The reasoning behind this somewhat random article was that I am considering a chapter in my future banking primer. My inflation primer manuscript is currently cowering in fear of Brent prices.
Public Banks
I will need to look at the literature to get a better handle on “public banks,” but I am aware of two models. The first is a national bank, often in the form of “postal banking” (a savings-oriented bank where you can use post offices as branches). The second are local banks owned by sub-national governments. The usual objectives for the government owning a bank is to ensure that banking facilities are open to all members of the public, that there be lending for publicly desired activities, and possibly general anti-bank populism.
For this article, I am only considering local banks, since postal banks are agents of national governments, and national governments have many tools to interact with sub-national finance.
How Many Bonds Could a Local Public Bank Buy?
The usefulness of a public bank for a state or municipality’s finances depends upon how many1 bonds it can buy. The amount is driven by the following equation (definition):
Amount of Municipals Bought = (Balance Sheet Size)×(Proportion of Bank Assets That are Municipals).
We can take these two components on the right hand side in turn.
The initial stumbling block to using a local bank to finance a municipality is that a new bank without a capital infusion starts out with $0 in assets. The municipality would presumably have to raise equity — which is a drain on the municipality’s finances. This means that the plan has to be long-term to allow the public bank to grow.
The asset weighting on a particular municipal bond is an important constraint for a bank. Municipal bonds are illiquid, and the issuer probably wants to have a much longer duration than a bank wants for assets. Banks are thinly capitalised, they cannot afford to take large losses adjusting their liquidity portfolio. Also, a public bank presumably has other mandates for its lending activities, which will consume the bulk of the bank’s asset mix. Finally, banks need to attract deposits — and the typical way to attract deposits is to make loans to the depositors and hope they stick around after they pay off the loan.
The liquidity issue is not trivial. If you are a major holder of a risky bond, any attempt to make major sales will move the market price — the other market participants know you need to raise liquidity, and they are not charitable institutions. Instead, they will exploit your weakness, and drop their bids on the bonds that you are trying to sell. This is generally not a feature of central government bonds due to the depth of the markets and the central bank backstop.
Suksawat & Ferguson Description Too Optimistic
If we look at the description in the article by Suksawat and Ferguson, they write:
Therefore, when a public bank purchases its own city’s municipal debt, the result is not a closed loop in which a finite amount of money is passed back and forth. Because the public bank actively generates money to purchase the debt, the operation dramatically enlarges the city’s fiscal space. In such an arrangement, the municipal government acquires funds in the short term to meet community needs. The public bank grows its holdings by receiving interest payments from the city. The loops, then, are not redundant; they are kinetic. Far from an inert circuit, a public bank that purchases city debt is a dynamic design that defies the artificial gravity of austerity.
The problem with their description is that they are describing the transactions at the point of purchase of the bonds, not the steady state. A municipality is not going to issue bonds so that it can build up its cash balances, they will need to spend that cash. Even if a municipality’s “subsidiaries” (library, fire department, etc.) bank with the public bank, their employees and suppliers will most likely not. The cash inflow would then turn into expected cash outflows. Which would then have to be met by selling the municipal bonds, which may be a problem.
For example, imagine that a city sells a convenient $100 in municipal bonds that are bought by the public bank. The public bank started with liquidity and capital ratios that were near its target. The public bank will end up with the following transactions (or some equivalent).
Sell $100 of liquid bonds (e.g., Treasurys) out of its liquidity portfolio to raise cash to pay for the municipals at “auction.” (The issuance of non-govvie bonds — the primary market2 — is typically via placements with dealers, which are less structured than the auctions for central government bonds. I will call this an “auction” for simplicity.)
Buy $100 in municipals. Although these might be placed in the “liquidity portfolio,” they are certainly less liquid than Treasurys. The city will deposit $100 into the public bank, allowing it to buy back $100 in Treasurys. It has expanded its balance sheet by $100, with +$100 assets in municipal bonds, and +$100 deposit liabilities.
Some time later, it will face up to $100 in outflows as the city (or “subsidiaries”) spend the raised cash. To avoid the transaction costs on the municipal bonds, it would likely sell Treasurys. The public bank’s balance sheet is back to its original size, but it has replaced up to $100 in Treasurys with municipals.
In order for the city to raise liquidity, it needs to issue municipal bonds to “outsiders” — which was exactly the situation without a public bank. The public bank bidding at the bond “auction” might help it go through at decent pricing, but that then pushes the problem of finding buyers in the secondary market — which is typically harder.
