Product Management

Product Management 2011
Page 1 of 133
Contents
1
Financial Markets
Product Management: Influence Factors
2.1
Parties Involved . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2
Asset Classes, Securities and Risks . . . . . . . . . . . . . . . . . .
19
2.3
Handling Ratios and Characteristics . . . . . . . . . . . . . . . . .
26
Zero Coupon Bonds
33
Coupon Bonds (Straight Bonds)
35
Floating Rate Notes (FRNs, Floaters)
36
Swaps
37
6.1
Plain Vanilla Swaps (Straight Swaps) . . . . . . . . . . . . . . . . .
38
6.2
Basis Swap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
39
6.3
Cross Currency Swaps . . . . . . . . . . . . . . . . . . . . . . . . .
39
Forward Rate Agreements (FRAs)
40
Bond Futures
40
8.1
Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
44
8.2
Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
45
8.3
Conversion Factors . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
8.4
CHF-denominated Bonds . . . . . . . . . . . . . . . . . . . . . . .
47
8.5
EUR-denominated Bonds . . . . . . . . . . . . . . . . . . . . . . .
47
8.6
Euro-BTP Future . . . . . . . . . . . . . . . . . . . . . . . . . . . .
48
Eurodollar Future
Erich Janka
49
FH bfi
Page 2 of 133
10 Leverage and Margins
52
11 Arbitrage
55
12 Convertible Bonds
57
12.1 Replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
59
12.2 Valuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
59
13 Bootstrapping
60
13.1 Discount Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
60
13.2 Spot Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
61
14 Forward Rates
61
15 Interpolation of Yield Curves
61
15.1 Accrued Interests . . . . . . . . . . . . . . . . . . . . . . . . . . . .

16 Interest Rates and Returns
63
66
16.1 Discrete and Continuous Interests . . . . . . . . . . . . . . . . . .
66
16.2 Arbitrary Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . .
68
16.3 Interests and Return . . . . . . . . . . . . . . . . . . . . . . . . . .
69
16.4 Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
70
16.5 Multi Period Returns . . . . . . . . . . . . . . . . . . . . . . . . . .
71
17 Options
85
17.1 Put Call Parity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
85
17.2 Black Merton Scholes . . . . . . . . . . . . . . . . . . . . . . . . . .
85
17.3 Influence Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . .
85
17.4 American Options . . . . . . . . . . . . . . . . . . . . . . . . . . . .
86
17.5 Asian Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
87
17.6 Digital Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
88
17.7 Quanto Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
90
Erich Janka
FH bfi
Page 3 of 133
18 Payoff Diagrams
91
19 Payoff Formula
91
20 Structured Products
94
20.1 Capital Protection Certificate with Participation . . . . . . . . . .
94
20.2 Convertible Certificate . . . . . . . . . . . . . . . . . . . . . . . . .
96
20.3 Barrier Capital Protection Certificate . . . . . . . . . . . . . . . . .
97
20.4 Capital Protection Certificate with Coupon . . . . . . . . . . . . .
98
20.5 Discount Certificate . . . . . . . . . . . . . . . . . . . . . . . . . . .
99
20.6 Barrier Discount Certificate . . . . . . . . . . . . . . . . . . . . . . 100

20.7 Reverse Convertible . . . . . . . . . . . . . . . . . . . . . . . . . . 101
20.8 Barrier Reverse Convertible . . . . . . . . . . . . . . . . . . . . . . 102
20.9 Express Certificate . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
20.10Tracker Certificate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
20.11Out-performance certificate . . . . . . . . . . . . . . . . . . . . . . 106
20.12Bonus certificate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
20.13Bonus outperformance certificate . . . . . . . . . . . . . . . . . . . 108
20.14TwinWin certificate . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
20.15Reference Entity Certificate with Conditional Capital Protection . 110
20.16Reference Entity Certificate with Yield Enhancement . . . . . . . . 111
20.17Reference Entity Certificate with Participation . . . . . . . . . . . 112
21 Structured Products and Barrier Options
113
22 Option Strategies
114
22.1 Long Risk Reversal . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

22.2 Short Risk Reversal . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
22.3 Bull Spread . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Erich Janka
FH bfi
Page 4 of 133
22.4 Bear Spread . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

22.5 Long Butterfly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
22.6 Short Butterfly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
22.7 Long Condor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
22.8 Short Condor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
22.9 Long Straddle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
22.10Short Straddle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
22.11Long Strangle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
22.12Short Strangle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
22.13Ratio Call Spread . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
22.14Ratio Put Spread . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
22.15Call Ratio Backspread . . . . . . . . . . . . . . . . . . . . . . . . . 120
22.16Put Ratio Backspread . . . . . . . . . . . . . . . . . . . . . . . . . . 121
22.17Calender Spread . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
23 Securities Lending
121
24 Repurchase Agreements (REPOs)
122
25 Basel II, Basel III
123
25.1 Basel II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

25.2 Basel III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
26 Monte-Carlo-Simulation
128
26.1 Generate Correlated Random Variates . . . . . . . . . . . . . . . . 128

26.2 Product Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
26.3 Value at Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Erich Janka
FH bfi
Product Management
Erich Janka
2011
Financial Markets
Figure 1 uses data from [DB] showing the notional amount and the gross market
value for the derivatives outstanding as well as the market capitalization for
shares and bonds in the year 2007.
Figure 1: Size of markets: Derivatives, Shares and Bonds.
Figure 2 on the next page shows a customer breakdown of European dealers

revenues by underlying asset class in the year 2006 ([DB] cites McKinsey as
source of the data).
Figure 3 on the following page shows a breakdown of the global derivative
market 2007 by OTC versus on-exchange and by underlying asset class according
to [DB].
Page 6 of 133
Figure 2: Dealers revenue for OTC derivatives.
Figure 3: Global derivatives market.
Erich Janka
FH bfi
Page 7 of 133
Product Management: Influence Factors

price
expected return
risk
organizational issues
laws
taxes
investment objectives
2.1
Parties Involved
Parties Involved
Investor
Asset Manager
Portfolio manager
Analyst
Risk control
Reporting
Trading desk
Broker
Trading system
Custodian
Sales
Auditor
Erich Janka
FH bfi
Page 8 of 133
Authorities
Counterpart
Parties
This graph shows the relations between the involved parties:
The main parties are the investor who wants diligence for his money and the
asset manager who should provide it. Some other involved parties are also
shown: The custodian who is responsible for the safekeeping of the securities,
the auditor (internal/external) and the authorities ensure the diligence of the
asset manager. Every single party usually makes a sure profitexcept for the
investor who pays for it. Some parties can coincide but investor, counterpart,
trading system and custodian are usually always present.
In the following we will focus on the investor, the asset manager and their
relationship.
Portfolio Management Process
Erich Janka
FH bfi
Page 9 of 133
There can be two kinds of legal restrictions: Some hold for the investor and others
hold for the investment type. These restrictions do differ between countries
even within the EU where all legislators have the same directives.
Investment Planning
When someone wants to invest money, he should begin with his objectives and
constraints and check if they are realistic:
Is it a singular investment or periodic?
When is money needed, and how much?
Are there unique circumstances?
Are there tax, legal or regulatory issues?
Are the objectives realistic?
Which asset manager do I trust?
These questions should result in the Investment policy statement.
Price and Utility
price: The exchange value of an asset.

utility: The individual value of an asset.
The price of an asset needs not to be the value it has for individual investors.
The expected return and its risk should match the needs of the investors.
Below you will find two examples of assets with negative expected value
bought by many people:
Erich Janka
FH bfi
Page 10 of 133
Lottery (very little price)

Insurance (possible loss too big to take)
Choosing an Asset Manager
Investing can be very close to gambling if it is not planned carefully. It is not
new that asset managers promise riches and investors believe them:
I can calculate the motions of the heavenly bodies,
but not the madness of people.
(Sir Isaac Newton, 1720)
. . . on losing a fortune due to the South Sea Bubble.
Investment Policy Statement
With the investment policy statement the investor must write down hismaybe
quite vagueobjectives in a structured form, thereby forcing himself to reconsider his intentions. The clearer he does this, the fewer misconceptions will occur.
The investor also determines how much freedom he will grant the asset manager
when he defines many or few details of the asset management such as asset
allocation, investment style, . . . A statement typically includes:
Investment objectives and constraints.
Duties and responsibilities of all parties involved.
Benchmarks and performance measures.
Time and extent of reporting.
Time schedule for review investment process.
If the statement does not only include the objectives, duties, measures and so on
but also the investors intentions and explanations why he set it up in a certain
way, the investor has a reference and the other parties can make more adequate
decisions.
Investment Specifications
Some examples of what the investment policy statement can contain are:
Strategic asset allocation: benchmark
Erich Janka
FH bfi
Page 11 of 133
Investment universe
Withdrawal plan
Band-widths for asset manager
Risk figures with band-widths
Performance figures
Types of Investors
Small investors
Pension funds
Insurance companies
Financial institutions
Companies, counties, foundations, large scale investors
Investment Objectives
Investment objectives can be manifold. Below, there are some examples; of
course there may be many others and combinations are possible as well:
Preservation of capital
Absolute return
Islamic banking
Ethical and sustainable investments
Benchmark
Time horizon
Return on equity target
Level of risk
Erich Janka
FH bfi
Page 12 of 133
Ethical and Sustainable Investments

The attempt to define ethical and sustainable investment can lead to lengthy
discussions and in fact it is not really a part of asset management. But a prudent
asset manager should not just do some cover up. There are several approaches
on how to implement ethics in portfolios:
absolute: Really ethical investments.
relative: Best in class.
average: Point system for investments.
Usually a committee or a NGO affirms that the investment is in line with its
ethical restrictions.
Selecting an Asset Manager
Selecting an asset manager is not a simple task for the investor since the asset
manager always tries to present himself most favorably. If he can, he will choose
the ratios and their time horizons. Therefore the investor should watch out and
decide what information he wants. Besides heand not the asset manager
should:
Choose performance and risk ratios.
Choose the time domain of interest.
Check the company set-upespecially quality and the integration of risk
management.
Check the costs.
Be careful if everything looks too good.
Belief in the skill of the asset manager.
If the asset manager cant deliver the requested information, he should be able
to explain the reasons. If a an asset manager always makes profit, you should
be suspicious. There might be some hidden risk.
Performance Fees
There are several reasons against performance fees:
Asset managers are bound to do the best for their client.
Erich Janka
FH bfi
Page 13 of 133
They wont do any better if he gets a performance fee.

The asset manager might focus more on the fee than on the interests of the
clients.
If there is a performance fee, it should at least
have a high watermark and
depend on a risk adjusted characteristic.
If the maximum fee is reached, the asset manager might take less risk. If a
performance fee is unlikely because of bad performance, the manager might
increase the risk to improve his chances.
Creating a Product
Product design group

Risk management
Legal / tax aspects
Authorities check product documentation (e. g. Key Information Document
(KID))
Education of sales
Marketing
Portfolio Turnover Ratio (PTR)
A characteristic the investor should be interested in is the portfolio turnover ratio
(PTR). The PTR relates the amount of security transactions to the portfolio value.
It is an indicator for the amount of transaction costs. There are several definitions
which lead to similar values for a typical portfolio. Differences arise when large
purchases or redemptions take place.
The required sums comprise the period of one year.
PTR: Definitions
PTR =
Erich Janka
min (
X,
M
Y)
FH bfi
Page 14 of 133
P
PTR =
(X + Y)
M
( P + R)
X . . . Values of security buys

Y . . . Values of security sellings
M . . . (Monthly) average of the NAV
P . . . Investments
R . . . Redemptions
Equation (2) is from the Austrian Prospektinhalt-Verordnung - Anlage 2 zu BGBl.

