gives himself from pure delusional babble is the phrase “or buy other assets”—because that’s the one thing that banks can actually do with the excess reserves that QE has generated.

But rather than rescuing Central Bankers from folly, this escape clause is an unwitting pact with the devil: they are now caught in a Faustian bargain. Any attempt to terminate QE is likely to end in deflating the asset markets that it inflated in the first place, which will cause the Central Banks to once more come “riding to the rescue” on their monetary Rocinante.

While Central Bankers can personally still join Faust and ascend to Heaven—thanks to their comfortable public salaries and pensions—the rest of us have been thrust into the Hell of expanding and bursting speculative bubbles, hoist on the ill-designed lance of QE. Bernanke is a rich man’s incompetent Frank N. Furter: confronting a wet and shivering couple, he promises to remove the cause—but not the symptom:

So, come up to the lab,/ and see what's on the slab!/ I see you shiver with antici...  ...pation./ But maybe the rain/ isn't really to blame,/ so I'll remove the cause.../ [chuckles] but not the symptom. (“Sweet Transvestite”)

Bernanke’s QE instead maintains the symptoms of the crisis, but does nothing about its cause. It generates rampant inequality, drives asset prices sky-high and causes frequent financial panics—while doing nothing to reduce the far too high a level of private debt that caused the crisis.

Let’s put QE on the slab in my lab—my Minsky software—and track the logic of the monetary flows that QE can trigger. Figure 1 shows the full model; the Tables in this post go through it step by step.

Figure 1: The full Minsky model of QE (with no simulation settings as yet)

[Here’s a quick primer on Minsky’s accounting conventions for those who haven’t read the previous post. Minsky shows Assets as positive sums, and Liabilities and Equity as negative, so that any operation on a row sums to zero: therefore, positive entries increase Assets, and reduce Liabilities and Equity; while negative entries reduce Assets, and increase Liabilities and Assets. This might look strange on first glance, but it (a) enforces the “Fundamental Law of Accounting” that “Assets minus Liabilities equal Equity”, since when a transaction is properly recorded, each row sums to zero; and (b) since Minsky maintains interlocking “Godley Tables”—named in honour of Wynne Godley—it shows a Liability from one sector’s point of view as an Asset from another’s. The program can thus accurately track how money flows through the financial system.]

I make some genuine simplifying assumptions to make the Godley Tables simpler to read (genuine in the sense that my results don’t depend on the assumptions—unlike so-called “simplifying assumptions” in much of Mainstream economics). I treat QE as a loan (rather than a bond purchase), and I ignore all other transactions in the economy apart from those related to QE.

The first step in the model is the Central Bank loan of QE to Private Bank 1 (PB₁). That increases the Assets and Liabilities of the Central Bank: Loans_PB rise by QE, as do Reserves_B!. Since Central Banks now pay interest on excess reserves, I include the flow of interest payments Int_ER as well. This is financed from the essentially limitless equity of the Central Bank, and—from its point of view—adds to the Reserves of Private Bank 1 (if the rate is positive). These operations are shown in Table 1:

Table 1: QE and interest payments on Reserves from the Central Bank’s perspective

These same operations are recorded on Private Bank 1’s accounts as shown in Table 2. This confirms one of Joe Stiglitz’s points about QE from his article that I otherwise criticized in my last post: Banks have indeed been earning “earning nearly $30 billion – completely risk-free – during the last five years” thanks to QE.

Table 2: QE and interest payments on Reserves from the recipient Private Bank PB₁’s perspective

Now what happens to the QE funds once they’re in Private Bank 1 (PB₁ for short)? It can’t simply lend them out, as I showed in my last post—the belief that they can violates the Fundamental Law of Accounting. But PB₁ can buy assets with it, which I’ll show here as a purchase of shares from a broker who banks with another Private Bank PB₂. This does of course decrease PB₁’s reserves—mission half-accomplished.

Table 3: Buying shares from PB₁'s perspective

But Oh Dear! This increases PB₂‘s reserves by precisely the same amount as PB₁‘s reserves have fallen—see the first row of Table 4. However, finally, the QE money gets into the money supply, because by buying shares off the broker, the liabilities of the banking system rise, and therefore money has been created—money that mainly goes to the wealthy people (the Super-Rich) who generally own shares, from whom the brokers buy more shares—shown in the second row of Table 4.

The Super-Rich spend of course, but more slowly than anyone else relative to their wealth, simply because they have so much more of it than anyone else. And they spend a lot of that on each other—buying and selling assets and luxury possessions. Given the slower speed at which they spend, the money created by QE dwells here the longest—and therefore their wealth grows more than that of any other social group thanks to QE.

Table 4: PB₁'s purchase of shares from PB₂'s perspective

The inflow of money to the broker triggers two other flows that reduce PB₂‘s reserves: paying dividends back to PB₁ (conflating the broker and the company whose shares have been bought, again to save adding another column) and paying wages and bonuses to their staff, whom I assume bank at PB₁–and are Rich rather than Super-Rich. The first two rows of Table 5 show these operations from the point of view of PB₁.

Table 5: Payment of dividends and bonuses from PB₁'s perspective

Oh Dear! The reserves that PB₂ got rid of by paying dividends and bonuses out have ended back at PB₁. There has therefore been no net change in the aggregate level of reserves. But, finally, there is an operation that puts money into the hands of poorer households: the Rich households use part of their Bonus to buy something from the Poor households (far be it from me to insinuate that this might sometimes be something illicit!).

