Global Banking
Part 1:
1. Financial Intermediaries: Rationale and Risk
With complete markets, complete and symmetric information and so on, there is no need of FI to
match investors with savers. However, the real world is far from perfect and frictions are present.
These frictions are also known as agency costs: information is not complete. Therefore savers would
prefer not to lend and there are higher information costs.
There are 3 potential problems in a world without FI’s:
- Adverse Selection: Prior to purchasing a firm’s debt or equity, each individual must incur
costs to investigate its quality. If not, only the poorest (adverse) quality firms have the
greatest incentive to issue securities to unwary investors.
- Moral Hazard: Once managers have control of other people’s money, they may have the
incentive to spend it on excessively risky projects or perquisite consumption.
- Maturity, Liquidity and Denomination Mismatches: The firm’s debt or equity may have
characteristics, such as maturities, liquidity and denominations, that may not be attractive to
individual savers
FI’s can have major functions. The two most important are ‘broker’ and ‘asset transformer’.
- Broker: FI provides information about the quality of security issues. When only the broker
needs to incur costs to screen the quality of the firm’s securities, it reduces costs through
economies of scale.
- Asset Transformer: It transforms primary securities (e.g., loans, bonds, stocks) issued by
firms into secondary securities (e.g., bank deposit).
A: The average monitoring cost is lower for FIs since they exploit economies of scale
B: The FI can provide maturity intermediation: the maturities of its assets differ from the
maturities of its liabilities (Liabilities short, assets long)
Some other functions are: Transmission of monetary policy, credit allocation, payment services.
Because FI’s provide special services, they also receive special regulatory attention. The effects when
something goes wrong can be bad for many people and the economy as a whole. Because of the
potential effects, some institutions are too big to fail. Their failure could lead to the failure of many
other institutions.
- Bear Stearns (March 2008): NY Fed arranged a purchase by JP Morgan Chase (29 billion
billion in losses guaranteed by the Fed).
- AIG (September 2008): Government bought 79% stake (was heavily involved in credit default
swap market).
- Citigroup: Received 25 billion in October 2008 and then another 20 billion by TARP
,Regulation has to ensure the soundness of the system as a whole. It also has to prevent against unfair
practices like discrimination on geographical area’s or race.
Regulation imposes some private costs to FI’s: No more than 10% of equity to single borrower and
capital requirements for example. However, this also has benefits: Deposit insurance, access to TARP
funds to boost capital, too-big-to-fail protection.
FI’s face certain risks, otherwise they would not have existed. There are different types of risk.
- Interest Rate Risk: The mismatch in maturities (actually, durations) between assets and
liabilities.
Refinancing Risk: when assets have longer maturities than liabilities.
Reinvestment risk: when liabilities have longer maturities than assets.
The risk that the return on assets to be reinvested fall below the cost of liabilities.
Balance sheet hedge via matching maturities of assets and liabilities is problematic for FIs.
Inconsistent with asset transformation function.
- Market Risk: The risk of losses from the actively trading assets and liabilities (and
derivatives). This risk is related to changes in interest rates, exchange rates, and other asset
prices. However, this risk represents the incremental risk when other risks are combined with
an active (short term, even one day!) trading strategy.
Financial crisis illustrates that market (or trading) risk is present whenever a FI takes an open
(unhedged) long or short position in securities and prices change in a direction opposite to
what is expected.
- Credit Risk (Most important): Risk that promised cash flows on financial claims hold by FIs
(such as loans) are not paid in full.
• Firm-specific credit risk: associated with the specific types of project risk taken by that
firm.
• Systematic credit risk: associated with the macro-conditions affecting all borrowers (e.g.,
recession).
This can be managed by: screening before underwriting a loan, monitoring credit after
underwriting a loan, diversification, pricing higher credit risks with higher interest rates.
- Off-Balance-Sheets Risk: Off-balance sheet activities do not appear in the balance sheet
instead involve the creation of contingent A and L => affect the future balance sheets
(standby letters of credit, loan commitments, derivative positions).
- Liquidity Risk: Risk of being forced to borrow or sell assets in a very short period of time in
order to meet sudden increase in liability withdrawals. Low (fire sales) prices may result.
Runs may turn liquidity problem into solvency problem.
Contagion effect: once one bank goes down, people lose confidence in other banks. This
occurred somehow after Lehman Brother failed.
- Insolvency Risk: Risk of insufficient capital to offset sudden decline in value of assets relative
to liabilities.
Several banks the recent financial crisis became insolvent but were rescued by the TARP
funds and the Capital Purchase Program. Other insolvent banks were taken over by healthier
institutions. Citigroup became insolvent, but was considered too big to fail and rescued by
the government.
- Other Risks: These risks are not particularly proper of FIs, but they are classic topic in risk
management of any type of firm.
• Foreign Exchange Risks: To hedge foreign exchange risk, work with multiple currencies or
offset them with derivatives.
• Sovereign Risk: Exposure to foreign governments which may impose restrictions on
repayments to foreigners. Includes the risk that the government itself may not repay
(sovereign default). Very difficult to hedge.
, • Technology and Operational Risks: technology investment may fail to produce
anticipated cost savings or anticipated revenue enhancements. Operational risk that
existing technology or support systems may malfunction or break down.