II Nr. 237/2005.
PTR Example 1
Typical portfolio:
Month
Buys
Sellings
NAV
Investments
Redemptions
1
2
3
4
153
30
44
37
52
34
237
39
9 100
9 109
8 911
8 909
100
0
0
0
0
0
200
0
Sum
264
362
9 007
100
200
PTR 1: 2.9 %.
PTR 2: 3.6 %.
These example trades, investments and redemptions give similar results for both
formulas. Since the table only covers four months, the problem how to deal
with ratios when they are calculated for one time domain but needed for another
arises. In this case it seems very easy: When we assume that the trading style
is consistent, we just multiply our result by 3 to get annual values. Of course
there are situations where this assumption and our method for rescaling will
be wrong.
There is no typical range for the PTR. It can be a few percent or
many thousand. The higher the value, the more the transaction costs, which will
reduce the performance.
PTR Example 2
No investments or redemptions:
Erich Janka
FH bfi
Page 15 of 133
Month
Buys
Sellings
NAV
Investments
Redemptions
1
2
3
4
25
25
25
25
25
25
25
25
100
100
100
100
0
0
0
0
0
0
0
0
Sum
100
100
100
PTR 1: 100 %.
PTR 2: 200 %.
This example shows a fundamental difference between the formulas: After a
complete round-trip, the first formula yields 100 % while the second yields 200 %!
The first result might be more plausible but its just a matter of scale: a simple
division by 2 solves this problem. This type of ambiguous characteristics can
lead to misunderstanding. To avoid this, check the used formula.
PTR Example 3
Fewer buys than investments:
PTR 1:
Month
Buys
Sellings
NAV
Investments
Redemptions
1
2
3
4
100
50
0
0
0
0
0
100
100
200
200
100
100
100
0
0
0
0
0
100
Sum
150
100
150
200
100
67 %.
PTR 2: 33 %.
This example shows anothermore complicateddifference between the formulas. Formula 1 ignores investments and redemptions; the PTR is 67 %. Formula
2 33 % takes these aspects into account: The PTR is negative because not all
incoming money is used to buy securities. This could be because the investment
and the buys happened in different periods of interest (around the turn of the
year)this means the ratio is distortedor some money is left in cash. Negative PTRs and the corresponding formulas are sometimes criticized, ignoring
that negative values do carry information.
Erich Janka
FH bfi
Page 16 of 133
Transaction Costs
The transaction costs include more than just some fees, they consist of:
Fees. The fees include brokerage, stamp duty, settlement, . . .
Buy/sell spread. The spread usually depends on the traded volume at a
specific stock exchange. It can happen that a specific exchange has higher
fees but a smaller spread.
Impact of trade size. If the order size is much above average the other
market participants will notice and the price will change unfavorably. This
is closely related with the liquidity of the security. Small trades will have
no impact, huge trades will have severe impact.
The first two items are easy to determine, the third can only be estimated. The
trade changes the price and it is not possible to know the price if that trade never
happens.
Transaction Costs: Liquid Share
Transaction Costs: Illiquid Share
Erich Janka
FH bfi
Page 17 of 133
Total Expense Ratio (TER)

The Total Expense Ratio is the total costs divided by the total assets. Usually
large funds have a smaller TER because unlike the management fee some of the
other fees are not percentages of the net asset value. The Total Expense Ratio is
an indicator for how much a fund loses due to administration:
TER =
Total costs
Total Assets
Components of the TER

Typically the costs include:
Management fee.
Legal fees.
Auditors costs.
Custodian fees.
But it does not include transaction costs.
Austrian mutual funds have to calculate the TER according to the ProspektinhaltVerordnung - Anlage 1 zu BGBl. II Nr. 237/2005.
Erich Janka
FH bfi
Page 18 of 133
Asset Allocation
The investment objectives are the foundation of the strategic asset allocation which
is defined in the investment policy statement. Usually it gives the asset manager
the freedom to modify the asset allocation within a given range: the tactical asset
allocation.
Strategic asset allocation.
Tactical asset allocation.
Trading Team
When the asset allocation is fixed, the individual securities can be bought. This
can be done by a single person or it can be split according to different tasks:
Different asset classes
Various regions
Currency overlay
Hedge strategy
Intern or Extern Management
Since the asset management can be split, it is can be internal or external.
Internal management (competence)
External management (best in class)
Mutual funds
Sub asset managers
An investor should be interested in which part of the management is internal
and which external.
Active and Passive Management
Asset management styles can be classified as follows:
Passive asset management. In this strategy the manager makes as few
transactions as possible to minimize transaction costs.
Erich Janka
FH bfi
Page 19 of 133
Active asset management. In this strategy the manager tries do bring in

his knowledge and information to optimize the performance. This can be
done by
Market Timing. The manager predicts future prices and changes the
market exposure of this portfolio.
Stock Selection. The manager predicts which stocks will do better
(worse) than the benchmark and increases (decreases) their weight.
There is a smooth transition between active and passive management since the
investor can define how much the manager can deviate from his benchmark.
2.2
Asset Classes, Securities and Risks
Asset/Risk Classes
The following asset classes can also be considered as risk classes:
Equity
Interest (Fixed Income)
Foreign Exchange (FX)
Credit
Commodity, Energy
Real Estate
Inflation
Volatility
Weather
Acquire Asset Classes
If we want a specific asset/risk class in our portfolio, there might be several
securities we can choose from:
Erich Janka
FH bfi

Cash
FX
Page 20 of 133
Share
Bond
Loan
Deriv.
Cert.
Fund
Equity
Commodity
Volatility
Weather
Real Estate
Interest
Credit
Inflation
x
x
To the right there are the securities with less clear composition; insurances are
not included in this list.
Acquire Asset Classes Unintentionally
Securities my include several risksthe one we desire and maybe some others.
The following table shows which risks can be in the portfolio unintentionally:
Cash
FX
Share
Bond
Loan
Deriv.
Cert.
Fund
Equity
Commodity
Real Estate
Interest
Credit
Inflation
Volatility
Weather
Properties of Securities
OTCexchange traded. Bonds are usually traded over the counter; shares
via stock exchanges.
Erich Janka
FH bfi
Page 21 of 133
Common stockpreferred stock. To compensate the (usually) missing

voting rights, preferred stock often pay higher dividends. They are senior
to common stock but subordinate to bonds.
Debt securitiesequity securities. Very crude: With debt securities you
lend money to the company while equity holders own (a little) part of the
company.
Derivative. The price of a derivative depends on some other asset, event,
value or conditionthe underlying (asset).
More Properties of Securities
Leverage. Some securities/portfolios have more exposure to the market

than they are worth. (For example the margin of a future is only a fraction
of a futures market exposure.) Leverage can be achieved via credit or
derivatives.
Physical delivery. When a commodity future ends one usually owns the
commodity. Since asset managers are not interested in the commodity itself
(only its price changes) they buy contracts without physical delivery or
close the contract before it ends.
Path dependent. The price of a security at maturity can depend on the
historic prices of some underlying, for example knock-in/out options.
Examples for Securities
Shares
Bonds
Convertible (bond)
Credit default swap (CDS)
Asset-backed security (ABS)
Mortgage-backed security (MBS)
Collateralized debt obligation (CDO)
Erich Janka
FH bfi
Page 22 of 133
CDO-Squared
Collateralized loan obligation (CLO)
Real estate investment trust (REIT)
More Examples for Securities
Certificates
Mutual funds
Exchange traded funds (ETF)
Closed-end funds
Hedge-funds
Futures, forwards
Options, warrants
Swaps
Contracts for difference (CFD)
Diversification
The risk level of a portfolio is influences by
the number of securities,
their weights,
their volatilities
and the pairwise correlations.
Diversification: Estimation of Magnitude
Assumption: All N securities have equal weight (1/N), identical risk () and
pairwise the same correlation ().
Portfolio risk:
s

1
1
P =
+ 1
N
N
Erich Janka
FH bfi
Page 23 of 133
lim p =
ij
ij = 2
2
1
2 2
N
2 2 . . . 2 1
N
1
1
1
..
..
.. ..
.
N N ... N
.
.
.
.
. .
2
2
1
. . . 2
N
1
N 1
+
N
N
r
1 + N 1

N1
1
N
N
1
2
+
=
..
N
N
N
.
1
N 1
N + N
v
u
u
u
u
= w0 w = u
u
t
v
u
u
u
u
= u
u
t
1
N
1
N
...
Thus low correlations are an efficient way to reduce portfolio risk.

Diversification: Figure
A single big position of 1/m of the whole portfolio also limits the efficiency of
diversification:
1
1 2 2
.. mm1
..
2

.
1
. N 1
1
m 1
P2 = m1 m
.
N 1 N 1
.
.
.
..
. . 2 ..
..
m 1
2 2 1
N 1

2
2
( m 1)
2N 2
= 2 1 + 2( m 1) +
+ ( m 1)
N1
N1
m
q
lim p =
1 + ( m2 1)
m
N
Erich Janka
FH bfi
Page 24 of 133
Diversification: Figures
2.2.1
A Hedging Example
[Glass][p. 21ff] The mathematics behind diversification and hedging is identical.

An air cargo company
needs jet fuel (two million gallons)
in three months
it cant take an increasing fuel price
Therefore it has to hedge this risk. Since no jet fuel contracts are available, the
company has to use futures on heating oil instead (they are highly correlated).
We want to find answers to the following questions:
1. What is the price risk (volatility) without hedge?
2. What is the price risk with gallon-for-gallon hedge?
3. What is the hedge with minimum risk and how effective is it?
Spot price heating oil: 65 cents
Spot price jet fuel: 68 cents
Future price heating oil: 67 cents
Future contract size: 42 000 gallons
Erich Janka
FH bfi
Page 25 of 133
3-month vola of jet fuel: 18 %

3-month vola of heating oil future: 23 %
correlation: 0.82
We want the risk valued in USD:
X . . . change in heating oil futures price per gallon (3 months)
Y . . . change of jet fuel price (per gallon, 3 months)
X = 0.23 0.67 = 0.1541

Y = 0.18 0.68 = 0.1224
Now we can answer question 1:
The change of the unhedged position is: 2 000 000 Y and the volatility is
2 000 000 Y = 244 800USD
2.2.2
Gallon-for-Gallon Hedge
The company needs

2 000 000
= 47.6 48 contracts
42 000
and will close them before expiry.
The risky position is now:
48 42 000 X 2 000 000Y
With X = 0.1541, Y = 0.1224 and = 0.82 the volatility is
q
(48 42 000)2 X2 + 2 000 0002 Y2 2 48 42 000 2 000 000X Y = 178 092 USD
which is lower than the unhedged 244 800 USD.
Erich Janka
FH bfi

2.2.3
Page 26 of 133
Minimum-Variance Hedge
Since we dont know the optimal number of contracts c, we treat it as variable

and get as risky position:
42 000 c X 2 000 000Y
Setting the first derivative of the volatility to zero gives the optimal c . In our
example:
2 000 000 Y
c =
XY 31
42 000 X
This hedge reduces the volatility to 140 115 USD which is a reduction of the risk
115
by 140
244 800 = 57.25 %
This maximal risk reduction depends only ton the value of the correlation:
q
optimal hedge
=
1 2XY
no hedge
2.3
Handling Ratios and Characteristics
Ratios
The foundation of quantitative analysis are ratios, characteristics, statistics, . . . to
interpret the results one should understand their properties and pitfalls.
If you are interested in a specific portfolio you usually dont have the time
or opportunity to check every detail. Asset managers want to protect their
intellectual property and therefore provide only condensed information, a good
deal of which is (more or less) standardized ratios. Each ratio focuses on a single
aspect of the portfolio. Even if you consider quite a lot of ratios: you wont get
the whole picture. Ratios (of portfolios) . . .
are quantified characteristics.
are more ore less standardized.
focus on a single aspect.
Understanding Ratios
Understanding has many aspects:
Erich Janka
FH bfi
Page 27 of 133
Applying a formula.
Knowing the assumptions.
Being able to follow the derivation.
Being able to derive the formula.
Knowing why it works.
Knowing when and why it fails.
Just being able to copy a formula and calculate some values can be very dangerous if applied to situations where it is not suitable. Finding a formula in a book
or the Internet is very easy, good derivations are harder to find and you have to
be quite lucky if the pitfalls are explained as well.
Anscombes Quartet: Data
i
x
ii
y
10.0 8.04
8.0 6.95
13.0 7.58
9.0 8.81
11.0 8.33
14.0 9.96
6.0 7.24
4.0 4.26
12.0 10.84
7.0 4.82
5.0 5.68
iii
10.0
8.0
13.0
9.0
11.0
14.0
6.0
4.0
12.0
7.0
5.0
9.14
8.14
8.74
8.77
9.26
8.10
6.13
3.10
9.13
7.26
4.74
iv
y
10.0 7.46
8.0 6.77
13.0 12.74
9.0 7.11
11.0 7.81
14.0 8.84
6.0 6.08
4.0 5.39
12.0 8.15
7.0 6.42
5.0 5.73
8.0 6.58
8.0 5.76
8.0 7.71
8.0 8.84
8.0 8.47
8.0 7.04
8.0 5.25
19.0 12.50
8.0 5.56
8.0 7.91
8.0 6.89
Anscombes Quartet: Ratios

These four sets of data have all of the following ratios in common:
mean of each x variable: 9.0
mean of each y variable: 7.5
variance of each x variable: 10.0
variance of each y variable: 3.75
correlation between each x and y variable: 0.82
Erich Janka
FH bfi
Page 28 of 133
linear regression line: y = 3 + 0.5x

residual sum of squares: 13.8
multiple R2 : 0.67
Anscombes Quartet: Graphs
Usually one assumes statistical

data to look like the scatter plots displayed in the first figure. To ensure this the
knowledge of the correlation coefficient wont help. The second figure shows
a very strong functional relation which is not linear. The last two figures could
include statistical outliers which deserve a closer look.
Blind Persons Describing an Elephant
Buddha once told the tale of blind men who touched an elephant to learn what
it is like. They could not agree about its nature because each one grasped only a
single part of the elephant.
With ratios in asset management it might be even worse: often someone else
chooses the characteristics he presents to you and if he has the chance he will
only choose those you are in his interests.
Erich Janka
FH bfi
Page 29 of 133
Why Ratios?
Ratios can have several purposes:
Identify portfolio characteristics.
Check portfolio guidelines or restrictions.
Satisfy an authority, auditor, . . .
Keep them in stock for possibly future use.
Compare the portfolio with its former version (structural interruption).
Compare the portfolio with other portfolios from the same asset management company.
Compare the portfolio with any similar portfolio.
Compare the portfolio with an arbitrary portfolio.
Ratios: Limited Comparability
A comparison might provide just little insight:
Only meaningful for specific asset classes. Shares pay dividends, bonds
have got a duration.
Use of ordinary returns vs. log-returns.
Details of method of calculation are not public. (The value at risk is very
complicated and every implementation has its own specialties.)
Erich Janka
FH bfi
Page 30 of 133
Time series have different timescales.

