I’ve been thinking a lot about bank reform lately. Here’s the core of the problem: deposits are sticky in ordinary times, particularly once you have a guarantor of deposits like the FDIC. But for some banks, they look to other short term funding, whether it is short CDs or repo funding.
Now to me a lot of the issue is asset-liability mismatch. Banks borrow short and lend long. That leads to banking panics. Financing illiquid assets with liquid liabilities is unstable, and begs for bankruptcy at the first significant loss of confidence.
But there is a greater mismatch present, which I want to explore. Every asset is financed with some liability or equity. And, every liability is someone else’s asset, but not vice-versa, because assets owned free and clear are equity-financed.
Assets financed by debt are frequently mismatched short. Long mismatches are rare because of the cost of financing being too high. Now, if short mismatches are small, that’s not a problem. There is enough flexibility in financial balance sheets to accept small mismatches. Real disasters happen when long assets are financed in such a way that there is a risk that the financing will fail prior to the assets being paid off.
The fundamental mismatch in debts that finance assets is that the ultimate assets being financed are longer-dated than the financing. We fund land, houses, buildings, plant & equipment, and do it off of deposits, savings accounts and CDs. Some financial companies finance off of short-dated repo funding. The reason that this mismatch is hard to avoid is that average individuals who save want short-dated assets that can be used for transactions. That doesn’t fit well against the need to fund long-term assets.
The same problem exists in the municipal bond market. Much more money wants to invest short, while municipalities want to borrow long. This leads to a steep muni yield curve. Commercial insurers writing long tail business, and wealthy people that can tolerate interest rate volatility end up buying the long end, and lower taxes in the process.
If banks were required to approximately match cash flows for assets not financed by equity, yield curves would steepen for other areas of the fixed income markets. Areas of the financial market where there are long/strong balance sheets, such as Life Insurers, Commerical Insurers, Defined Benefit Pensions and Endowments would get higher yields for longer commitments. Banks would become a lower ROE business, and that would be good, as there would be many fewer failures, and there would be fewer banks; we are over-banked. Time to re-educate bankers for more productive activities.
Long dated floating-rate loans could be a solution for banks funding loans off of short-dated lending, or, using interest rate swaps to achieve the same result. The risk is that a bank locks in what proves to be a low spread on the asset, while funding costs are volatile.
A few final notes: 1) the standard of broadly matching asset and liability cash flows should be applied to all regulated financial institutions, including investment banks. Only surplus assets not needed to match liabilities can be used for investments with equity-like risk. 2) There must be an unpacking of complex vehicles with embedded leverage to do the Asset-Liability management. As examples:
would need to be reflected as looking through to the items ultimately financed. As an example, the AAA portion of a senior-sub securitization is long the loans, and short the certificates sold to the rest of the deal.
Repo funding has its own issues. In a crisis, haircuts rise as asset values fall. Institutions relying on that funding often fail at those times, and leaves losses to the repo lender. There would need to be something reflected for the risk of repo market failure, though the grand majority of the losses go to the borrower, and not the lender.
3) Even short lending to those getting loans that do not fully amortize should be reflected as loans that are longer-dated, because of the risk of rates being higher, and refinancing is not possible.
I have more to say, but I’m going to hit the publish now. Comments are welcome.