Primer: Bank Treasury Operations

My primer is aimed more at the macroeconomic aspects of banking, and would be a lot shorter (and cheaper). However, my plan is to lean on the Anthology for more complex topics.

I was amused by his introduction to Chapter 10, which discusses the topic. Given the length of the quotation, I have cut it down.

There are certain functions in a bank that are understood universally and which require very little effort in their definition. For example [he gives “corporate lending” as an example]. And while some departments are relatively recent […], some departments are as old as banks themselves.

The “Treasury” function is just such an ancient department. Whatever they may have been called at the time, the very first banks would have had an individual or team of individuals responsible for collecting all the deposits and for transmitting all the loans,… However, mention “Treasury” to a banker (or University economics department professor) and it will not be entirely clear to them quite exactly what the Treasury department does in any one specific bank.

Choudhry explains that there is not a single model for a bank Treasury department, the format will depend upon the jurisdiction and the configuration of the bank itself. For example, an international megabank is going to have a much more intricate Treasury operations than a mom and pop savings bank with a single branch in a small town.

Choudhry suggests that there will be an asset-liability committee (ALCO) that is responsible for setting and implementing asset-liability matching (ALM) policies. The committee has a number of missions, the typical components of which he lists.

• Formulating ALM strategy
• Management reporting
• ALCO agenda and minutes
• Assessing liquidity, gap and interest-rate risk reports
• Scenario planning and analysis
• Interest income projection
• Managing bank liquidity book (CDs, Bills)
• Managing FRN book
• Investing bank capital
• Yield Curve analysis
• Money market trading
• Liquidity policy
• Managing funding and liquidity risk
• Ensuring funding diversification
• Managing lending of funds
• Formulating hedging policy
• Interest-rate risk exposure management
• Implementing hedging policy using cash and derivative instruments
• Formulating transfer pricing system and level
• Funding group entities
• Calculating the cost of capital

Why do I find this list interesting? This list of missions corresponds very closely to the tasks that one can run into in the funding group of an investment firm. That is, the Treasury group has a role similar to other fixed income investors. There are some firm-specific tasks, and the liability structure faced by a bank is different than other fixed income firms, but the principles are the same. Liquidity management is far more intricate for a bank, but it is an area that cannot be completely ignored by most other fixed income investors — even a long only bond fund faces uncertainty about inflows.

The role of the bank treasury is to take a global view of the bank’s operations, while other operating groups are attempting to generate profits within the framework determined by the treasury team.

The rest of this article will discuss some of the implications of this structure. However, liquidity management is deferred to another stand alone article due to its importance.

Lending Fables

This division of responsibilities runs directly into one of the fairy tales told about banking by economists: that banks worry about their liquidity position as part of the decision process for making a loan. Unless a bank is extremely small, the loan officers who approve a loan application are not members of the bank treasury, and unless they are senior loan officers, may never talk to anyone in the treasury group. Instead, the loan officers face internal guidelines for making loans, and the treasury group only worries about the liquidity effects after the loans are extended (unless the loan is so large that it has to be dealt with a loan officer who will be discussing the deal with treasury).

Meanwhile, for large banks, lending officers are in a hierarchy — there are a lot of junior loan officers who have low lending limits, and then lending limits rise as one moves up the hierarchy. Given the dispersion of loan officers, there is no way a bank can quietly decide to cut back on lending because of liquidity issues. The news would leak almost immediately, and the bank would be in a lot of trouble. The only time a bank can get away with a sudden stop is when everybody else is doing it — as happened in 2008.

In the real world, banks will have a large liquidity buffer to deal with the uncertainty about the pace of loan origination and deposit loss. Since it takes time to process loan applications, it is unlikely that there will be a sudden surge that needs to be dealt with. But if a bank has been aggressively expanding its loan book, its liquidity buffer will drop, and steps would have to be taken. This does not necessarily imply slowing lending, it could be dealt with via the issuance of longer-dated securities (debt and capital instruments).

The lesson is that we need to reject two common views about banking.

1. Although bank lending does create liabilities that pop into existence when they come into force (that is, “they come out of thin air”), this does not mean that a particular bank has an infinite capacity to create loans (or is somehow acting a fraudulent manner).

2. Attempting to micromanage lending via central bank operations does not work. Lending decisions are made by loan officers that are insulated from the details of bank treasury operations. One might argue that interest rate policy influences loan growth, but that has almost no connection to open market operations in practice (despite misleading models in economics textbooks).

Transfer Pricing

The Moorad Choudhry Anthology has a long discussion of “transfer pricing,” which I may or may not expand upon. However, it is the means by which the treasury group interacts with the other groups within a bank.

The treasury group will periodically produce a “price grid”: a menu of interest rate levels based on the instrument, typically expressed in terms of a spread. Groups within the bank “trade” with the Treasury at the specified price.

For example, a lending group will issue a floating rate loan to a customer at a spread above the transfer price spread for that category of loan. The treasury group ensures that the customer gets the funding. The profit for the lending group is determined by the return on the loan (after any credit losses) less the cost of funding as determined by transfer pricing.

The same principle holds for groups that generate funding: they supply funds to the Treasury at the transfer price, and their profit is the gap between the cost of funds externally raised versus the transfer price.

