Janet Yellen and Daniel Tarullo are probably OK with this.

Banks Turn Cash Into Mortgages and Back Into Cash

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

Here is a quite wonderful post by Tracy Alloway at FT Alphaville about the Federal Home Loan Banks and the Liquidity Coverage Ratio. The issue, as it always is, is that:

  • everyone wants banks to be safe, but
  • everyone wants banks to do banking stuff.

The particular flavor of "safe" here is: Look, the big systemic risk of banks is runs on the bank. The way (a way) to reduce the risk of runs is for banks to just have a bunch of cash lying around. That way, if everyone comes into the bank yelling for their money back, you can just give them their money back.

But of course there'd be no point to a bank that just has a bunch of cash lying around. The whole point of a bank is to take cash from depositors and use it to do stuff. The particular flavor of stuff at issue here is mortgage loans. Banks take cash from depositors and lend it to people to buy houses. That's good, that's what they're supposed to do. But then the banks aren't safe, because if there's a run on the bank all the money is in mortgage loans, and you can't give depositors back mortgage loans, there is actually a movie about that.

So obviously what you should do is find a way for banks to take in cash, use it to make mortgage loans and then magically transmute it back into cash. And the U.S. financial system, being wonderful, has a way to do just that. It's called an advance from a Federal Home Loan Bank. Basically the FHLBs are government-sponsored enterprises that lend banks cash, at very low subsidized variable rates, secured by the banks' mortgages. The FHLBs get the money by selling bonds, which are implicitly government guaranteed.

So the system is, schematically:

  1. Depositor deposits $1 at bank.
  2. Bank makes $1 mortgage loan.
  3. Bank borrows $1 against mortgage loan from FHLB.
  4. Bank has $1 in cash.
  5. FHLB borrows $1 from the market for safe agency bonds.

Here is a fake balance sheet for the system:

The magical result is that:

  • from a depositor's perspective, the bank is safe, because it has lots of cash;
  • from a mortgage borrower's perspective, the bank is nice, because it is giving the borrower a loan.

Everyone wins! And this is not a sneaky piece of evildoing. This is what banking is. The point of banking is to conceal risk. Banking is the business of issuing risk-free liabilities and using them to purchase risky assets. That's the whole business. The system exists to perform the magic in the lower-right-hand corner of that diagram.

You can get outraged about this but to what end? Imagine that banks couldn't do this. Then they would either:

  1. make fewer mortgage loans (bad!),
  2. have less liquidity and be more vulnerable to runs (bad!), or
  3. find some other way to transform risky mortgages into safe cash.

I didn't supply a "(bad!)" at the end of that last option, but you can probably supply your own. "Let's package our risky mortgages into mortgage-backed securities and bamboozle ratings agencies into giving those securities AAA ratings so that investors think they're risk-free," is more or less a thing that people thought and then did and then regretted. At least with the FHLBs, the magical transformation is conducted by a willing participant and explicitly blessed by the government.

Generally speaking there are about three ways for banks to perform their magic function of transmuting risky assets into safe liabilities. One is to, like, be good at the job: take in deposits and use them to make careful diversified loans while managing liquidity carefully. Another is to trick people: just tell them that stuff is safe even when it isn't. And the third is to rely on government guarantees, since the best way to ensure that liabilities are safe is to be the issuer of the currency in which they're denominated.

That last one is pretty controversial -- why are the banks gambling with taxpayer dollars? -- but it is also the only entirely reliable one. The point of banking is to conceal risk, but banks often need the government to help with that. Fortunately the government often obliges! The widespread sense that banks relied too heavily on trickery before the last financial crisis -- and on government backstops during that crisis -- makes it hard for anyone to trust them to do the safety transmutation on their own these days. So you can't be too surprised that, when they're told to do a better job of managing their liquidity, they end up falling back on governmental backstops to do it. If you have access to magic that works, why wouldn't you use it?

  1. See page 306 of this JPMorgan 10-K, which shows $61.9 billion of FHLB advances, at (mostly floating) interest rates ranging from 0.16 to 2.04 percent.

  2. The FHLBs are government-sponsored entities, and while they are not officially guaranteed by the U.S. government, everyone assumes "that the FHLB system would get bailed out if it ever ran into trouble," as Alloway puts it. A point that I gloss over in the text is that the "cash" that I describe isn't just "cash"; you can count high-quality liquid bonds as being basically like cash. In fact you can count FHLB bonds as being basically like cash, albeit they're only "Level 2A Liquid Assets" and so subject to a haircut and a cap. See pages 69-75 of the Fed's release on the liquidity coverage ratio.

    While you're there you might as well also read pages 257-259, on outflow rates. The idea of the LCR is that you need to have enough liquid assets to cover your potential outflows in a crisis: The more you rely on short-term risky funding, the more liquid assets you need. FHLB advances can be short-term risky funding, so you have to assume that they will partially vanish in a crisis. Except that FHLB advances with a maturity greater than 30 days don't count as short-term risky funding -- and as that JPMorgan 10-K makes clear, most of the big banks' FHLB funding is longer than 30 days.

    Incidentally! All of this is separate from the other big mortgage-subsidy program, the Fannie Mae and Freddie Mac program of guaranteeing mortgages and plopping them into (liquid, government-backed) agency securities that can then serve as high-quality liquid assets. In the FHLB version, the bank retains the credit risk. But if you don't think too hard about it the two things can have similar LCR effects.

  3. I ignore that the FHLB advance will come with some haircut. (Also capital requirements, etc.) The point is just schematic. But actually the magic can be even deeper than that. See footnote 2: The bank can count the FHLB bonds as being high-quality liquid assets, albeit at a 15 percent haircut. So the FHLB can get its funding from the bank. So the balance sheet of the bank can look like this:

    [imgviz image_id:iaFDSBSZXNvE type:image]

    And the FHLB is doing nothing in this scenario -- the bank gives it money and gets money back. This is an exaggerated oversimplification but a fun one to consider, because it elides the economically uninteresting role of the FHLB bond investors. The bank does it all itself; the FHLB's only role is as the dispenser of magic pixie dust, turning liquidity-risky mortgages into safe FHLB bonds.

  4. You don't have to be. Alloway's outrage is different and more interesting:

    But we can’t help but feel a lingering sense of righteous disbelief that the biggest banks are taking advantage of low borrowing costs that come about through an FHLB system that carries an implicit government guarantee to satisfy the new LCR rules. That’s worth remembering the next time the biggest banks claim they don’t receive funding advantages over smaller competitors.

    The point is that the use of FHLBs to satisfy liquidity requirements seems to be concentrated among the big banks; there's a report from the Federal Housing Finance Agency's inspector general to that effect. One question you might ask is why the smaller banks don't get in on this more. Part of it is structural; bigger banks have access to FHLBs in different jurisdictions while smaller banks are more limited. But I suspect that bigger banks are also just more focused on, and better at, optimizing funding than smaller banks are. (That's why they're bigger!) Which may explain part of the too-big-to-fail subsidy on its own.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Matthew S Levine at mlevine51@bloomberg.net

To contact the editor on this story:
Zara Kessler at zkessler@bloomberg.net