Time series do differ in length.
Different time zone used for end-of-day-quotes.
Days of time series dont match perfectly. (e. g. time-lag.)
Accuracy of ratiocalculation is very noisy. (E. g.: maximum draw-down.)
Calculation is erroneous.
Data are erroneous.
Ratio doesnt measure what you would expect.
There are many reasons why ratios might fail. Some ratios are only meaningful
for specific asset classes, others allow for some freedom in the calculation thus
hindering comparison with ratios or portfolios which use a different methodology. This may apply to different asset management companies but even for the
portfolio itself when the calculation method has changed.
Some ratios are so noisy that it is quite impossible to get the desired information
and comparing them doesnt make anything better . . .
(To become statistically meaningful noisy ratios need very long time series.)
Definition: Correlation
Asses/Product management deals a lot with all kinds of time series and their
dependencies.
Pearsons correlation coefficient is widely used to model dependence between
two quantities:
XY = XY [1, 1].
X Y
The sample estimates are:
v
u n
n
X
X
1
1 u
t ( xi x )2
s xy =
( xi x )(yi y )
sx =
n1
n1
i =1
i =1
The correlation coefficient is the covariance of X and Y divided by the product

of the standard deviations which ensures that the value of lies in the interval
[1, 1].
1 . . . perfect correlation
XY =
0 . . . no correlation
1 . . . perfect negative-correlation.
Erich Janka
FH bfi
Page 31 of 133
Correlation Examples
The following examples show that the correlation just measures the linear relationship between two data sets.
While the figures of the first two rows show images one might expect, the third
row displays only data sets with no linear correlation.
Origin of Correlation
Statistics can find correlations between sets of data X and Y but it wont help to
identify the origin of it.
Correlation occurs when . . .
Y depends on X.
X depends on Y.
X and Y share a common reason.
There is no underlying reason.
The data or the calculation is erroneous.
To decide which option is true, statistical analysis is not enough: First you should
have an idea of which connections might exist. It is dangerous to start a statistical
analysis and interpret the results afterward!
Logical Fallacies
Cum hoc ergo propter hoc With this, therefore because of this
Texas Sharpshooter
Fall for Fallacies
The best way to fall for those fallacies:
Erich Janka
FH bfi
Page 32 of 133
1. Do lots of data mining.

2. Find some correlations.
3. Set up a theory about the relationship.
Perfidy of Statistics
Statistics will find correlations
when there is a causal relationship but also
when there is no causal relationship and
statistics will find no correlations
when there is no causal relationship but also
when there is a causal relationship.
Beware of Statistics
He uses statistics as a drunken man uses lamppostsfor support

rather than for illumination.
(Andrew Lang)
Statistics can be made to prove anythingeven the truth.
(Author Unknown)
Torture numbers, and theyll confess to anything.
(Gregg Easterbrook)
Statistics can be very useful if used properly but can be misleading if not questioned.
When someone tells you how good his trading idea is, be aware that he might
just trick himself and those who believe them.
Prudent Statistics
In order to avoid those fallacies the following order should be followed:
Erich Janka
FH bfi
Page 33 of 133
1. Set up a theory about the relationship of the issues.

2. Do (statistical) analysis to check your assumption.
3. Reconsider your theory.
If you do this, you can still be wrong but you avoid many pitfalls.
Accuracy of Ratios: Standard Error
Since statistical data are usually incomplete one cant expect statistical ratios to
be exact.
Even if the true distribution with its ratios is known, the ratios derived from
a finite sample will always differ from it. The more observations the sample
contains, the smaller these differences will be. For a given distribution it is often
possible to calculate the magnitude in subject to the number of observations n.
The standard error is the standard deviation of a ratio. It measures how much
the ratio fluctuates statistically.
Zero Coupon Bonds
[OeNB]
Single fixed cash flow
at a fixed maturity
The profit results from the difference between the issue price and the redemption
price.
The yield on this bond is called spot rate
F ( T ) = 100 % = P( T ) 1 + s( T )
F ( T ) = 100 % = P( T )e
T
or
r(T )T
P( T ) . . . present value in % of the principal

F ( T ) . . . future value (at maturity)
s( T ) . . . discretely compounded annual spot rate
Erich Janka
FH bfi
Page 34 of 133
r ( T ) . . . continuously compounded annual spot rate

T . . . maturity
The spot rates vary with the credit risk of the issuer for identical maturities. The
basis is often a risk-free spot rate with a credit spread added.
An investor knowing the price of the bond can calculate its annual yield:

s( T ) =
F(T )
P( T )
1
1
or

1
F(T )
r ( T ) = ln
T
P( T )
When issuing a new bond, to determine its price, the current interest rate must
be known:
T
P( T ) = F ( T ) 1 + s( T )
or
P ( T ) = F ( T ) e r ( T ) T
Example 1 (Zero Bond)
Maturity
Interest rate
Issue price
Currency
Principal
10 years
0%
38.55 %
EUR
1 000 000
On the issue date, the price of the bond is EUR 385 500 per bond. If the issuer
is solvent, he has to pay back 1 million. Therefore the achieved interest rate
is:
r
100
10
1 10 %
or
s( T ) =
38.55

1
100
r(T ) =
9.53 %
ln
10
38.55
Foreign currency zero bonds also

1. First valuate, then convert:
Use the respective spot rate of the foreign currency for valuation and
convert the outcome at the current exchange rate.
Erich Janka
FH bfi
Page 35 of 133
2. First convert, then valuate:

Convert the known cash flows into euro at forward exchange rates and value
the resulting euro cash flows at the euro spot rates.
Both methods must result in (almost) the same outcome; otherwise, arbitrage
opportunities would exist.
If for some other products, the future cash-flows are not known, the first technique is th only possible.
To understand zero bonds the following must be mastered:
discrete and continuous interests
spot rates
credit risk
spot exchange rates
forward exchange rates
Coupon Bonds (Straight Bonds)
A bond pays
fixed interests (derived from the principal)
on specific days (chiefly annually or semiannually)
to its holder.
Example 2 (Coupon Bond)
Time to maturity:
Nominal interest rate:
Coupons:
Redemption:
Principal:
2 years
8%
semiannually
100 %
EUR 10 000
The holder of this bond will receive EUR 400 (= 82% 10 000) twice a year until
the bond matures. At maturity the principal is redeemed in full.
Erich Janka
FH bfi
Page 36 of 133
A long position in a coupon bond can be decomposed into several long zero
coupon bonds:
Coupon bond =
n
X
zero coupon bond(i )
i =1
The maturities of the zero coupon bonds are the coupon payment days and the
maturity of the bond: Every single cash flow is transformed into its own zero
coupon bond: The time of the cash flow is its maturity and the amount is its
principal.
Example 3
The bond given in the example above can be replicated using the following
zero coupon bonds:
0.5 years maturity and EUR 400 face value,
1 year maturity and EUR 400 face value,
1.5 years maturity and EUR 400 face value and
2 years maturity and EUR 10 400 face value,
The present value of the bond is:
Pbond = 400 P(0.5) + 400 P(1) + 400 P(1.5) + 10400 P(2)
Pbond . . . present value of the bond
P(t) . . . present value of a zero coupon bond with maturity t.
For bonds we therefore need many spot rates.

Bonds need not be redeemed at 100 % but a balance with the coupon rate must
hold: A higher redemption results in smaller coupons and vice versa; otherwise
arbitrage would be possible.
Even if redemption is 100 %, the coupon rate usually is not the spot rate corresponding to the maturity!
Floating Rate Notes (FRNs, Floaters)
Unlike straight bonds, floaters do not carry a fixed nominal interest rate. The
coupon payments are linked to the movement in a reference interest rate. These
Erich Janka
FH bfi
Page 37 of 133
often are money market rates like the EURIBOR. The adjustment of the coupons
happens in specific intervals, usually on the coupon date for the next coupon
period.
Example 4
A typical FRN could have features like this:
Maturity:
2 years
Nominal interest rate:6-month EURIBOR
Reset period:
every 6 months
Coupons:
semiannual, in arrears
Redemption:
100 %
Denomination:
EUR 10 000
Initial coupon corresponds to the 6-month EURIBOR at issue date. After half a
year, the coupon is paid and the actual 6-month EURIBOR is used to determine
the next coupon due in another half of a year; and so on.
The coupon frequently is defined as a reference interest rate plus a spread of x
basis points.
A FRN can be replicated by a zero coupon bond with
face value: Principal of the FRN plus the forthcoming coupon.
maturity: date of the forthcoming coupon.
This is because the value of a floater will be exactly 100 % at the coupon dates.
Even if a floater has a very long maturity, because of the coupon adjustments,
the interest rate risk is limited to the coming adjustment. Therefore they are
considered as money market instruments.
Of course these adjustments effect the credit risk only little.
Swaps
Same Currency
Plain Vanilla Swap: Fixed for floating
Basis Swap: Floating for floating
Erich Janka
FH bfi
Page 38 of 133
Cross Currency Swaps (CCR)

Fixed for fixed
Fixed for floating
Floating for floating
6.1
Plain Vanilla Swaps (Straight Swaps)
Interest rate swaps are a contractual agreement to exchange fixed interest payments for floating interest payments at specific dates on an agreed notional
amount. The floating interest rate is reference rate; usually a money market
rate such as the EURIBOR. In order to avoid settlement payments, the fixed rate
is chosen in a way that the initial value of the contract is zero. Generally, the
difference between the coupons at coupon dates are the only cash flowsthere
is no exchange of principals during the life of the swap.
Example 5 (Payers Swap)
Being long a Payers swap means paying the fixed coupons; otherwise its a
Receivers swap.
Term
Reference interest rate
Interest payment dates:
Swap rate:
Principal:
10 years
12-month EURIBOR; fist payment: 4.75 %
annual
5%
EUR 100
On the first payment date, the fixed-rate payer has to transfer EUR 5 per
contract and the floating-rate payer 4.75. Netting results in a payment of EUR
0.25 per contract to the counterpart (who is being short the swap). At the
same time the floating rate for the next period is determined.
Since swaps have a kind of symmetry, being long and short is not so clear;
therefore its better to think twice.
Interest rate swaps can be replicated as follows:
+Payers swap = +FRN coupon bond

+Receivers swap = FRN + coupon bond
+Receivers swap = Payers swap
Erich Janka
FH bfi
Page 39 of 133
The + and represent long and short positions; usually the + is omitted. Of
course, the FRN and the coupon bond themselves can be replicated using zero
coupon bonds.
6.2
Basis Swap
[Bier] and [Tuck] Basis swaps exchange floating rates for floating rates. This pair
must have a correlation less than 1 in order to be meaningful. They can be based
on different indices or have different time horizons. Examples in USD are:
T-Bills
vs. 3-month LIBOR
3-month LIBOR vs. 6-month LIBOR
To get an initial price of zero, there will be an adjustment to one of the rates. For
example, 1-month LIBOR + 1.5 basis points might be exchanged for 6-month
LIBOR.
6.3
Cross Currency Swaps
[Tuck] The two main differences to single currency swaps are:

The interest rates refer to different currencies and
the principals are exchanged
Example 6
Cross Currency Swap On October 14, 2011 the Canadian dollar traded for
0.98 US dollars and a trader might have agreed to the following:
At initiation: Pay 1 CAD and receive 0.98 USD
During lifetime: Receive 3-month CDOR plus 10 basis points on 1 CAD and
pay 3-month USD LIBOR quarterly on 0.98 USD for three years.
At expiration: Receive 1 CAD and pay 0.98 USD
Assuming only risk free rates, there wont be a premium. But when the USD
LIBOR bears more credit risk than CDOR, guaranteed cash flows in USD have
more value and therefore the CAD cash flows must be increased.
Erich Janka
FH bfi
Page 40 of 133
Forward Rate Agreements (FRAs)
A FRA is an over-the-counter (OTC) agreement that

a fixed interest rate (contract rate)
for a fixed principal (contract amount) will be paid for
a specified future period of time (contract period).
This is an agreement about a future money market transaction: the seller guarantees the buyer a future loan at the specified conditions.
Example 7 (3 12 FRA)
Contract period:
Reference interest rate:
Contract rate:
Contract amount:
in 3 months time for a period of 12 months

12-month EURIBOR
5%
EUR 100
Generally a m n FRA starts in m months and ends in n months; therefore the

length of the interest period will be n m months.
Bond Futures
This section (including examples) is based on [FT]. The basis value of these
futures are synthetic bonds with
a fixed time to maturity and
a fixed coupon.
This makes the quotes of bond futures comparable over time. They are very
popular instruments and havedepending on the marketvarious names:
Erich Janka
FH bfi

Future
T2M in y
Euro-Schatz
1.752.25
Euro-Bobl
4.55.5
Euro-Bund
8.510.5
Euro-Buxl
2435
ST Euro-BTP
23.25
LT Euro-BTP
8.511
CONF
813
US T-Note 10y
6.510
Page 41 of 133
Coupon
Curr
Tick size
Contr. Val.
from
6
6
6
4
6
6
6
6
EUR
EUR
EUR
EUR
EUR
EUR
CHF
USD
0.005
0.01
0.01
0.02
0.01
0.01
0.01
1
1
64 or 128
100,000
100,000
100,000
100,000
100,000
100,000
100,000
100,000
GER
GER
GER
GER
ITA
ITA
SUI
USA
Data of this table are from [EX] and [CME].

The buyer of a bond future is obliged to buy
specific bonds
with a notional equal to the contract value
to a specific price
at maturity of the future.
There is an asymmetry between buyer and seller of the bond future because the
seller can choose which bond he will sellthe cheapest to deliver. This might
for details.
give the seller the possibility of arbitrage (see [Wost]
The quotes of bond futures move similar to those of bonds: They increase with
falling interest rates and vice versa.
The tick size is the minimal movement of the futures price and corresponds to a
fixed value.
tick value = contract value tick size
for the Bund-future this would be:
EUR 10 = EUR 100 000 0.0001
Daly movements of more than 50 bp are quite common ([FTC]). The initial
margin is EUR 3 540 (Oct., 20 2011) and the maintenance margin should be about
2 832 which gives a possible leverage of 28 or 35.
According to [FTC], more than 97 % of all bond futures are evened before maturity. For the rest, things are a bit more complicated because
Erich Janka
FH bfi
Page 42 of 133
there is physical delivery

the basis is a synthetic bond.
At maturity, the Exchange Delivery Settlement Price (EDSP) is calculated a volumeweighted average of the last prices.
Example 8 (EDSP for Bund Futures)
The EDSP is the volume-weighted average of
all trades of the last trading minute if there were more than 10 trades or
otherwise:
the last 10 trades during the last half trading hour.
To make the possible bonds comparable to the synthetic one, conversion factors
are introduced:
PVclean
f conv =
100
f conv . . . conversion factor
PVclean . . . theoretical present value of the bond calculated with the yield given by the synthetic bond.
The conversion factor is used mainly to determine the Cash amount to be paid
for the delivered bond:

Cash =

EDSP
f
V + Acc
100 conf
EDSP . . . Exchange delivery settlement price

f conv . . . Conversion factor
V . . . Contract value
Acc . . . Accrued interests
(
f conf =
> 1 if Coupon(real) > Coupon(synthetic)

< 1 else
Example 9 (Bond Future)

Deliverable Bonds (exp: Dec. 2011, acc: Oct. 20 2011):
ISIN
DE0001135408
DE0001135416
DE0001135424
DE0001135440
ErichDE0001135457
Janka
Coupon Rate (%)
Maturity Date
3.00
2.25
2.50
3.25
2.25
04.07.2020
04.09.2020
04.01.2021
04.07.2021
04.09.2021
Conversion Factor
0.803418
0.750685
0.760622
0.803821
0.729389
FH bfi
Page 43 of 133
Example 10 (Bund Future)

The Bund future contract ends with a price of 107.72. The seller can deliver
one of the following bonds:
Bond A:
Maturity:
Coupon:
Quote:
f conv :
Acc:
10 years
5.375 %
102.90
0.9539995
0
The seller of the Bund future has to deliver bonds with a face value of 100 000
and gets:
Cash = 1.0772 0.9539995 100 000 EUR
= 102.764 EUR
To get the bond to the current price, he has to pay 102 900 which results in a
loss of
Loss = 102 764 102 900 = 136
Bond B:
Maturity:
Coupon:
Quote:
f conv :
Acc:
10 years
7.000 %
115.44
1.0736009
0
The seller of the Bund future has to deliver bonds with a face value of 100 000
and gets:
Cash = 1.0772 1.0736009 100 000 EUR
= 115.648 EUR
To get the bond to the current price, he has to pay 115 440 which results in a
profit of
Profit = 115 648 115 440 = 208
Since the seller of the bond future can decide, he will deliver bond B.
This calculated profit/loss is only because of an imperfect conversation factor
and adds to the profit/loss accumulated during the open position.
Erich Janka
FH bfi
Page 44 of 133
This imperfect calculation gives a different P/L for each deliverable bond. Therefore there will be a bond with the highest P/L, the cheapest to deliver (CDT).
A quick guess which bond is CDT is by calculating:

Spoti
min
f conv,i
i
which ignores the effect of accrued interests.
i . . . The index of the i-th bond
Spoti . . . the spot price of the i-th bond
8.1
Pricing
The seller of a bond future has to buy a bond (the CDT) an refinance this. Higher
refinancing cost will increase the price of the future. Coupons paid during the
open future bring a profit an reduce therefore the price of the future.
Using this replication, the future can be priced (no arbitrage).
The influencing factors are:
current price of the CDT
accrued interests
open interests till maturity of the future
refinancing costs
potential coupon payments
profit from reinvestments
Theor. futures price =
CP + FK E
f conv
CP . . . Clean price of CDT

FK . . . Refinancing costs for CDT
E . . . Fraction of coupon of CDT from current date to maturity of the future.
f conv . . . Conversion factor for CDT.
FK = (CP + Acc) refinancing yield Maturity

Erich Janka
FH bfi
Page 45 of 133
Example 11
We want to estimate the price of a future maturing in 150 days. The CDT
pays no coupons during this period and it holds:
Clean price:
Accrued interests:
Interests till maturity:
Coupon:
f conv :
Refinancing yield:
EUR 104
EUR 3
2.25
5.25 %
0.948594
4.0 % p.a.
Then the financing costs are:

FK = 107 0.04
150
= 1.78
360
and the price will be:

Price =
104 + 1.78 2.25

= 109.14
0.948594
This example demonstrates that in contrast to money market futures, bond

futures can have values greater than 100.
8.2
8.2.1
Usage
Hedging or Leveraging a Portfolio
In order to get smaller or higher exposure to yield changes, bond futures can be
useful if
The bonds are quite illiquid and have a high bid/ask spread.
The hedging/leverage is supposed to be for a short period.
Short selling of bonds is not possible.
8.2.2
Hedging a New Emission of Bonds
A company (bank) plans to issue a new bond with known timing. To hedge
against until the day of issue, bond futures can be uses.
Erich Janka
FH bfi
8.3
Page 46 of 133
Conversion Factors
The conversion factor makes it possible to compare real bonds with the notional
bond underlying of a bond future contract. To calculate the conversion factor
published on Eurex various formulas are used, depending on the characteristics
of the deliverable bonds. The basic idea is to discount all cash flows with a flat
interest rate equal to the coupon of the notional bond.
If the coupons are paid annually and there is no long or short coupon, an
investment of 1 results in the following present value which is equal to the
discount-factor:
f conv = V0 =
c
V1 (1 f )

f
cN
1 + 100
|
{z 100}
1
ACC
V1 is the future value at the time of the next coupon payment:

V1 =
c
c
c
c
+q
+ q2
+ qn
+ qn
100
100
100
100
with q =
1
cN
1 + 100
c 1 q n +1
+ qn
100 1 q
c
=
100
cN
(1+ 100
)
1
1 cN
n +1
1
c N n
1 + 100
1+ 100
!
cN
1 + 100
c
1
1
=
+
1

c
c N n
N
c N n +1
100
1 + 100
1
1 + 100
1 + 100
!
c
1
cN
1
=
1+
n +
c
c N n
N
cN
100
1 + 100
1 + 100
The discounting factors are based on the coupon of the notional bond and
therefore assuming a flat yield curve with a constant yield. Usually this doesnt
hold and there will be some mis-pricing: The factor introduced to make all valid
bonds equal doesnt hold its promise: There will be a cheapest to deliver (CTD)
bond. This mis-pricing can lead to arbitrage possibilities as demonstrated in
[Wost].
Erich Janka
FH bfi
8.4
Page 47 of 133
CHF-denominated Bonds
f conv =
1
1+
cN f
100
c
cN
cN
1
1+
cN n
100 (1 + 100
)
1
+
c N n
1 + 100
c (1 f )
100
n . . . Number of integer years until maturity of the bond.

f . . . Number of full months until the next coupon date, divided by 12 (except for f = 0,
where f 1 and n n 1).
c . . . Bond coupon.
c N . . . Notional coupon of futures contract.
8.5
EUR-denominated Bonds
The type of future allows for long and short coupons.
f conv =
1
1+
cN f
100
c
c i
+
100 act2 c N
!

1
c
e
cN
i
1
1+
cN n +
c N n
100 (1 + 100
100 act2 act1
)
1 + 100
e = NCD1 DD
(
NCD NCD1
act1 =
NCD1 NCD2
if e < 0
else
f = 1+
i = NCD1 LCD
(
NCD NCD1
act2 =
NCD1 NCD2
if i < 0
else
e
act1
DD . . . Delivery date.
NCD . . . Next coupon date after delivery date.
NCD1 . . . One year before the NCD.
NCD2 . . . Two years before the NCD.
LCD . . . Last coupon date before the delivery date. Start interest period if last coupon
date not available.
c . . . Coupon.
n . . . Integer years from the NCD until the maturity date of the bond.
c N . . . Notional coupon of future contract.
Erich Janka
FH bfi
8.6
Page 48 of 133
Euro-BTP Future
This future allows for adjustments to ensure coupons are paid on trading days.
f conv =
1+
cN c f
100
FCA +
N
X
i =0
100 c f 1 +
ci
i
cN c f + c
100
delay(ci )
f dc(d(ci ),d(ci +1 ))
+
1+
cN
100
N
cf
delay( N )
+ c dc(d( N ),d( N +1))
f
ACC
e = NCD1 DD
(
NCD NCD1
act1 =
NCD1 NCD2
e
f = 1+
act1
if e < 0
else
c i
FCA =
100 c f act2
i = NCD1 LCD
(
NCD NCD1
act2 =
NCD1 NCD2
c
ACC =
100 c f
if i < 0
else
i
e
act2 act1
c . . . Coupon Rate (e. g. 4 % = 4.0)

c N . . . Notional coupon
c f . . . Coupon frequency (1 . . . annual, 2 . . . semi-annual, 4 . . . quarterly)
DD . . . Delivery date
LCD . . . Last coupon date (if the first coupon date of the bond is later than the delivery
date DD, is the start of the accrual period)
NCD . . . Next coupon date (after the delivery date DD).
NCD1 . . . One coupon period after NCD.
NCD2 . . . Two coupon periods after NCD.
N . . . Number of full coupon periods between NCD.
ci . . . i-th coupon payment after NCD (c0 . . . Coupon payable at NCD).
d(ci ) . . . Regular payment date of i-th coupon payment.
delay(ci ) . . . Number of days between d(ci ) and the actual value date of the i-th coupon
payment (adjusts for d(ci ) falling on a weekend or holiday).
d( N ) . . . Maturity date; Regular payment date of the redemption value.
delay( N ) . . . Number of days between d( N ) and the actual value date of the redemption
payment.
d( N 1) . . . d( N ) plus one coupon period.
Erich Janka
FH bfi
Page 49 of 133
Eurodollar Future
Eurodollar:
Underlying: Eurodollar Time Deposit having a principal value of USD $1,000,000
with a three-month maturity.
Price Quote: Quoted in IMM Three-Month LIBOR index points or 100 minus
the rate on an annual basis over a 360 day year (e. g., a rate of 2.5 % shall
be quoted as 97.50). 1 basis point = .01 = $25.
Tick Size: One-quarter of one basis point (0.0025 = $6.25 per contract) in the
nearest expiring contract month; One-half of one basis point (0.005 = $12.50
per contract) in all other contract months. The new front-month contract
begins trading in 0.0025 increments on the same Trade Date as the Last
Trading Day of the expiring old front-month contract.
Contract months: Mar, Jun, Sep, Dec, extending out 10 years (total of 40 contracts) plus the four nearest serial expirations (months that are not in the
March quarterly cycle). The new contract month terminating 10 years in the
future is listed on the Tuesday following expiration of the front quarterly
contract month.
Last Trading Day: The second London bank business day prior to the third
Wednesday of the contract expiry month. Trading in the expiring contract
closes at 11:00 a.m. London Time on the last trading day.
Final Settlement: Expiring contracts are cash settled to 100 minus the British
Bankers Association survey of 3-month U.S. Dollar LIBOR on the last
trading day. Final settlement price will be rounded to four decimal places,
equal to 1/10,000 of a percent, or $0.25 per contract.
Margin requirements by July 2011:
Erich Janka
Start Month
End Month
Initial
Maintenance
07/2011
09/2012
09/2014
09/2015
06/2012
06/2014
06/2015
06/2021
608 USD
743 USD
743 USD
1013 USD
450 USD
550 USD
550 USD
750 USD
FH bfi
Page 50 of 133
[FTC]
An airline will buy new airplanes in half a year. It will lock in the current
level of interest rates. (Either it believes in raising interest or cant afford them.)
Therefore it will buy Eurodollar futures.
Current Quote: 97.3, Contract size: 1 million, 15 contracts. What are the initial
costs?
What are the costs and capital requirements?
What is to do and what happens? At the begin of the contract:
Capital requirement
Begin
Active
End
Erich Janka
Costs
Initial margin
Maintenance margin Variation margin
Last variation margin
FH bfi
Page 51 of 133
TED-Spread, Kreditlaufzeiten ublicherweise

langer als 3 monate parallelverschiebungen der Zinskurve
What happens? What is the amount we have to reserve now to get 1 million in
three months time with respect to the current three-months LIBOR?
We have to discount the million.
P2.7 % =
=
=
=
1, 000, 000 1, 000, 000 2.7 % 90/360

1, 000, 000 1, 000, 000 0.00675
1, 000, 000 6, 750
993, 250
Interest rate increases by 1 basis point:

P2.71 % = 1, 000, 000 1, 000, 000 2.71 % 90/360
= 1, 000, 000 1, 000, 000 0.006775
= 1, 000, 000 6, 775
Erich Janka
FH bfi
Page 52 of 133
= 993, 225
This results in a loss of:
P2.70 % P2.71 % = 993, 250 993, 225
= 25 USD

90
90
P/L(k BP) = 1MM 1MM r0
1MM 1MM (r0 k |{z}
BP )
360
360
0.01
= 1MM (r0 /4 r0 /4 0.0025 k)

= 25 k USD
10
Leverage and Margins
Investors can increase their exposure to the market by borrowing money and
investing it as well.
An investor owns 1000 shares worth 20 each. If the stocks rise to 25, he makes a
5000
profit of 1000 (25 20) = 5000, which is a return of 1000
20 = 25 %.
If the investor can borrow another 20 000, he could double his profit to 10 000,
which is a return of 50 %.
Of course if the stock declines, his loss will also double; The investor has a
leverage of 2 or 200 %.
On the other hand if the investor is not fully invested and he has some spare
cash, his leverage will be smaller than 1:
Our investor now owns only 500 shares and has 10 000 cash. Then his profit will
be only 500 5 = 2500, which is a return of 12.5 %. His leverage is 50%.
Short selling can result in even negative leverage.
The leverage is the lever of the the arbitrage pricing model: Depending on the
utility function of the investor he will choose his leverage of the market portfolio.
As explained in [Fab], investors can borrow money from brokers: The possible
amount is restricted: There are laws, exchange regulations and the brokers own
rules, resulting in the margin requirements.
Erich Janka
FH bfi
Page 53 of 133
The initial margin requirement is the proportion of the total market value of the
securities that the investor must pay as an equity share, and the remainder is
borrowed from the broker.
There is also a maintenance margin requirement which is the minimum proportion
of
the equity in the investors margin account
the total market value.
If the investors margin falls below the minimum maintenance margin (which
could happen if the share price fell), the investor is required to put up additional
cash; the investor receives a margin call from the broker specifying the additional
cash to be put into the investors margin account. If the investor fails to do so,
the broker may sell securities in the investors account.
possible leverage =
1
margin (in %)
(1)
Assume the initial margin is 50 % and the maintenance margin is 25 %, then with
20 000 an investor could buy 2000 stocks worth 20 each. If the share price drops
to 15, the stock value is 30 000 and the market value would be 2000 15 20 000 =
000
10 000 and their ratio is 10
30 000 33 % > 25 %. Therefore this drop does not lead
to a margin call. If the share price drops further to 13, the stock value is 26000
and the market value is 6000 with a ratio of 266000
000 = 23.08 %.
The following ratio has to be above the margin requirements
market value
> margin requirement
stock value
MV = S p L + C
SV = S p
and therefore
LC
1
> mr
S p
MV . . . Market value of portfolio
SV . . . Total value of shares
mr . . . Margin requirement
Erich Janka
FH bfi
Page 54 of 133
S . . . Number of shares
p . . . Price of one share
L . . . Liabilities
C . . . Cash
The product manager now wants to avoid margin calls until the share price falls
below 13. He has to sell some of his shares or must add cash. How much he has
to add depends on whether he wants to keep the cash on his portfolio account
or invests in more shares.
If he invests the money, the following must hold:
1
LC
= mr
S p + Sa p
Sa . . . additional shares
And therefore he needs this amount:

Sa p =
LC
S p
1 mr
Continuing our example, we would need

20 000 0
2000 13 = 666.67
1 0.25
(Sa = 52)
This amount is only valid when the price is exactly 13. Buying the 52 shares
earlier is more expensive.
If we want the cash being on the account, this condition must hold:
1
L C Ca
= mr
S p
Ca . . . additional cash
solving this for Ca gives:

Ca = L C (1 mr) S p
With the numbers from the example:
20 000 0 (1 0.25) 2000 13 = 500
If it comes to a margin call, it is not enough to comply with the maintenance margin because then consecutive margin calls would be highly probable. Therefore
Erich Janka
FH bfi
Page 55 of 133
the initial margin requirements are used. When there is no reaction, parts of the
portfolio will be sold.
Margins are not only valid when buying stocks but also when trading derivatives.
The exchanges publish their margin requirements. The brokers can use these or
use even stronger requirements. Using derivatives much higher leverages than 2
are possible.
The value of futures is adjusted every trading day: When there was a loss, the
account will be reduced by the corresponding amount of cash. Profits on futures
lead to a cash payment. This cash transfers are called variation margin.
We now know the following types of margins:
Initial margin
Maintenance margin
Variation margin
The exact values of the initial and maintenance margin are set by the exchange
and can vary depending on the current volatility of the underlying.
Up to this point we ignored the interest the investor has to pay when he borrows
money. For exact calculations these must also be taken into account.
The leverage can be used as a measure of relative risk. A leverage of 1.5 means
that the portfolio has 50 % more risk than a when exactly the NAV is invested.
To know if this is high risk or not, one has to know the basis investment. For
example, a money market portfolio with a leverage of 2 usually is much less
risky than a share portfolio with no leverage at all.
11
Arbitrage
All current methods of pricing derivative assets utilize the notion of arbitrage
([Nevt]):
Arbitrage pricing methods: Asset prices are obtained from conditions that
preclude arbitrage opportunities.
Equilibrium pricing methods: Lack of arbitrage opportunities is part of
general equilibrium conditions.
Erich Janka
FH bfi
Page 56 of 133
First kind: Starting with a portfolio with no net commitment and ending
with a sure positive profit.
Second kind: Starting with a negative net commitment and yielding nonnegative profits in the future.
[Wilm] gives another classification of arbitrage:
Static arbitrage: Re-balancing of positions is not required.
Dynamic arbitrage: Depending on market circumstances, trading instruments will be required.
Model-dependent arbitrage: The arbitrage opportunity depends on the
validity of a model (e. g. Black-Scholes).
Statistical arbitrage: Statistical arbitrage is not an arbitrage! Based on past
statistics, probable future market conditions are exploited.
Most arbitrage opportunities cant be exploited for various reasons ([Wilm]):
Others already exploited the opportunity.
Quoted prices are wrong or not tradable.
Compared prices were not quoted synchronously.
There is a bid-offer spread which was not accounted for.
The Model is wrong or there is an overlooked risk factor.
Even if arbitrage opportunities exist, arbitrage-free prices are still of interest
([Nevt]):
Creating new products: There are no market prices yet and therefore a
reasonable price is needed.
Risk Management: Stress-tests often assume worst case scenarios. Since
these scenarios are fictional, there are no observable market prices.
Marking to market of assets: When the back-office or a trader wants to
know the current price of a non-liquid asset which has not been traded
lately.
Trading opportunities: Comparing arbitrage-free prices with those traded,
might detect underpricedand therefore cheapassets.
Erich Janka
FH bfi
12
Page 57 of 133
Convertible Bonds
[OeNB] Convertible bonds are corporate bonds including the right to exchange
the bond
for a specific number of shares of the issuing company according to the
conversion ratio
at predetermined dates
to predetermined prices.
There can be several call dates with individual prices.
A convertible bond provides debt capital to the issuing company at first which
might change to equity capital later. The conversion right is just a type of option
which can be of European, Bermudan or American style.
For American and Bermudan options there are two additional aspects to conversion:
The investor looses the accrued interests upon conversion
The conversion ratio is predetermined, but the applicable strike price is not
known in advance; because it depends on the current value of the bond.
European style convertibles are more straightforward containing a simple European option. But since this option is a warrant, it has influence on the price of
the underlying share price: When the holder of the convertible bond chooses to
convert, he makes a profit while the issuer has to take a loss and this should be
reflected in the companys stock price.
When the stock price increases, also does the probability of conversion and on
conversion, the bond holder takes her share of the profits. Thus the price of the
stock haves differently in the presence of conversion rights.
Example 12
A company has issued a convertible bond with face value Debt(= 20) and a
conversion rate c of 45:1. The equity (= 80) of the company is represented by
80 shares of stock.
The value of the share immediately before the conversion date is 100, what is
the proper price p of the share? Since the debt is 20, the equity of the company
Erich Janka
FH bfi
Page 58 of 133
is Equity = 80. Without the conversion right, the price would be:
p =
Equity
80
=
= 1
Shares
80
With the conversion rights, the price immediately before and after conversion
must be equal: The company issues 45 new shares and the debt is converted
to equity. The company value is still 100 but now there are 80 + 45 = 125
shares. The price must be
p =
Equity + Debt
100
=
= 0.8
Shares + c
125
The following table shows the value of the convertible bond immediately
before conversion depending on varying quotes of the share:
Quote
Converted Value
Bond Redemption
0.1
0.2
0.4
0.8
1.0
4.5
9.0
20.0
36.0
45.0
20
20
20
20
20
Instead of issuing new shares, the company can also buy the 45 shares on the
stock market and then give them to the investor. What would in this case the
fair price be? The companys value after buying the shares would be initial
value minus the amount paid for the shares. Since the number of shares is
unchanged, the price must fulfill:
100 45 p
80
p = 0.8
p =
and therefore
20
a conversion will happen.
When the share price is above Debt/c = 45
= 0.4,
When the price of the share is below this threshold, only the shareholder
45
profit from increasing prices. When the price is above, Equityc+Debt = 125
36 %
of the increase will go to the holder of the convertible.
Erich Janka
FH bfi
12.1
Page 59 of 133
Replication
A European-style convertible bond can be replicated like this:

Convertible bond = Zero coupon bonds + Warrants
Bermudan and American convertibles cant be replicated this way since after
conversion no further coupons are paid. The strike price of the coupon is also
not determined at the beginning since it is
strike = time value of the bond conversion ratio
12.2
Valuation
The value of a convertible bond is not path dependent; former prices have no
influence to the current price. Therefore binomial trees can be used for valuation.
For European-style convertibles, the value of the warrant can be found using the
following formula:

Nx
ZN (d1 ) X exp(rT ) N (d2 )
N + Mx
M
Z = S+ W
N

2
Z
ln X
+ r + 2Z T
d1 =
Z T
d2 = d1 Z T
W =
(2)
W . . . the premium on warrants on the companys own stock

S . . . price of the stock
Z . . . price of the stock plus the proportionate value of the warrants
X . . . strike price
M . . . number of warrants outstanding
N . . . number of shares outstanding
x . . . exercise ratio (shares per warrant)
T . . . option maturity
r . . . risk free interest rate
Z . . . volatility of Z
Please note, that W is on both sides of equation (2) and there is no closed form for
W. Therefore the solution must be found numerically using Newtons method
or the regula falsi.
Erich Janka
FH bfi
13
Page 60 of 133
Bootstrapping
To get spot rates yield curve the best would be to have lots of government zero
coupon bonds. But most of those bonds bear coupons and the rate of the coupon
is usually not the spot rate corresponding to its maturity.
It is possible to estimate the spot yield curve from those bonds. In the general
case this requires quite sophisticated methods.
But there is a simple case, where the solution is very easy and called bootstrapping.
If one has a sequence of coupon bearing bonds:
All bonds have an identical coupon frequency.
The maturity of the first bond is exactly the length of a coupon period.
(This is equivalent to a zero coupon bond.)
Each bond has exactly on coupon more than its predecessor.
The dates of coupons are concurrent.
Bootstrapping can be used if there are enough government bonds available or
one can transform a bond yield curve to spot rates.
13.1
Discount Factors
Starting with the first bond, it is possible to calculate the discount factors (which
correspond to yields) by induction.
dn
d1
t0
d2
t1
d3
t2
dn-1
t3
1
dT =
Erich Janka
tn-1
rT
m
tn
dt
t< T
1 + rmT
FH bfi
Page 61 of 133
dt . . . Discount factor at t
rt . . . interest rate p. a.
m . . . Number of coupons p. a.
13.2
Spot Rates
Since the discount factors correspond tho the spot rates in a known way, it is
possible to derive the latter:
PV = dt FV

st mt
= FV 1 +
m
st . . . Spot rate for time interval t.

st mt
dt = 1 +
m

1
mt
st = m dt 1
14
Forward Rates
s T mT
m

st mt
f t T m( T t)
= 1+
1+
m
m
FV =
1+
f tT . . . Forward rate beginning at t and ending at T.
!1
s T T T t
1+ m
f t T = m
1

st t
1+ m
15
Interpolation of Yield Curves
Usually yield curves have only values for distinctive time horizons. In order to
get also values in between or even outside the range, interpolation is necessary.
Suppose we know the yields y1 and y2 for times t1 and t2 and need an estimate
of the yield y at time t1 t2 . There are several methods to do this:
Erich Janka
FH bfi
Page 62 of 133
Linear interpolation on yields,

linear interpolation on discount factors,
Raw inerpolation (linear on the log of discount factors),
interpolation with polynomials,
cubic splines,
monotone convex inerpolation,
...
The figures show the difference between linear interpolation on yields and linear
interpolation on dicount factors.
The raw interpolation corresponds to piecewise constant forward curves. It is
easy to implement and has good properties.
y =
y2 t2 ( t1 ) + y2 t2 ( t2 )
t2 t1
Interpolation using polynomials usually gives bad results.

Maybe the best results are obtained using monotone convex interpolation. Details on this method can be found in [Hagan]. There is also a list of criteria to
check the quality of the methods
Are forwards positive? If not, the curve might imply arbitrage.
Are forwards smooth? Forward curves usually dont show jumps.
Are the forwards stable? Small changes of the inputs should result in small
changes of the forward rates.
Erich Janka
FH bfi
Page 63 of 133
Is the method local? Input changes on a certain location should not affect the
curve far away.
Are derived hedges stable? If hedges ar calculated using the yield curve, small
changes of the inputs should only have small consequences on the hedge.
FW pos.
Raw
Other linear
Natural cubic
Monotone convex
15.1
!
%
%
!
FW smooth
FW stable
local
Hedge stable
not cont.
not cont.
smooth
cont.
+++
+++
++
+++
+++
+
+
++
++
Accrued Interests
Between coupon payments the holder of a bond is entitled to a fraction of the

next coupon, the accrued interests. Bonds are often quoted without these accrued
interests, the clean price; when the accrued interests are added, one gets the
dirty price. There are many conventions how these accrued interests must be
calculated, depending on the way the days are counted and when trades are
settled. This section deals with the rules valid for the SWX Swiss Exchange and
is based on [Chri].
To calculate the accrued interests (if they apply for the trade) we need:
The settlement date.
The date (D1.M1.Y1) where accrued interest starts.
Usually this is the start of the interest period within which the settlement
date falls.
The date (D2.M2.Y2) to which accrued interest is calculated.
By default this is the settlement date, unless it is equal to or later than the
maturity. Then it is set to the maturity.
Determine the number of days for which interest is to be accrued.
The day counting conventions.
The relevant dates are depicted in figure 4 on the following page.
In order to find the settlement date which usually is the date where entitlement
for accrued interests ends, the following has to be taken into account:
Erich Janka
FH bfi
Page 64 of 133
starting date for accrued interest
date of next coupon
length of interest period
number of interest bearing days

time
D1.M1.Y1
coupon n
D2.M2.Y2
settlement
D3.M3.Y3
coupon n
Figure 4: Relevant dates for accrued interests
Settled in
USD
USD
CHF
Trade date
Settlement date
Comment
Monday, July 1st

Monday, July 2nd
Monday, July 1st
Friday, July 5th

Friday, July 8th
Friday, July 4th
July, 4th is currency holiday

July, 4th is currency holiday
Table 1: Settlement dates: examples
Usually the settlement date is the trade date plus 3 business days (T+3).
The weekend day calender
The settlement currency holiday calender for the security in question.
The clearing organization holiday calendar for the security in question.
The market holiday calendar is irrelevant.
Table 1 gives some examples of how to determine the settlement date.
If one of the following conditions holds, no accrued interests are paid:
Settlement date first date of interest entitlement.
The settlement date falls on an interest payment date.
Settlement date maturity.
trade date in default from date.
The null day count method is used.
Erich Janka
FH bfi

15.1.1
Page 65 of 133
Day Counting Conventions and Holiday Calendars
15.1.2
30/360 (German): N = ( D2 D1) + 30 ( M2 M1) + 360 (Y2 Y1)
The year has 360 days.
If D1 or D2 is 31, then use the value 31 instead. Thus, on the 31st of a
month the number of accrued interest days equals those of the 30th .
The last day of February is treated is treated as the 30th day of the
month.
N
A = coupon 360
30S/360 (Special German): N = ( D2 D1) + 30 ( M2 M1) + 360 (Y2 Y1)

the year has 360 days.
If D1 or D2 is 31, then use the value 31 instead. Thus, on the 31st of a
month the number of accrued interest days equals those of the 30th .
The last day of February is not treated specially.
N
A = coupon 360
act/365 (English): N = Number of days between D1.M1.Y1 and D2.M2.Y2

N
A = coupon 365
act/360 (French): N = Number of days between D1.M1.Y1 and D2.M2.Y2

N
the year has 360 days. A = coupon 360
30U/360 (US): N = ( D2 D1) + 30 ( M2 M1) + 360 (Y2 Y1)

1. If D2 is the last day of February and D1 is the last day of February,
change D2 to 30.
2. If D1 is the last day of February, change it to 30.
3. If D2 is 31 and D1 is 30 or 31, change D2 to 30.
4. If D1 is 31, change D1 to 30.
Erich Janka
FH bfi
Page 66 of 133
N
A = coupon 360
act/365L (ISMA-Year): N = number of days between D1.M1.Y1 and D2.M2.Y2

A = coupon N
Y
If the coupon frequency is annual, then Y = 366 if February 29th is
included in the interest period, else Y = 365.
If the coupon frequency is not annual, then Y = 366 for each interest
period where D3.M3.Y3 falls in a leap year, else Y = 365.
act/act (ISMA-99): For details see [Chri].
act/act: NL = # days between D1.M1.Y1 and D2.M2.Y2 which fall in a leap year.
NN = number of days which fall in a non leap year.
The interest bearing days are split whether they belong to leap years
or non leap years.

NN
NL
A = coupon 366 + 365
The last listed method is from [Para] where many more methods are mentioned
and alternative names are given.
16
16.1
Interest Rates and Returns

Discrete and Continuous Interests
If the interests are paid at discrete times, e. g. once a year as usual for saving
accounts, the following formulas hold for these discrete interests which are given
on an annualized basis.
With an initial capital of V0 the following holds:
after 1 year:
after 2 years:
V1 = V0 + rV0 = V0 (1 + r )
(3)
V2 = V1 + rV1 = V1 (1 + r ) = V0 (1 + r )(1 + r ) = V0 (1 + r )
(4)
after n years:
Erich Janka
Vn = V0 (1 + r )n
for integers n.
(5)
FH bfi
Page 67 of 133
If the interest rate is unknown but the future value Vn is,

r =
Vn
V0
1
(6)
Interests are paid semi-annually (r 1 is an annual interest rate):

2

r1
V1 = V0 1 + 2
2
2

r 1 2
V1 = V0 1 + 2
2
2
1
after year:
2
after 1 year:
(7)
(8)
Since r 6= r 1 , it is important to know

2
the time interval the interest rate refers to and

the frequency of the interest payments.
This information wont be necessary for discrete interests.
Interests are paid monthly:
after 1 year:

r 1 12
V1 = V0 1 + 12
12
(9)
Even if the interests are added every single moment, the value is limited and
resulting in continuous interests:
Interests are added m times per year:

V1 = V0 1 +
r 1 m
m
(10)
When letting m
V1 = V0 erc
(11)
Vn = V0 enrc
(12)
For n years holds:
Erich Janka
FH bfi
Page 68 of 133
This holds even for any period a (in years):

Va = V0 earc
(13)
This is how the exponential function and with it the logarithm as its inverse
function enters finance:

1
Va
rc = ln
(14)
a
V0
If only one type of interest rate is known, the other can be calculated as follows:
V1 = V0 (1 + r )
and
V1 = V0 erc
(15)
rc = ln(1 + r )
(16)
and therefore
r = erc 1
16.2
and
Arbitrary Intervals
The value of an investment is determined at any time when based on discrete

interests. This doesnt hold for discrete interests: it is still not clear what to do
when the period of interest is not a whole year or the valuation date differs from
dates when interests are paid. Then one has to check up the products details or
the markets conventions.
Erich Janka
FH bfi
Page 69 of 133
For the fractional part the interests can be supposed to grow linear or also
exponential:
(17)
after a years:
V0 (1 + r )
?
Va = V0 (1 + ra)
(18)
Saving accounts, credit account and bonds usually the fractional part of a is
supposed to be linear.
Between the dates with interest payments, especially before the first payment,
the calculation of the value can be ambiguous. For periods smaller than one year
it is common to assume a linear growth while for longer periods, equation (17)
is more appropriate, as shown in equation (18).
If the date of interest t differs from a date with interest payment, the calculation of equation (18) depends on the day counting conventions explained in
section 15.1.1 on page 65.
16.3
Interests and Return
Equation (6) on page 67 and equation (14) on the previous page are also the
formulas to calculate the annualized returns for all kind of financial products: If
an asset is worth Vt0 at the begin of a period and Vt1 at its end, the period return
is