Now let’s take stock—to coin a phrase. All these operations have done nothing to reduce the aggregate level of Reserves: they’ve simply shuffled them from one bank to another. In fact, even if QE stops, reserves grow at the rate of interest on reserves.

Banks come out of this as rather innocent: they can’t do what Ben Bernanke (and now Joe Stiglitz) are berating them for: not only can’t they lend out Excess Reserves, they can’t get rid of them either. It’s not their fault that the only way they can individually attempt to do so amounts to playing “hot potato”— trying to reduce their reserves while simultaneously boosting another bank’s reserves—by buying assets, since this is what Don Quixote (sorry, Bernanke) intended them to do in the first place. However, these misunderstood institutions are nonetheless “crying all the way to the bank” with the risk-free earnings that QE has given them.

The only ways that Reserves can fall are (a) if the Central Bank takes them back or (b) if some part of the public withdraws its money from deposit accounts as cash. Under the dynamics of QE itself, Reserves can only rise (so long as the interest on Excess Reserves is positive—which may be one unstated reason why Central Banks are implementing negative rates now, if anyone inside one of them has worked this out).

QE gets into the money supply—not via lending, which is impossible, but via asset purchases, which far and away benefit rich households more than poor ones. Rich households also benefit from the income the share transactions generate. And finally, some of that money gets to poor households when the rich ones—made richer still by QE—buy some services off them.

The real economy has thus received some impetus from QE, but only a relatively trivial amount of the money created has got into circulation in Main Street. As Michael Hudson puts it, Bernanke’s helicopter dumped money on Wall Street, not Main Street.

The bubble before the financial crisis had already exaggerated income inequality past what is sustainable in a capitalist society. Central Bank meddling via QE has made this problem worse, and without the illusion of a boom (like the Internet and Subprime Bubbles) to make it seem somehow palatable.

It has done this, not only by giving money to the Super-Rich, but by inflating asset prices, driving them well above the increase in consumer prices and wages. This is yet another symptom of the bubbles before the crisis, when speculative lending by banks drove asset prices (shares in the DotCom Bubble, house prices in the SubPrime Bubble) into the stratosphere in the first place. The S&P 500 fell from 45 at the peak of the DotCom Bubble to around 25 when the SubPrime Crisis began, and 15 in the depth of the crisis—see Figure 2. QE drove it back up to 25.

Now Stockmarkets worldwide are falling again, partly in response to the The Fed’s tentative and foolish attempts to unwind its original bold and foolish intervention into inflating asset prices. But it will be dragged back in again if it attempts to unwind QE, because all the dynamics explained in the preceding Tables works in reverse as well. So long as it continues to believe that sky-high asset prices are actually good for the economy, it can’t afford not to continue meddling in them.

Figure 2: Robert Shiller's Cyclically Adjusted PE Ratio

Figure 3: Shiller's long term real house price index

All the while, the Fed has not only ignored the real cause of both the Asset Price Bubbles and the crisis itself—the private debt bubble that financed the DotCom and SubPrime Bubbles—its Impossible Dream was that QE would cause this debt bubble to rise again too. Quoting Bernanke from 2009 again, “The idea behind quantitative easing is to provide banks with substantial excess liquidity in the hope that they will choose to use some part of that liquidity to make loans”.

Bernanke and his Central Bank colleagues around the world have made the symptoms of the crisis worse, in the hope that the cause would get worse, thus curing the patient.

Take a good look at Figure 4. Anyone who thinks than more private debt in America is a good thing is as delusional as Don Quixote.

Figure 4: Estimated long term US private debt to GDP ratio using Fed data since 1945 and Census data since 1834

The Man of La Mancha’s magnificent song “The Impossible Dream” ends with this stirring paean to chivalry:

And the world will be better for this
That one man, scorned and covered with scars,
Still strove with his last ounce of courage
To reach the unreachable star.

Central Bankers, the Don Quixote’s of our modern age, are making the world a far worse place than it would be without their foolish gallantry.

PS: I note that Frances Coppola states on her Forbes blog that lending can reduce excess reserves—which is quite correct (“It Was The Financial Crisis That Stopped Banks Lending, Not Interest On Excess Reserves”). If banks lend and have excess reserves, then part of these become required reserves and excess reserves fall. But the total amount of reserves remains constant—as Frances and I both observe. So bank lending can reduce excess reserves, but banks still can’t “lend out reserves” themselves, as Frances explains there and as I did in my previous post (“Hey Joe, Banks Can't Lend Out Reserves”). And Frances is also quite right that the financial crisis brought bank lending to a halt,  not interest being paid on excess reserves.

Technical Appendix

Here’s a screenshot of a “quick and dirty” simulation using this model, including the ability to turn off or reverse QE. It would be fun to add an asset price component here (assume fixed supply and the flow of new money from QE driving price change) and see what happens.

If you’d like to check this out, here are the links: Minsky model before simulation added (Model; LaTeX code; Matlab code); model after simulation added (Model; LaTeX code; Matlab code). To run Minsky, download it from here. The current Windows-only beta build overcomes speed issues with Godley Tables in the current release version (but it might introduce some other bugs)

The equations for the model (generated by Minsky as LaTeX output from the file menu) are: