T8-R44-P2-Hull-RMFI-Ch24-v3 - Study Notes
T8-R44-P2-Hull-RMFI-Ch24-v3 - Study Notes
T8-R44-P2-Hull-RMFI-Ch24-v3 - Study Notes
The information provided in this document is intended solely for you. Please do not freely distribute.
2
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Identify liquidity funding risk, funding sources, and lessons learned from real cases:
Northern Rock, Ashanti Goldfields, and Metallgesellschaft.
Evaluate Basel III liquidity risk ratios and BIS principles for sound liquidity risk
management.
Explain liquidity black holes and identify the causes of positive feedback trading.
Explain and calculate liquidity trading risk via cost of liquidation and
liquidity-adjusted VaR (LVaR).
Cost of liquidation
Financial managers are often interested in what it would cost
to liquidate a book of assets. Measuring the cost of
liquidation is often the first method professionals turn to
when answering this question.
Suppose the difference between the asking price (offer
price) and the bid price for an asset is its spread, noted by s,
as shown below1.
= ( )−
The mid-market price is simply this equation divided by the mid-market price, as shown in the
following equation1.
−
=
−
The proportional bid-offer spread1, s, has a cost associated with it if a financial institution opts to
liquidate a position. That cost is equal to
= ∙
where is the value of the position measured in dollars. The above equation is the cost to
liquidate a position. If the financial institution wants to know what it would cost to liquidate an
entire book, the cost of liquidation is simply the sum of all the costs associated with each
individual position in the firm’s book, as indicated by the following equation,
=∑ (Eq 24.1)1
where N is the number of positions in the book. The above equation is the cost of liquidation
formula related to liquidity trading risk.
1 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018).
Equation 24.1
3
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Two important notes are relevant from the equation for the cost of liquidation.
First, diversification may reduce market risk, it does not necessarily follow that it will
reduce liquidity trading risk2.
Second, the spread, s, is affected by many factors, including the size of the book. As the
size of the book increases, the bid-offer spread generally increases. Why? Because a
concentrated portfolio with large amounts of only a few stocks is usually harder to unload
than a few shares of many stocks. As shown below3, the spread as a function of quantity
held. The spread widens as the quantity of the asset an investor is trying to sell
increases.
An example of calculating the cost of liquidation4. Suppose that a financial institution holds
100 million shares of Company Gamma and 200 million ounces of a commodity. Further,
suppose that the bid price for the 100 million shares is $100 and the offer price is $100.50 and
the bid price for the commodity is $20 and an offer price of $20.10. The mid-market values for
the two positions are $10.025 billion and $4.010 billion, as shown in the following two equations.
$ . +$ .
− =( )∙ , ,
=$ , , ,
$ . +$ .
− =( )∙ , , =$ , , ,
2 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
3 Variation of Hull’s Figure 24.1. John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ:
John Wiley & Sons, 2018)
4 Variation of Hull’s Example 24.1. John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken,
4
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
From these, the proportional offer for the shares and commodities are5
100,000,000
ℎ = = 0.00998
10,025,000,000
200,000,000
= = 0.04988
4,010,000,000
With the proportional offers known, the cost of liquidation in a normal market is $105,011,875 for
a book value of $14,035,000,000, or 0.7% of the value of the book.
0.00998 0.04988
= 10,025,000,000 + 4,010,000,000
2 2
= $105,011,875.
( )∙
( )= ∑ (Equation 24.2)5
The value is the deviate (aka, quantile) associated with the desired confidence level; in
practice, we assume a normal distribution. If the confidence level is 99.0% (aka, significance
level is 1.0%), then assuming normally distributed spreads, the = 2.326.6 For statisticians, the
2.326 should be familiar. It is the z-score, found using Excel’s NORMSINV function.
5 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
6 Like VaR, this stressed spread always assumes a one-tailed deviate; i.e., at 99.0%, we use 2.36 and we do not
use 2.58, because only spread widening increases the cost.
7 The formula for conversion from annual volatility to daily volatility is = .
√
5
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
The VaR amount is now known. The next step is to calculate the LVaR. This is done at the
bottom of the following table assuming a 99.0% confidence level. Suppose the mean bid-ask
spread is $0.220 (0.458%). Further, assume that the spread volatility is 0.200%. Under these
assumptions, the worst expected spread is 0.92%, which is 0.458% plus the 0.200% multiplied
by 2.33 (remember, 99% confidence level). Next, to the actual LVaR calculation.
The liquidity cost (LC) with a static spread assuming equal half is 0.229% (0.5 times 0.458%).
The liquidity cost with a volatile spread is about twice as big at 0.462% (0.5 times (0.458% +
0.20% times 2.33)). In dollars, the liquidity cost using the static spread in this example is
$11,000, while the liquidity cost using the volatile spread is $22,166. This leads to the final
LVaR for the static spread of $3,199,264 and the final LVaR for the volatile spread at
$3,210,431.
6
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
7
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Next, suppose that mid-market changes are normally distributed with a standard deviation of
each day and that trading occurs at the beginning of each day. The total variance associated
with the trader’s position, assuming each day’s return is independent of the other, is, therefore,
the sum of ∙ , where is the initial position size8.
In this framework, the unwinding trader simply minimizes the VaR after trading costs. The formal
equation is to minimize the first equation subject to the second.
( )
∙ ∙ + ∙
2
Once one knows the coefficients and equations to plug into the above equations, then the
optimal trading strategy results.
8 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
8
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Funding sources
Main sources of liquidity include the following six9:
1. Cash and Treasury securities: Cash is always liquid and Treasury securities can
usually be converted to cash on short notice without problems.
a. Advantage: Easy source of liquidity on a short timeframe
b. Disadvantage: Low return (expensive). For financial institutions to be profitable,
they much allocate their assets to higher-yielding loans, such as to corporations.
2. Ability to liquidate trading book positions: Banks hold some of their assets in their
trading book. When liquidity needs arise, the bank can sell trading book assets.
a. Advantage: Generally, earn a higher rate of return under normal market
conditions.
b. Disadvantage: Has a higher risk than cash and Treasury securities, and the
potential that the bank will be selling assets at undesirable prices due to stressed
market conditions.
9 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
9
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
3. Ability to borrow money on short notice: Banks must be well aware of the assets they
could use as collateral when loans are needed on short notice. In order to mitigate
potential funding risks during a liquidity crisis, banks typically arrange for lines of credit.
a. Advantages: Avoids holding assets in low-returning options such as Treasury
securities.
b. Disadvantage: During times of stress, the bank risks having its collateral assets
devaluated by the market and potentially being unable to access funding beyond
its lines of credit.
4. Ability to attract retail and wholesale deposits on short notice with favorable offer
terms: Banks can offer higher interest rates to depositors in an attempt to bring in more
retail and wholesale deposits.
a. Advantage: Potentially attracts capital that is sensitive to interest rate
movements.
b. Disadvantage: Typically, liquidity problems tend to occur simultaneously across
the entire market, affecting many banks at the same time. This means that when
a firm plans to attract capital by offering higher interest rates than other banks,
other banks are likely thinking the same thing.
5. Securitizing assets on short notice: Prior to August 2007, a significant source of
liquidity for banks was the originate-to-distribute model for such things as loans. Rather
than hold loans or mortgages on their balance sheets, banks securitized the assets and
sold them to investors. In a liquidity crisis, banks could securitize assets with investor
demand.
a. Advantage: Offers banks a source of funding on short-term notice while being
able to hold the assets on its balance sheet.
b. Disadvantage: If the market becomes too risk-averse or lack confidence in
certain assets that the bank had planned to securitize, this potential source of
short-term funding liquidity disappears.
6. Ability to borrow from the central bank: This is typically not the most desirable option
because central banks usually charge higher interest rates on short-term loans than
other options. Central banks also typically require a haircut on the pledged collateral (i.e.
the central bank often will not lend at the full value of the pledged assets). Still, should
liquidity become an issue, banks can typically borrow from their central banks. The rules
for borrowing from a country’s central bank differ. For instance, in August 2007, the
haircut required by the European Central Bank (ECB) was lower than what a bank would
have to take when borrowing from the Bank of England or the Federal Reserve. Many
banks considered (and some actually did) setting up operations in Ireland to take
advantage of the lower haircut offered by the ECB.
a. Advantage: A safe source of liquidity financing.
b. Disadvantage: Borrowing from the central bank can be more expensive.