Vt1
Vt0
1
and its log-return:
rlog = ln
(19)
r=
Vt0
Vt0
or annualized, when t1 t0 equals a years:

r=
Vt0
Vt0
1
and its log-return:
rlog
1
= ln
a
Vt1
Vt0

(20)
(21)
Interest rates and returns are quite similar; we will even use the same symbol r.
Usually the return of financial products is given as a discrete return or interest
rate for one year (p. a.). For periods smaller than one year the returns can be
Erich Janka
FH bfi
Page 70 of 133
reported for that specific period. This is done to avoid the impression of an extrapolation of the returns which could be interpreted as a prognosis. Sometimes
the overall return accumulated for several years is reported because this looks
more attractive.
Log-Returns are the are similar to discrete interests. Log-Returns are easier to
handle and used in essential finance-mathematical frameworks like the BlackScholes formula; so the logarithms of the returns are supposed to be normally
distributed and not the simple returns. This can be interpreted as follows: Capital
gains are not additive but multiplicative.
16.4
Approximation
1
Let w = V
V0 be the wealth ratio ; Simple returns are a first order Taylor approximation of log-returns at r = 0 or w = 1:
f (w) = ln(w)
f 0 (w) =
1
w
f 00 =
1
w2
(22)
and
f 0 (1)
f 00 (1)
( w 1) +
( w 1)2 +
1!
2!
ln(w) = 0 + 1(w 1) +
f ( w ) = f (1) +
(23)
(24)
and therefore
rlog = ln(
V1
V
) 1 1 = r
V0
V0
(25)
This means that for small returns, the difference between the two will be also
very small. A similar relationship holds for the accumulation factors for discrete
and continuous interests:
rc ln(1 + r )
or
r erc 1
(26)
This approximation can be used to justify the usage of simple returns instead of
log-returns for small movements or short time intervals.
Erich Janka
FH bfi
16.5
Page 71 of 133
Multi Period Returns
Compounded Return
The return of multiple periods can be compounded very easily when there are
no external cash flows. Let VS = V0 , V1 , . . . , Vn1 , Vn = VE the values of the
portfolios at distinct times.
1+r =
V
V2
VE
V1
V
n 1 E
=
VS
VS
V1
Vn2 Vn1
|{z}
|{z}
| {z
} | {z }
1+r1
1+r2
1 + r n 1
1+r n
1 + r = (1 + r1 )(1 + r2 ) (1 + rn1 )(1 + rn )

This process of combining the returns of sub-periods to the return of the entire
period is called geometric or chain linking.
Example: Chain Linking
Begin of year
End 1st quarter
End 2nd quarter
End 3rd quarter
End of year
Market value (e)
Return (%)
100
97
95
102
108
3.00
2.06
7.37
5.88
VS = V0
V1
V2
V3
V4 = VE
Total
r =
8.00
97
95 102 108
108
1 =
1 = 8 %
100
97
95
102
100
|{z} |{z} |{z} |{z}
|{z}
0.970
0.979
1.074
1.059
1.08
External Cash Flow

When new money is invested in the portfolio or is withdrawn, this is called
external cash flow C and it must be considered when calculating the return since
the portfolio manager has usually no influence on that.
VE C
VS
VE
1+r =
VS + C
1+r =
Erich Janka
FH bfi
Page 72 of 133
1+r =
VE C/2
VS + C/2
(Dietz Method)
When there is just a single external cash flow C, the formula for calculating the
return depends on the time when C was available for the portfolio manager.
The first formula is valid if the money was available at the end of the period,
the second formula if it was available right at the beginning of the period and
the third holds for the case when it came in just in the middle. When there
are more external cash flows at various times, calculating the portfolio return is
more complicated.
Portfolio Return
Internal rate of return

Modified Dietz
Linked modified Dietz
True time-weighted
Money Weighted Return
VE = VS (1 + r ) +
n
X
Ci (1 + r )
T ti
T
i =1
r . . . internal rate of return,
Ci . . . i-th external cash flow,
ti . . . time of i-th cash flow (t0 = tS = 0),
T . . . end of total period (T = t E ).
Internal Rate of Return: Example 1

The internal rate of return is a hypothetical constant interest rate for all assets.
Date
Dec, 31
Jan, 17
Feb, 10
May, 19
Jun, 30
Erich Janka
Value
0
17
41
139
100
220
Cash flow
With interests
100
200
200
106.70
106.05
210.28
203.03
220.00
FH bfi
Page 73 of 133
IRR = r = 6.70 %
IRR Example 1
Internal Rate of Return: Example 2
Value
0.0
0.2
0.8
0.9
T=1
1000
(
IRR = r =
Cash flow
1200
200
100
110
With interests
r = 10.26 % r = 87.04 %
1102.58
1297.51
203.94
100.98
110.00
129.60
234.03
132.91
81.52
110.00
+10.26 %
87.04 %
IRR Example 2
Erich Janka
FH bfi
Page 74 of 133
Modified Dietz Method

Austrian pension funds have used this method to calculate their performance
since 2000.
VE VS C
r=
P
ti
VS + Ci T
T
C . . . sum of all cash flows within the period.
Since this method assumes that all cash
flows happen at the same time, the approximation is growing worse for longer
periods. To reduce the effect, modified Dietz can be used for sub-periods and
chain-linking the results.
Modified Dietz: Example
Value
Dec, 31
0
Jan, 17
17
Feb, 10
41
May, 19 139
Jun, 30 181
100
Date
Cash flow
100
200
200
220
r = 6.69 %
Linked Modified Dietz: Example
As a rule of thumb, a new period should be started at the latest when the external
cash flows exceed 10 % of the assets.
Erich Janka
FH bfi
Page 75 of 133
t
Value
Dec, 31
0
Jan, 17
17
Feb, 10
41
Mar, 31
90
May, 19 139
Jun, 30 181
100
Date
Cash flow
Return
100
200
450
17.24 %
200
7.43 %
220
r = 8.53 %
Modified Dietz: Linear Approximation of IRR
When the exponential interest of the IRR is replaced by linear interests, the result
is the modified Dietz method. Solving the second equation for r shows this.
VE = VS (1 + r ) +
n
X
Ci (1 + r )
T ti
T
i =1
VE = VS (1 + r ) +
n
X
i =1
Ci (1 +
T ti
r)
T
Since the function of example 1 is very close to a line, it is not surprising that
both methods give very similar returns (6.7 % and 6.69 %).
Newton Iteration
This is the first step of Newtons method.
x n +1 = x n
Newtons method:
f (r ) = VS (1 + r ) +
n
X
f ( xn )
f 0 ( xn )
ti
Ci (1 + r )1 T VE
i =1
f 0 (r ) = VS +
n
X
i =1
r0 = 0 :
Erich Janka
r1 = 0
Ci
ti
T ti
(1 + r ) T
T
VS + C VE
VE VS C
=
Pn T t i
P
ti
VS + i=1 T Ci
VS + in=1 T
T Ci
FH bfi
Page 76 of 133
Modified Dietz and IRR: Illustration
Time-Weighted Return
When each time period has equal weightregardless of the NAVwhen calculating the rates, the returns are called time-weighted.
This is usually the return calculation methodology asset managers prefer since
they cant influence the amounts invested in their portfolios.
V1 C1 V2 C2
V
Cn1 Vn Cn
n 1
V0
V1
Vn2
Vn1
1+r =
Example: Time-Weighted Return
Value
Dec, 31
0
Jan, 17
17
Feb, 10
41
Mar, 31
90
May, 19 139
Jun, 30 181
100
175
420
450
210
220
Date
Cash flow
Return
Cumulated
25.00 %
25.71 %
7.14 %
200 8.89 %
4.76 %
25.00 %
5.71 %
1.02 %
7.96 %
3.58 %
100
200
r = 3.58 %
Example: Returns
Erich Janka
FH bfi

Method
IRR
Dietz
Linked Dietz
Modified Dietz
Linked M. Dietz
Time Weighted
Page 77 of 133
Return
6.70 %
13.33 %
9.71 %
6.69 %
8.53 %
3.58 %
Time-Weighted vs. Money-Weighted Returns
Money weighted: Each money unit has the same weight.

Time weighted: Each time interval has the same impact.
An investor might be interested in comparing his profit/loss with a constant interest rate and therefore in the IRR. The portfolio manager is usually not interested
in money weighted returns because he cant influence when the investor adds
money or withdraws it. He is measured by the time weighted performance.
Annualized Returns
The length of periods can be very different. For better comparability periods
longer than one year usually are annualized.
r m t = (1 + r t ) m 1
(time scaled return)
rt is the return calculated for the original time interval t.

m multiple of t.
Annualized Return: Example

After three years a portfolio has a return of 11 %. The annualized return is:
r1year = (1 + r3years )1/3 1

3.54 % = (1 + 0.11)1/3 1
Continuously Compounded Returns

Simple returns are positively biased, whereas continuously compounded returns
are not. The interests on a bank account compound over timewe receive
Erich Janka
FH bfi
Page 78 of 133
interests for our interests. The more often interests are paid, the higher is the
effective return r. The compounded return for n payments:

r n
1 + r = 1 + 1/n
n
Continuous case:
r1/n n
1 + r = lim 1 +
= erc
n
n
rc = ln(1 + r )

Continuous compounded returns are additive:

ln(1 + r) = ln(1 + r1 ) + ln(1 + r2 ) + + ln(1 + rn )
Compounded Returns: Example
Time Value
0 100.00
1
2 106.00
1 112.36
Total
112.36
Interests
Interest rate
6.00
6.36
12.00 %
12.00 %
12.36
12.36 %
Figure: Discrete and Continuous Returns
Erich Janka
FH bfi
Page 79 of 133
Discrete Returns approximate Continuous Returns

The simple (discrete) return is the first step of a Taylor series approximation of
the continuous return at r = 0. For sufficient small r the difference therefore is
insignificant.
f (r ) = ln(1 + r )
f 0 (0)
f 00 (0) 2
T f (r ) = f (0) +
r+
r +
|{z}
1! }
2!
| {z
0
= r r /2 +
Approximation of Continuous Returns
Fees
A portfolio pays lots of fees. These fees must be considered properly when
evaluating and comparing the performance. There are three basic types of fees:
Transaction fees. These fees are directly related to trading assets.
Portfolio management fee. The payment of the asset manager.
Custody and other administrative fees. These include audit fees, legal fees,
...
Erich Janka
FH bfi
Page 80 of 133
Gross- and Net-Of-Fee Returns

The performance can be calculated before (or gross of) fees or after them (net of
fees). The portfolio manager should be measured net of all fees he can control
such as transaction fees.
Since portfolio management fees can be negotiated or paid from another source,
comparison with other portfolios should be done gross of them.
Of course the investor is interested in values net of all fees.
The best way to calculate gross returns is to treat the fees as external cash flows.
If only limited data are available, they can be estimated as follows:
Grossing up net returns:
rG = (1 + r N )(1 + f ) 1
Netting down gross returns:
rN =
1 + rG
1
1+ f
rG . . . return gross of fees,

r N . . . return net of fees,
f . . . periods fee rate.
Portfolio Component Returns

To understand the return of a portfolio, its overall return is not enough. The
return of relevant subsets is also essential. These subsets can be defined by
sectors, asset classes, fund managers, . . .
Internal Rate of Return no
Modified Dietz
yes
Time Weighted
yes
Since the IRR assumes a constant rate of return for all assets, calculating the
return of sub-portfolios is of no use.
Modified Dietz: Portfolio Component Returns
Component return:
Vi,E Vi,S Ci
ri =
P
Vi,S + tTi Ci,t TTt
Portfolio return:
r =
n
X
wi r i
i =1
Erich Janka
FH bfi
Page 81 of 133
Weights:
wi =
Vi,S +
Ci,t TTt
n
X
VS +
Ct TTt
i =1
wi = 1
Vi,E . . . end value of component i,

Vi,S . . . start value of component i,
Ci . . . total cash flow in in sector i,
Ci,t . . . cash flow in sector i at time t,
Ti . . . all cash flow times of i-th component.

wi . . . weight of the i-th component in the portfolio.
Component Returns: External Cash Flows

Now external Cash flows must be defined for each component. Since there
are cash flows between components, daily valuation will be necessary for time
weighted performance calculation.
Shifting funds between the components generates external cash flows for
sub-portfolios but not for the total portfolio.
Dividends appear in the asset class cash but must be treated as external
cash flows, while they increase the performance of the asset class shares.
Especially the asset class cash must be considered carefully. It might be useful
to define several accounts for some components otherwise it might be difficult
to do a proper valuation of the sub-portfolios.
Single-Period Component Returns
Portfolio mangers will change the asset allocation over time: money will be
shifted between sub-portfolios. The timing of these decisions will have an
impact on the portfolio return:
r1 =
rt =
n
X
i =1
n
X
i =1
wi1 ri1
wit rit
n
X
V 1 V 0+
i
i =1
n
X
i =1
V 0+
(t1)+
Vit Vi
V (t1)+
wit . . . weight of the i-th component in the portfolio for the t-th period and
rit . . . its return.
Vit . . . Value of the i-th component for the t-th period immediately before the external cash flow.
Vit + . . . Value of the i-th component for the t-th period immediately after the external cash flow.
V t . . . Value of the portfolio for the t-th period immediately before the external cash flow.
Erich Janka
FH bfi
Page 82 of 133
The second version of the formula is more general: it even holds when an initial
weight is 0. This could happen when derivatives are involved.
Multi-Period Component Contribution
The returns of the individual periods can be chain-lined to their return for the
total period:
m
Y
1 + ri =
(1 + rit )
t =1
Multi-Period Component Returns: Timing

Its not only the the performance of the sub-portfolios is important, the timing
when how much money is assigned to them is also crucial.
Multi-Period Component Returns: Example
Period
Weight (%)
Europe Asia
Return (%)
Europe Asia
Total return (%)
Q1
Q2
Q3
Q4
30
50
80
40
70
50
20
60
17.3
3.5
12.8
12.7
4.2
1.3
9.8
2.9
2.3
2.4
8.3
3.3
Year
50
50
19.3
3.5
0.8
Due to bad timing, the portfolio return is negative while both sectors have a
positive performance.
Elements of the Return
This example shows: Two factors are vital for a good performance:
the performance of the components and
the timing.
Stock Selection and Timing
These two elements of return result in two styles of active management:
Stock selection: The portfolio manager tries to find assets that will outperform
the benchmark and
Erich Janka
FH bfi
Page 83 of 133
Timing: The portfolio manager overweights sectors he thinks will outperform.

