If you've been in finance circles online you hear the Fed described as the "lender of last resort". Smart people with careers in finance. They're incorrect.

Sometimes it takes a scholar, such as the illustrious money man Perry Mehrling, to correct the meatheads on Wall st.. I've been following Perry's work for the last 5 years or so. I took a course of his online (from where, I can't remember), which covers essentially the same content as the book we're discussing here: The New Lombard Street.

There are maybe three people who publicly understand financial plumbing. That would be Perry, Conks, and Zoltan Pozsar. I would hope some people inside the Fed as well.

"The New Lombard Street"' is an ode to the book "Lombard Street" by Walter Bagehot, who was a banker and one of the first writers on what you might call financial plumbing. Perry very obviously loves Bagehot and owes his intellectual lineage to him.

Bagehot's advice in his book was that central banks should be lending freely to solvent firms, against good collateral, at high rates. The most brittle part of the financial system during times of crises for Bagehot et al was "funding liquidity". Because in his time and place, the banks held short-term commercial paper. A British merchant ships cotton from America to Liverpool, a bank finances the voyage, the cotton sells, the bill pays itself off. These were considered 'self-liquidating' in that they would naturally convert to cash once the underlying transaction completed. The funding liquidity problem was about panicked depositors, but the solution was that the banks just needed to bridge the gap. In this case, central banks were the lenders of last resort.

That world is gone. The critical shift that Mehrling traces is from "commercial loan theory" to "shiftability". The biggest factor was actually WWI. After WWI, the balance sheets of the banks were stuffed with government bonds, because someone needed to fund the war. This is long-term debt that doesn't self-liquidate on any useful timeframe. So when you need liquidity, what do you do? You sell them. The liquidity is no longer intrinsic to the asset, it depends entirely on there being a functioning market to sell into.

Moulton first articulates "shiftability" in 1918, and it's clearly the right framework. Everyone can see that banks are holding government war debt now, not self-liquidating commercial paper. But the Fed's enabling legislation was written around commercial loan theory. The institution was designed for a world of real bills. You get this awkward period where practice has moved on but doctrine hasn't.

Then the 1920s happen. The Fed, trying to be helpful, smooths out seasonal and cyclical fluctuations in money rates. Low, stable interest rates. Sounds great, right? Except it unintentionally supports an asset price bubble. The intertwining of money markets and securities markets—which Mehrling emphasizes is the defining American feature—means you can't stabilize one without affecting the other. Cheap funding liquidity translates into speculation. The Twenties roar, and then crash.

BUT! The craziest insight from this book is almost a footnote

The important point, for our story, is that under American conditions, the money market and the securities markets have always been completely intertwined and, as a consequence, it has never been possible to distinguish speculative from productive credit. This intertwining predates the Fed, having its origins in bank reliance on shiftability under the National Banking System

This insight basically invalidates 80 years of monetary policy debate. The Taylor Rule types, the inflation targeters, the whole modern central banking framework... all of it assumes you can separate "monetary policy" from "asset prices."

This helps explain a lot of otherwise unexplainable phenomena:

• Why QE was supposed to stimulate the "real economy" but mostly inflated asset prices.
• Why tightening to fight inflation also crashes markets and causes bank runs
• Why the 2000s housing bubble happened right after the Fed cut rates to cushion the dot-com crash
• Why tightening to fight inflation also crashes markets and causes bank runs (2023 SVB)

There doesn't seem to be a stop on the elevator: only up or down.

Continuing on with our chronology...

The Depression is basically the shiftability doctrine failing its first real stress test. Everyone tries to sell at once and there's nobody on the other side. Market liquidity evaporates. The 1935 Banking Act finally codifies shiftability doctrine. The Fed will now accept any "sound" asset, not just commercial paper.

Then WWII makes it all moot. The Fed basically becomes an arm of Treasury until 1952, pegging bond prices to finance the war. Private capital markets disappear entirely.

In 1952, McChesney Martin establishes the "bills only" framework. The idea was that the Fed handles short-term money markets. Private dealers handle everything else. Funding liquidity provided by the Fed would translate into market liquidity through private arbitrage, but it rested on the assumption that dealers would always want to make markets!

This assumption actually held a surprisingly long time, 56 years until 2008, when the assumption completely shatters with the collapse of the Asset-Backed-Commercial-Paper market, Lehman Brothers, etc, and dealers stopped making markets.

A brief aside about Dealers.... they are middleman with balance sheets. They post bid and ask prices and pocket the spread. When you want to buy your mortgage backed securities, you don't sell directly to another willing buying, you pass through dealers first. This is unlike the housing market, where sellers and buyers go direct and inherit the timing risk. (I suppose not if opendoor becomes a dealer in the near future). Dealers are the hidden infrastructure that makes markets liquid.

The critical thing is that dealers are not charities. They make profits in normal times because the bid ask spread compensates them for holding inventory. Here's what happens in the crisis - "subprime MBS, who the fuck knows what those are made of, lemme just widen the bid ask here to 90c/101c". And now you have a problem, because the entire banking operation requires shifting assets around to meet obligations. When dealers step back, there's no one to shift to.

This is the thing the funding-liquidity crowd doesnt get. The Fed can pump unlimited reserves into the banking system. Great. The banks are flush with cash that they can use to lend, which will make its way into an asset bubble somewhere. But if the dealer aren't making the markets to shift assets, it doesn't matter, the cash sits there in MMFs. Funding liquidity and market liquidity are not the same thing. That's why the "bills only" framework worked for 56 years until it didn't.

In 2008, the Fed had to abandon bills only policy to avoid the complete destruction of the financial system. It started actually buying MBS outright and became the dealer of last resort as a matter of necessity, and put a floor under prices.

In March 2020, it backstopped the market again. Mehrling's point is that we need to actually think through what it means for the central bank to backstop market liquidity, not just funding liquidity. What assets? At what prices? Under what conditions? The Bank of England figured out lender of last resort doctrine after they'd been doing it for decades.

We still don't really have a framework for all of this. The Fed has acted as dealer of last resort twice now in twelve years, and most certainly will again in the next crisis. But the policy debates and bloomberg are still talking about the federal funds rate like it matters. And financial experts are arguing about QE happening in Japan and America as if they have the same outcomes, because the plumbing is the same, but this is a lazy conclusion.

The conceptual frameworks in the minds of everyone but the plumbers are two generations behind reality. The rest of finance is still arguing about a system that stopped existing in 2008. 5/5, thank you Mehrling!