Additionally, there is the risk of reputational loss when a bank borrows from a
central bank. The reason being is that the market typically considers borrowing
from a central bank as a last resort option (lender of last resort), and thereby
must be experiencing financial problems.
10
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Lessons learned
Liquidity risk has led to the downfall of some of the world’s largest and respected financial
institutions, including, among others, Northern Rock, Ashanti Goldfields, and Metallgesellschaft.
Northern Rock10. One of the most famous examples of a financial
institution failing because of short-term liquidity problems was the
UK-based Northern Rock.
Up until its liquidity crisis in August 2007, Northern Rock had no
problems financing its large mortgage portfolio with wholesale
deposits and selling short-term debt instruments. Most of its funding
came from these two sources. But then came the global financial
crisis of which started in 2007. During this time, Northern Rock
couldn’t replace maturing financial instruments with new ones
because the crisis caused institutional investors to panic about the
potential credit risk associated with mortgage lenders.
Interestingly, Northern Rock was not short on assets that would
cover their liabilities. The Financial Services Authority (FSA) said as
such back in September 2007: “The FSA judges that Northern Rock
is solvent, exceeds its regulatory capital requirement, and has a
good quality loan book.10”
Even with this reassurance, when Robert Peston, who was the BBC business editor, revealed
that Northern Rock had borrowed from the Bank of England, Northern Rock experienced the
first bank run in England in 150 years10. From September 12th to the 17th, depositors withdrew
₤2 billion. By the end of the liquidity crisis, Northern Rock had borrowed an incredible ₤25 billion
from UK’s central bank. At that point (February 2008), the British government bought the bank
and removed management. Eventually, in November 2011, Northern Rock was purchased by
the Virgin Group for ₤747 million (led by Sir Richard Branson). This points to the importance of
managing the potential cascading effects of liquidity problems.
Ashanti Goldfields10. Another example of a company that
experienced liquidity problems was Ashanti Goldfields. The Ghana-
based company produced large amounts of the world’s gold supply
(it is still one of the top 10 gold producing mines in the world). As a
gold producer, Ashanti was concerned about the future price of gold.
To hedge its position, Ashanti worked with Goldman Sachs and
other investment banks to hedge its position in gold by selling gold
forward. Ashanti did this because they did not want their
shareholders to experience any adverse effects should the price of
gold decline.
The hedging backfired due to a lack of liquidity. In September 1999,
numerous European banks (fifteen in total) agreed that they would
reduce the sale of gold over the future span of five years. During this
time, gold prices increased more than 25%, which caused the
inability for Ashanti was unable to meet margin calls. This inability
led to a liquidity crisis at Ashanti. The gold producer was pushed to
the brink of bankruptcy, eventually selling a mine, diluting its equity
shareholders, going through major corporate restructuring, and
restructuring its “hedged” gold positions.
10 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
11
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Evaluate Basel III liquidity risk ratios and BIS principles for sound
liquidity risk management.
Basel III Liquidity Risk Ratio
Basel III introduced two new liquidity risk ratios:
Liquidity coverage ratio (LCR): The LCR is the result of the high-quality liquid assets
divided by net cash outflows in a 30-day period. The Basel III committee required this to
be equal to or greater than 100%. The LCR was slowly increased to 100% from 60% in
2015.
( )= (LCR11)
11 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
12
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Net stable funding ratio (NSFR): The NSFR shifts bank management’s attention to
liquidity management over one year. The ratio is simply the amount of stable funding
divided by the required amount of stable funding. The NSF should be above 100%.
( )= (NSFR12)
o Required amount of stable funding: The denominator stems from the items
requiring funding. Each category is multiplied by the required stable funding
(RSF) factor. The RSF factors are shown in the following table.