If these transfers are within a single legal jurisdiction, they all net out and so they only matter from an accounting purpose for things like group profits for determining bonus pools. If the transfers are between groups in different jurisdictions, they matter for income taxes, and thus are of interest to tax authorities.

I see two reasons to highlight this topic.

1. It helps explain how the different departments of banks operate — they have an internal cost of funds, and they are insulated from what is happening in funding markets that they are not directly involved in.

2. It is possible for the internal cost of funding to be set at inappropriate levels. The Moorad Choudhry Anthology discusses how badly set internal pricing led to banks ending up taking large positions with very poor liquidity characteristics going into the 2008 Financial Crisis. There is an incentive for groups within the bank to structure novel products in a way to arbitrage transfer pricing, and risk management needs to understand those products — which is hard to do when the financial engineers are running amok and churning out new products.

Hedging

Another area where economists spread fairy tales about banks is in the area of hedging: both interest rate and currency hedges.

If we go back in time, banks did not hedge interest rates (other than issuing long-term liabilities on a haphazard basis), and this notoriously was a driver behind the Savings & Loan crisis in the United States in the 1980s-early 1990s, which was echoed elsewhere. In the case of savings and loans, the celebrated 3-5-3 model (pay 3% on deposits, charge 5% on mortgages, hit the golf course at 3 PM) blew up courtesy of Fed Chairman Volcker. They needed to pay higher interest rate on deposits than what their long-term mortgage portfolio produced, and so they either went bust or gambled on high risk loans (making an even bigger bust).

However, the interest rate carnage in the 1994 bond bear market ended the era of ignoring interest rate risk. (New fangled exotic interest rate derivates were new and exciting, and they all blew up in the 1994 bond bear market. Exotic fixed income derivatives were largely side-lined in favour of vanilla derivatives thereafter. (Vanilla fixed income derivatives include caps, floors, fixed-floating swaps, swaptions, futures and futures options.)

Although different banks can operate in different manners, the usual assumption is that interest rate and currency risk is hedged via transfer pricing for all business units other than the bank treasury. The treasury department implements the external transactions, and ends up with the net exposure of all the other units — which they then manage via transacting on their own account.

(Why centralise interest rate/currency risk at treasury? Imagine what happens when you do not. Different groups could take the opposite sides of the same trade. Since they net to zero, this does nothing for the bank. But if the hedged instrument moves, one group makes a profit, the other a loss. Given the way bonuses work, the increase in bonus pay at the group that was correct will typically be larger than the drop in bonuses at the other group.)

The first thing to note is that the bank treasury would not “micromanage” positions — they look at the global exposure, and execute large adjustments when necessary. Within financial mathematics, instruments are priced on the basis that they instantaneously hedged as soon as risk parameters change — but that is just a convention to allow us to price instruments on an arbitrage-free basis, but not a behavioural requirement. Banks have a lot of ugly interest rate risk courtesy of embedded options in their portfolios (e.g., a loan that can be pre-paid early), and it would be extremely expensive to try to find an option portfolio to hedge out those risks perfectly (and very difficult to find counterparties).

When we realise that profits and losses are allocated among business units, a glance at the responsibilities of the Treasury shows us that some group that is implementing the treasury functions is going to be absorbing the gains and losses due to interest rate and currency changes. The treasury is acting like a fixed income manager, and thus also has to manage duration. If we look at recent decades, the consensus position has been to be bearish on bonds — and bank treasury groups are not going to be immune to that herding. With the understanding that embedded optionality may or may not be hedged out, we should not expect banks to be running the massive duration mismatches that economist/market commentator folklore suggests.

Finally, we run into currency mismatches. For multinational banks in the developed world, they similarly run their national banking businesses on a currency hedged basis. Groups may provide funding to each other, but this funding is done on a currency hedged basis. For example, the most dangerous aspect of the 2008 Financial Crisis was the tendency of foreign (mainly European) banks funding U.S. dollar-denominated asset-backed securities. As questions arose about the health of those groups, they were locked out of U.S. dollar money markets, and were forced to swap foreign (mainly euro) funding for dollars in the currency swap market — and the currency swap market was in the process of disintegrating until the central banks intervened.

One can verify this by counting up the number of developed country banks that have blown up due to currency valuation changes in recent decades (I am not an expert, but I am unaware of any). However, this is not true for developed economies: beliefs in the stability of currency pegs often leads banks to run currency mismatches (which was an issue during the 1997 Asian Crisis). This experience leads economists with experience in emerging markets to generalise this to developed banks. Furthermore, simplistic post-Keynesian multi-country models build in currency mismatches, leading some of the more dogmatic post-Keynesians to insist that this is how banks operate.

Since the mechanics of how a bank operates on a currency-hedged manner is perhaps not obvious, I will expand upon the possibly cryptic comments in the previous paragraph in a later article.

Concluding Remarks

Although bank treasury operations are a matter of the internal organisation of banks, understanding the structure helps explain the economic incentives of decision makers within banks. These incentives are obscured by primers that just look at aggregated bank balance sheets, and puts aside the question of hedging operations as being too complex.

Follow up articles will include bank liquidity management, as well as currency hedging.