In summary, owning a public bank might help a municipality, but that public bank needs to be a profitable going concern before it would offer much help to the municipality’s finances. This underlying logic explains the structure of most financial subsidiaries of firms. They exist as a way of raising money to finance customer’s purchases, not to draw in deposits. About the only corporations that are interested in setting up deposit-gathering banking subsidiaries are ones with extremely large retail footprints (like some tech companies or even grocery stores). They have the equivalent of a branching network, so why not use it?
Why are American Municipal Bonds Illiquid?
The bond market for American state and local finance — collectively called “municipals” — is an illiquid mess. The reason for this is that it is separated from global fixed income markets by their tax treatment. For example, my job title at one point was analyst for a large Canadian provincial bond portfolio, but my team looked at sub-national bonds for other currencies as well — but American municipals were not looked at.
American municipal bond interest income are income tax free for residents. Which is a subsidy for the municipalities, and allows them to borrow at much lower rates. The reason is that the buyers are comparing the after-tax yield on other bonds versus the raw (tax-free) yield on the municipal. As a result, municipalities with decent credit ratings can typically borrow at rates below U.S. Treasury yields.
Although this subsidy is nice, it also means that municipals are effectively uninvestible for any entity that does not pay income tax in that local jurisdiction — which is most of real money fixed income investors. (There are some dedicated municipal bond funds.) This means that only a few local institutions and upper/middle class retail are buyers, and so the market is buy-and-hold investors. Which means that there is very little secondary market activity. Even if the municipal bonds get really cheap so that they have higher yields than other bonds does not help much — responsible bond managers have to do credit analysis before buying, and global investors do not have any American municipal bond experts on staff.
By contrast, Canadian provincial bonds have no special tax treatment, and provincial governments have a large economic footprint. As such, they have issuance departments that are as sophisticated as large corporations (for example, they issue foreign currency denominated bonds that are swapped back into Canadian dollars).
Pension Funds a Better Match
The problem with trying to place your bonds with banks is that they inherently have a short-term liquidity focus. You need buy-and-hold investors. On the institutional side, these are mainly life insurers and pension funds. An in Quebec, a model exists — a Caisse de dépôt et placement du Québec. It manages public pension plans, including Quebec Pension Plan — which was carved out of the national Canada Pension Plan (equivalent to Social Security in the United States). (It also manages some public liquidity portfolios and pension plans that would normally be private, such as for the construction industry after some financial scandals in the past.)
La Caisse is not just a garbage dump of provincial and municipal debt — it is more a fund manager that the depositors (weighted towards pension funds) have asset allocations to meet their investment needs. (E.g., depositors with liquidity portfolios are mainly money markets with some bonds, pension funds are mainly in risk assets.) Nevertheless, it is a large buyer of Quebec provincial bonds (and municipals3), and helps stabilise the market from the whims of the bond traders down the 401 Highway in Toronto. (This was a concern in the 1970s when separatism fears peaked.)
However, the tax treatment of municipals in the United States is likely a problem. (Please note that these comments are an educated guess on my part.) Pension funds are pass-through vehicles that are not directly taxed — the payments to plan members are taxed on their individual tax forms when funds are withdrawn (years after intermediate interest income was earned). As a result, municipal tax-free status does not appear to help pension funds holding them.
Risk
The problem with this type of strategy is that everyone involved is losing diversification risk. If the local economy goes into the toilet, the local public bank may face insolvency risk. At the same time, municipal finances would likely go all to heck. In such a situation, having the local public loaded up with local municipal debt would just be blood in the water for the financial sharks.
This is of course a problem for pension funds, which is why that allocations to local bonds is not going to be too large (unless the “local economy” is relatively large).
Concluding Remarks
The central government has a natural monopoly on currency issuance, in the sense that it is hard for competing currencies to gain market share. This applies to both private and sub-national government currencies.
I am using “how many bonds” (as opposed to “how much bonds”) on the theory that bonds are sold in $1,000 pieces, so that they are countable. This is despite the fact that analysis typically treats the amounts as a continuous variable.
The “primary market” is the initial issuance of bonds, which is where the issuer borrows money. The “secondary market” is the trading of existing bonds by bond market participants. The issuer is not involved, although the prices in the secondary market matter for its next primary market issuance.
Canadian municipal finance is boring because municipalities are creatures of provincial legislation. Nobody really trusts the leadership too much, so finance is very simple. They just assess everyone’s property value, and decide how much they want to spend. The property tax rate is set based on those two numbers so that the resulting budget is nearly balanced. American municipalities that tried that would probably face a revolt.
No comments:
Post a Comment
Note: Posts are manually moderated, with a varying delay. Some disappear.
The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.
Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.