RSF Factors by Category13
Category RSF Factor
Cash 0%
Short-term instruments, securities, loans to financial entities if
they have a residual maturity of less than 1 year
Marketable securities with a residual maturity greater than 1 5%
year if they are claims on sovereign governments or similar
bodies with a 0% risk weight
Corporate bonds with a rating of AA- or higher and a residual 20%
maturity greater than 1 year
Claims on sovereign governments or similar bodies with a risk
weight of 20%
Gold, equity securities, bonds rated A+ to A- 50%
Residential mortgages 65%
Loans to retail and small business customers with a 85%
remaining maturity less than 1 year
All other assets 100%
12John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
13Adapted from John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley &
Sons, 2018)
13
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
An example. As an example, consider the situation where a bank has the following
balance sheet. It has $1 billion in cash, $3 billion in Treasury bonds, $1 billion in
mortgages, $1 billion in small business loans, and $3 billion in fixed assets. On the
liability side, the bank has $40 billion in retail deposits, $48 billion in wholesale deposits,
$4 billion in Tier 2 capital, and $8 billion in Tier 1 capital.
o The amount of stable funding is $72.
40 ∙ 0.9 + 48 ∙ 0.5 + 4 ∙ 1 + 8 ∙ 1 = $72
o The amount required amount of stable funding is $74.25.
5 ∙ 0 + 5 ∙ 0.05 + 20 ∙ 0.65 + 60 ∙ 0.85 + 10 ∙ 1 = $74.25
o The NSFR is then 0.97, or 97%. In this example, the bank fails the NSFR
requirement of 100%.
72
= = = 0.97
74.25
14
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
14 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
15
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
15 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
16
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
At times, negative feedback traders control the trading activity, whereas, at other times, positive
feedback traders control the market’s movements.
Negative feedback traders control market movements when markets are liquid. When
prices drop to theoretically unreasonable levels, negative feedback traders will buy the
stock. If the price moves above a reasonable valuation, then negative feedback traders
sell. The net result of the activity of negative feedback traders is a generally reasonable
balance between buyers and sellers.
Positive feedback traders control market activity when trading becomes weighted and
relatively illiquid. When asset prices fall, positive feedback traders sell, causing prices to
fall even more. The same co-movement happens in the opposite direction. Essentially,
when positive feedback traders dominate market activity, market prices are generally
more volatile and illiquid.
16 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
17
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
17 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
18
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
5. Margins. When markets move “big”, it can often lead to massive unrealized gains or
losses for investors that are heavily margined. When markets experience a big move
that is opposite of what a given investor was expecting, it can lead to a margin call.
When faced with margin calls, investors are forced to sell leveraged positions at large
losses, which can reinforce market volatility. The situation is further exacerbated by
exchanges responding to increasing volatility by increasing market requirements. The
positive feedback loop continues until it is broken.
6. Predatory trading. Trading is partly a zero-sum game. If a trader knows of an entity that
will have to sell a substantial portion of stock, then they will know that there will be a
drop in the price of that stock. To profit off this anticipated price drop, an investor may
take a short position, which reinforces the anticipated price decline. This type of
predatory trading often means that institutions may need to unwind large positions slowly
over time or use other discrete trading entities to accomplish unloading the position.
LTCM. One of the famous examples of positive
feedback trading destroying a fund is Long Term
Capital Management (LTCM). Perhaps the reason the
LTCM failure is so famous is that it was founded by
Nobel Prize-winning economists and other well-known
financiers. They built elegant mathematical models to
theoretically show the advantage of their trading
strategy. The strategy worked for a while. A popular
trading strategy was “relative value fixed income”.
This trading strategy involved shorting a liquid bond
and going long a similar bond that was illiquid. The
fund would then wait until the prices moved closer
together. What went wrong? In 1998, Russia
experienced financial turmoil and defaulted on its
bonds. This default had the effect of lowering the
value of illiquid instruments comparative to liquid
instruments that were analogous and caused massive
unrealized losses at LTCM. The problem was that
LTCM was severely leveraged and they were not able
to meet their margin calls. As a result, LTCM was forced to liquidate its positions. This
involved purchasing the securities they were short (the short-term liquid bonds) and
selling the long-term illiquid bonds, essentially buying high and selling low. This
leverage-driven requirement led to a further flight to quality18 and further divergence in
price, with the liquid bonds increasing in price and the illiquid bonds decreasing in value.
As is often the case, black swan events can lead to strong positive feedback loops.
7. Leveraging and Deleveraging. Positive feedback also stems from the cycles of
leveraging and deleveraging, depicted below. The first figure is the leveraging process.
When times are good, banks are often awash in liquidity, and credit spreads tighten.
Investors are allowed to increase leverage. In response to the increase in leverage,
investors buy more assets. In response, asset prices increase. When the increase in
asset prices is “sufficient enough”, the leverage of investors decreases. The positive
feedback loop then starts again until the “bubble” pops.
18 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
19
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
The deleveraging process is similar. When economic times are tough, credit spreads widen and
banks are less willing to lend, especially on margin. As a result, investors must decrease
leverage by selling assets. As a result, asset prices decline. When the decrease in asset prices
is “sufficient enough”, the leverage of investors increases. The process then starts again until
the asset markets reach an overall bottom.
19 Both diagrams on this page were adapted from John C. Hull, Risk Management and Financial Institutions, 5th
20
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Irrational Exuberance
The phrase “irrational exuberance” was first coined by former Federal Reserve chairman and
world-renowned economist Alan Greenspan in a December 1996 speech. He was originally
referring to equity markets, when he asked, “How do we know when irrational exuberance has
unduly escalated asset values?”20 The comment sparked a hotly contested debate in the
economics world between behavioral economists and efficient market hypothesis theorists.
On the one side of the debate are economists, such as Alan
Greenspan, who argue that markets are not as “efficient” as
some economists portray them to be. They argue that
investors have many psychological biases, such as
anchoring bias, risk aversion, herd following behavior
(bandwagon effect), attribute substitution bias, backfire
effect, framing effect, among many others. A well-known
quote from the ex-CEO of one of the largest banks in the
world sums up the business response to the behavioral
economics point of view21:
“When the music stops, in terms of liquidity, things will be
complicated. But as long as the music is playing, you’ve got
to get up and dance. We’re still dancing.”
On the other side of the debate are efficient market
hypothesists, who argue that markets simply incorporate all
available information at any given time. When markets
experience large swings in volatility, it is because new
information became available, not because traders
somehow behaved as crazed lunatics one minute and
switched the next.
The debate often comes down to one’s view on what causes bubbles. Is it irrational behavior on
behalf of market participants or does it stem from information changes? Likely the debate will
never be resolved because the two sides are really just framing the conservation through
alternative lenses. In any event, some of the most well-known “bubbles” that behavioral
economists mention are the 1987 stock market collapse, the 1994 bond market crash, the 1997-
1998 Asian financial crisis, the 1998 Long-Term Capital Management failure, and the 2007-
2008 subprime mortgage crisis20.
20 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
21 Chuck Prince, ex-CEO of Citigroup (July 2007)
21
Guru. Downloaded May 11, 2020.
The information provided in this document is intended solely for you. Please do not freely distribute.
Regulation. As is often the case, regulation plays an important role in financial market liquidity.
In the past two decades, financial regulators have worked to make regulations consistent across
countries. This move towards consistency comes with drawbacks.
First, uniform regulations tend to force banks to react in a similar way to external
events22. This homogeneous behavior may lead to similar risk profiles, positions, and
trades, which may result in a liquidity black hole – greater volatility and systemic risk
rather than less.
Second, the push towards uniform regulation has pushed credit into a black hole-like
situation. When the economy is at its bottom, default probabilities are high and loan
capital requirements tend to be high22. Banks are less likely to provide loans to small and
mid-sized businesses, the opposite of what a regulator would want. Similarly, when the
economy is booming, default probabilities tend to be low and banks are more interested
in providing loans. This procyclical nature of lending stemming partially from regulation
was noticed by the Basel Committee and has been somewhat addressed. Rather than
using current period default probabilities, banks use the average default probability
across the business cycle to estimate required reserves.
22 John C. Hull, Risk Management and Financial Institutions, 5th Edition (Hoboken, NJ: John Wiley & Sons, 2018)
22