In an earlier article, I did an over-simplified discussion of how a local public bank would interact with municipal bond issuance. There was some ambiguities that I skipped over in order to keep things brief. This drew a variety of questions, and so it is clear that I need to expand on what I wrote.
Since these articles are expected to be bound into a banking primer manuscript, I was going to need to cover some of these basics long before I got to a section on public banks. But I will cover the basics herein, and not worry about the manuscript logic. There are also some assertions about technical issues which I would have to reconsider and dig into if they stay in the manuscript.
Bond versus Loan Basics
Bonds can be thought of as “negotiable loans,” but they behave differently than loans. They are both types of credit instruments, and therefore both have the property of growing financial assets and liabilities “out of thin air” when created. However, bank loan creation is mirrored by deposit creation and bank deposits are considered “money,” so that “bank loan creation creates money,” while this is not true for bonds. Furthermore, traditional banks obviously have a privileged position in bank loan creation, but have no special privileges with respect to bond creation. Bonds have underwriters, which historically were investment banks. (The central bank is typically the “underwriter” of central government bonds, although the procedures are different.) With the collapse of the “pillar system” (and equivalents), bank holding companies may own traditional banking and investment banking subsidiaries. However, for my discussions, when I refer to “banks” I am referring to the traditional banking subsidiaries, and not the holding companies (since we might as well refer to “the financial sector” in that case, since the holding companies may have their fingers in every type of financial intermediary role).
Being a publicly traded security is not automatic. There are legal steps to follow, and the securities have to be priced on an arm’s length basis so that they acceptable investments for retail investors, who are assumed in securities laws to not be sophisticated. For our purposes herein, there is a presumption that bonds are widely marketed, so a single buyer is not buying up the bulk of the issue. (We can ignore such concerns when discussing central government bonds, as they are a special case within securities laws.)
“Private debt” are fixed income instruments where the investors are all presumed to be large, sophisticated institutional investors, and the deal is structured with less safeguards. They do not qualify as public securities under securities law. At the time of writing, there is hand-wringing in the financial press about private debt. Since private debt are expected to be illiquid, I am biased to not worry about them until defaults are realised.
Underwriting, No Public Bank Purchase
We will now imagine a city issuing $900 of municipal bonds via a single underwriter, which is an investment bank that handles municipal bonds. (Yes, $900 is comically small for a bond issue, but I want nice easy numbers to visualise.)
What happens is that the lawyers at the underwriter set up the contracts for $900 in bonds, which represent $900 in debt for the municipality. It will then attempt to sell the $900 in bonds to the underwriter’s clients, although it will keep a certain amount of bonds for its own account on its trading books, as other clients may wish to buy some bonds later. If there is difficulty in finding clients willing to commit to buy, the pricing (e.g., the coupon rate on the bond) may need to be adjusted so as to bring in more sales.
Once the pricing is set and committed investors pay, the investment bank will wire the municipality the proceeds of issuing all the bonds ($900) less the fees that are charged to pay all the hard-working, salt-of-the-earth investment bankers and sales people (plus the investment bank’s profit). That is, there is a wedge between the proceeds received by the municipality and the par value of the debt, which is presumably amortised as an additional interest expense over the life of the bond.
If we ignore the fees paid for the bond issuance, the municipality grows its balance sheet by $900. It has $900 in bond debt added to the liability side of its balance sheet, and receives $900 in cash inflows from the underwriter.
If we assume that the city is using a public bank as its bank, the public bank gets $900 wired via the payments system as a new asset, but has an increase of $900 in deposit liabilities (held by the city). It cannot safely buy illiquid assets with that $900, since the city presumably did not borrow just to plop money into its bank account, it will spend down its cash balance.
Underwriting With Public Bank Purchase
We then imagine that the public bank buys $100 of the city’s bond issuance. Since there is a new, big buyer, the underwriter could increase the size of the bond issuance by $100 to $1000.
However, it appears unlikely that the public bank can just say to the underwriter “give me $100 of the bond issue, I will settle it up with the city directly” instead, it would need to pay the underwriter $100 to be allocated a piece. If there is any lag between the payment to the underwriter before the city is paid, the public bank has an immediate liquidity outflow that is then reversed by the $1000 inflow (less fees) to the city. (My original article assumed that there was such a lag.)
Once all the transactions associated with the bond issuance settle, we have the following situation.
-
The city has $1000 in new deposits, which it will presumably spend.
-
The public bank has $900 in net settlement balances, plus a $100 in relatively illiquid municipal bond. Since the city is unlikely to spend the entire amount immediately, the settlement balances would likely be re-allocated to other higher-yielding assets.
Although the bond issuance improved the liquidity position of the city (and the public bank), that is due to the fact that it drew in $900 of cash from third parties. In order to get those inflows, the bond issue has to be seen as legitimate. Nobody sensible would want to buy into a bond issue where a subsidiary of the issuer (which is what the public bank is) is effectively drawing down 50% of the issue (for example). If the subsidiary needs to sell, the price of the bond in secondary market trading will tank — destroying the market value of the holdings of the other bond holders.
It should be noted that the public bank is now in a worse liquidity position than the previous case. In both cases it gets a deposit inflow. In the first case this is entirely matched by a settlement balance inflow, in the second, some of those settlement balances are replaced by illiquid municipal bonds. Since the city is probably not going to spend 100% of the proceeds immediately, this is manageable.
If the idea is that the public bank is going to offer lending support to the city, it is probably going to have to provide loans. However, those loans would run into concentration limits, never mind concerns about non-arm’s length dealings. Central governments do not worry about such rules when they deal with their central bank because they are the ones who set and enforce the rules.
Concluding Remarks
Even though bank loans and bonds are debt instruments, private banks have no special status for buying bonds. They need to wire funds to the entity selling the bond, like everyone else (although non-banks have their bank wire the funds on their behalf).