• Macroeconomic Risk: Examples: increased (volatility of) inflation, which affect interest
rates; increased unemployment, which affects credit risk.
2: Financial Intermediaries as Delegated Monitors: Theory and Empirical Evidence
Why do investors lend to banks who then lend to borrowers, instead of lending directly? To
rationalize FIs, it is important to describe the features of financial markets where they operate and
highlight what allows FIs to provide beneficial services. We must understand the type of financial
contracts written by FI’s and why these are optimal.
Douglas W. Diamond, 1996, “Financial Intermediation as Delegated Monitoring: A Simple Example”,
Federal Reserve Bank of Richmond Economic Quarterly, 83(2): 51-66.
This paper only focusses on asset services. FI’s only invest in assets in which it has a comparative
advantage. The model shows that reduced monitoring costs are a source of this competitive
advantage.
The cost and benefit of monitoring:
K = physical cost S = saving from monitoring m = number of lenders per borrower
Total Monitoring Cost: m*K > K
To save monitoring cost you can delegate monitoring to only one lender.
This however creates a new problem: Delegation Cost per borrower (D). Not sure if the delegator
(bank) really monitors.
Intermediation has to lead to an improvement. Therefore delegated monitoring can be optimal if and
only if K + D < min [S, m*K]
Example: Assume a borrower needs $1 million (1 unit). The borrower has a positive NPV project, with
uncertain realization V. The distribution of the V is known to everybody.
H = 1.4 with probability p = 0.8
L = 1 with 1-p = 0.2
But only the borrower observes the realization of the project.
There are m=10,000 investors, and each has 1/m of the amount needed by the borrower. Therefore
each investor has $100 (or 0.0001 unit).
K = $200 m*K = $2 million
Monitoring is expensive, so let’s see what is the best contract without monitoring.
Can equity be optimal? No… The profit-sharing depends on the value Z of the project reported by the
borrower. Let a*Z the fraction of reported value that goes to investors.
Borrower payoff is therefore V – a*Z. This implies Z=1, i.e. the borrower always report the smallest
value of Z, because this maximizes his V – a*Z. He always cheats on you. Nothing induces higher
payment in case of equity contract.
The investor could induce higher payments from the borrower by imposing a penalty for low
payments liquidation of the borrower’s assets. In case of liquidation/bankruptcy there are no
proceed both for the borrower and the investors.
, Liquidation works as follow: if the lender liquidates for a given payment f, he also liquidate for all
lower payments. Otherwise, the borrower pays the lowest amount that avoids liquidation and keep
the rest.
Therefore the optimal contract without monitoring is a debt contract with face value f. (A payment
equal to f avoids liquidation, a lower payment instead induces liquidation).
Assume investors require a 5% return on their lending. Can f=1 be possible? The answer is no.
When V=1 the borrower pays 1 and get 0.
When V=1.4 the borrower pays 1 (avoid liquidation) and get 0.4.
This means investors always get 1 < 1.05! Not acceptable.
To compute the minimum repayment f observe that any value 1 < f < 1.4 forces the borrower to
liquidate when V=1 and to repay in full when V=1.4.
Therefore investors have an expected return of 0.8f, which has to be equal their required return.
That is:
0.8*f = 1.05 f = 1.3125
The borrower gets 1.4 - 1.3125 = 0.0825 when V=1.4.
So, when investors do not observe borrower’s cash flows and they cannot monitor then:
- Equity contracts do not work.
- Debt contracts are best, they induce borrower to repay because default/liquidation is a
penalty to avoid.
- In our example, the cost of unmonitored debt is 31.25%
Monitoring means the investors observe the true V. Then when V=1 is now possible to avoid
liquidation (the borrower declares the low outcome and investors monitor).
The value of monitoring is the expected savings in bankruptcy/liquidation costs: S = 0.2(1) = 0.2 (or
$200,000).
The expected savings in bankruptcy/liquidation costs have to be compared with the expected
monitoring costs, which depend on how many monitor:
0.2*K must be smaller than S 0.2*200 =$40 < $200.000 TRUE
0.2*m*K must be smaller than S 0.2*2m = $400.000 > 200.000 NOT TRUE
Only when one investor monitors it is convenient to monitor, otherwise not!
But recall that delegating implies the delegating cost D (bank is delegated to monitor. How to
delegate without monitor the monitor (e.g. the bank)?
Answer: the bank has to face liquidation cost as a function of the amount paid to the 10,000
investors (depositors). If the bank does not repay its obligations with depositors it incurs liquidation
that yields nothing to the bank and depositors. The bank is better off paying a minimum amount to
depositors to avoid liquidation.
Suppose the bank has only one borrower/loan to monitor. When V=1, the bank monitors and
collects 1, avoiding liquidation for the borrower. However, the bank itself is liquidated since the
depositors expect 1.05 (5% return on their investment) < 1. The bank is liquidated whenever the
borrower is liquidated (would he used unmonitored debt)! The one-loan bank will default as often as
the borrower.
Suppose the bank monitors two loans (each from a different borrower). Each loan is worth $1 million
(or 1). To finance the loans the bank has $2 million (or 2) deposits from 20,000 investors. The returns
of each loan are independently distributed, otherwise are just like the single loan (i.e., V=1 w.p. 0.2
and V=1.4 w.p. 0.8).