Of course these two styles can be combined.
Multi Period Contributions are not Additive
Now we are interested in the contribution of a component during a specific
period. Maybe its manager was on holiday while a colleague did his work. If the
overall returns were just a simple sum this would be easy: We choose the term
of the relevant component at the period of interest. But unfortunately this is not
so easy . . .
r = 1 +
r 6=
n
Y
(1 + r i )
i =1
n
m
XX
wit rit
i =1 t =1
Multi Period Contribution is not Additive: Example
r = 1 + (1 +
n
X
wi1 ri1 )(1 +
n
X
wi2 ri2 )
i =1
|i=1{z }
r1
n
X
wi1 ri1
+ (1 + r1 )
n
X
i =1
r 6=
n
X
i =1
wi2 ri2
i =1
wi1 ri1 +
n
X
wi2 ri2
i =1
Multi Period Contribution

Multi period contribution needs quite complicated formulas which can be found
using e. g. [Fis].
Base Currency and Local Returns
When the portfolio contains securities denominated in foreign currencies, the
impact of the currency return has to be considered as well:
Erich Janka
FH bfi
Page 84 of 133
Exchange Rate
$:e
Start value
End value
Return
Portfolio value
$
e
1.4
1.2
100
110
71.43
91.67
16.67 %
10.00 %
28.33 %
1 + r = (1 + r L )(1 + rC )
1.2833 = 1.1667 1.1
r L . . . Return in local currency,
rC . . . currency return
Local Return and Currency Return Dont Add

Compound portfolio return:
1 + r = (1 + r L )(1 + rC ) = 1 + r L + rC + r L rC
1.2833 = 1 + 0.1 + 0.1667 + 0.0167
Portfolio return in e:
0.1667} + 71.43
{z
0.01667}
20.24 = 71.43
|
{z
|
| {z 0.1} + 71.43
local return
currency return
combined impact
= 7.14 + 11.9 + 1.19

Portfolio Return in Foreign Currency
The base currency returns of a portfolio can be transformed to any other currency:
rC = (1 + r )(1 + c) 1
rC . . . Portfolio return expressed in currency C,
c . . . currency return of C relative to the base currency.
Local Return Portfolios

Local currency returns dont take the currency exchange rates into account.
Therefore local currency returns are not realbut they are useful for intermediate calculations. Local return of multi-currency portfolio:
rL =
n
X
wi r Li
i =1
wi . . . Weight of currency i,
r Li . . . local return of currency i.
Erich Janka
This is the weighted average local return.

FH bfi
17
17.1
Page 85 of 133
Options
Put Call Parity
Call + Riskfree Zero = Underlying + Put
17.2
Black Merton Scholes
Volatility of underlying is constant over time

Underlying can be traded continuously
Price of underlying is log-normally distributed
Short selling of the underlying is always possible
No transaction costs, fees or taxes
Any fraction of the underlying is tradable
Risk-free interest rate is constant; borrowing and lending is possible
Constant and continuous dividend yield from the underlying
17.3
Influence Factors
17.3.1
Volatility
Trading Volatility: Straddle, Strangle, Index Futures (VIX)

17.3.2
Implied Volatility Skew
[Blum],
[Bouz] Time series analysis indicated thatin contrast to BSvolatility
is not constant over time (view back in time).
And this also holds for the implied volatilities of options (view forward in time).
But in addition the implied volatility also depends on the given strike price.
The volatility skew refers to the implied vola difference between option with the
Erich Janka
FH bfi
Page 86 of 133
same asset
same maturity
different strike price
Also called half smile or smile. Especially for FX we have a symmetrical situation
and therefore usually smiles.
Explanations:
Imbalance between supply and demand: Common strategy: Hedge with
OTM puts and finance it with short calls.
Skewness of distribution of returns: Big down moves come often with
increased volatility, and therefore increasing the probability of another big
move.
Measuring skew in theory:
Slope =
dimp (K )
dK
Problem: only available for a distinct set of strikes. Therefore skew is measured
as the approximation derived from the points 0.9K and K (90 % and 100 % of the
strike price):
imp (0.9K ) imp (K )
Slope =
10 %
The slope usually is negative; skews are positive (change sign).
Trading skew: Long bull call spread: Long skew, therefore more expensive Long
bear put spread: Short skew, therefore cheaper
17.3.3
Greeks
17.4
American Options
American options can be exercised at any time. The holder has more freedom in
comparison to European options, therefore they must have a higher price:
C A (t, K, T ) CE (t, K, T )
Erich Janka
and
PA (t, K, T ) PE (t, K, T )
FH bfi
Page 87 of 133
Figure 5: Delta changing with time.
Condition when to exercise an American option:

D
> r ( T t)
K
t . . . time of valuation
K . . . Strike price
T . . . Maturity
D . . . Future dividend value at T
r . . . interest rate
17.5
Asian Options
[Bouz, p. 43f] Asian options are European style with an average Savg instead of a
single fixed strike and have the following payoff structure:
Asian Callpayoff = max(0, Savg K )
Asian Putpayoff = max(0, K Savg )
The average reduces the influence of the volatility and therefore Asian options
have a lower price than European options.
Erich Janka
FH bfi
Page 88 of 133
Figure 6: Gamma changing with time.
The more observations the lower the price

averaging in: Averaging periods are uniformly distributed during the options
life time. (less risky)
averaging out: The dated for taking the average are at the end of the options
life time (e. g. quarterly)
17.6
Digital Options
Cash or Nothingpayoff
Cash or Nothingpayoff = C 1{ST > H }

(
C if ST > H
=
0 textotherwise
C . . . coupon
H . . . barrier level.
Erich Janka
FH bfi
Page 89 of 133
Figure 7: Rho changing with time.
Pricing: closed Black-Scholes formula.

Vanilla Callprice = SeqT N (d1 ) KerT N (d2 )
{z
} |
{z
}
|
Asset-or-nothing
Cash-or-nothing
q . . . dividend yield
r . . . risk-free interest rate
N ( ) . . . normal cdf
K . . . Strike = coupon (C) of digital
Digital options can be hedged using bull spreads.

Chart: Delta:
flatmountainflat
(composed of S and Z (shifted to the right))
Chart: Vega:
Flatmountainwider valleyflat
(composed of long mountain and a short mountain (shifted to the right))
Erich Janka
FH bfi
Page 90 of 133
Figure 8: Theta changing with time.
17.7
Quanto Options
[Bouz] Quantos are assets denominatd in a currency other than the one in
which it is normally traded.
They enable FX-returns in the base currency.
European Quanto:
Quanto Callpayoff = FX0 max(ST K, 0)
FX0 . . . exchange rate at time 0
There is a closed Black-Scholes formula for quanto European option: Just add
the following term to the dividend yield:
S,FX S FX
. . . corrlation between underlying equity and FX rate
S . . . vola of equity
FX . . . vola of FX rate
Erich Janka
FH bfi
Page 91 of 133
Figure 9: Vega changing with time.
18
Payoff Diagrams
[Blum]
One way to represent complex or structured products is the payoff
diagram. It represents the profit/loss by the price of the underlying at maturity.
Example 13
Capital guarantee with cap
It is not always possible to draw a payoff diagram. This can be the case when the
product references to underlyings with various maturities or is path dependent.
If there is a payoff diagram, the price of the product is determined: If two
products share the payoff diagram, it is not possible to distinguish between them
(in theory).
19
Payoff Formula
[Blum]
The payoff formula is an exact mathematical expression describing how
to calculate the precise value of the product at maturity.
Erich Janka
FH bfi
Page 92 of 133
Example 14 (Capital guarantee with cap)

payoff = max K1 , min(S, K2 )
K1 S < K1
payoff = S
K1 S < K2
K2 K2 S
Example 15 (Capital protected note on an index)

IndexT Indext0
Payoff = Nominal 95 % + 90 % max 0,
Indext0
There is no general rule where to start interpreting payoff formulas: The Nominal
is multiplied by a more complex term which are interpreted separately: The
95 % give the level of capital guaranteed (right now we dont know that it is a
guarantee because this depends on the other terms). Combined with the nominal
this is a zero coupon bond paying 95 % of the nominal at maturity.

IndexT Indext0
max 0,
Indext0
Is either 0 or the simple performance of the index. If the index value decreases,
the value is 0, otherwise its performance is used. This is just a long at the money
call option (the strike is the initial value of the index).
This maximum is multiplied by 0.9 and therefore the product participates only
in part on the positive index returns. The slope of the line will be less than 1, the
angle is less than 45 degree.
The result is a product which can loose at most 5 % of its initial value and rises
with the positive return of the underlying index.
Example 16

5 400 3 600
Payoff = 250 000 95 % + 90 % max 0,
3 600

= 250 000 95 % + 90 % max(0, 0.5)

= 250 000(|{z}
0.95 + |{z}
0.45 ) = 250 000 1.4
fix
Erich Janka
variable
FH bfi
Page 93 of 133
= 350 000

IndexT Indext0
Payoff = Nominal G + P max 0,
Indext0

G . . . Level of guarantee (in %)

P . . . Level of participation
T . . . Time at maturity
t0 . . . Time at inception.
How much must the index rise in order to break even?

1G
P
5%
= 5.56 %
=
0.9
Rbreak-even =
How much must the index rise in order to get a return of r = 3 % (which might
be the risk free interest rate)?
1+rG
P
8%
=
= 8.89 %
0.9
Rr =
R3 %
Example 17
Another investor places 150 000 in a product with G = 102 % and P = 85 %.

In this case the return of the product is always positive but the level of
participation is distinctly lower.
The initial value of the index is 1205 and after four years (maturity) it has
decreased to 980.

980 1 250
Payoff = 150 000 102 % + 85 % max 0,
1 250

= 150 000 102 % + 85 % max(0, 0.216)
= 150 000(|{z}
1.02 + |{z}
0.0 ) = 150 000 1.02
fix
variable
= 153 000
Erich Janka
FH bfi
Page 94 of 133
Regardless how low the index is after four years, the return p. a. is approximately 0.5 %.
20
Structured Products
The Swiss Structured Products Association ([SVSP]) classifies structured products:

1. Capital guarantee
2. Yield Enhancement
3. Participation
4. Leverage
These types are ordered by increasing risk/expected return.
20.1
Capital Protection Certificate with Participation
Capital protection
Example : Plan:
Maturity 4 years
equity exposure
no capital at risk
Constraints:
Interest rate: 3 %
Funding spread: 40bp p. a.
The bond price is 87.48 %
Product Costs: 52bp
Available for option: 100 % 87.48 % 0.52 % = 12 %
Erich Janka
FH bfi
Page 95 of 133
A 4 year vanilla ATM option on EuroStoxx50 gives an option price of 15 % which is more
than the 12 % we have.
One solution is by reduction of the participation:
Participation =
12
= 0.8
15
Therefore when the index rises by e. g. 10 %, the structure would only rise by 10 % 0.8 =
8%
Notepayoff

Index( T ) Index(0)
= 100 % + 0.8 0 %,
Index(0)
If a participation of less than 100 % is not desired, it is possible to use an (averaging out)
Asian option instead. The option can be chosen with as many quarterly observation
points at the end necessary in order to get the options price to 12 %.
Another solution could be to use an European option that is enough out of the money to
reduce its price to 12 %
Maybe the investor is willing to participate in losses as well then he could use short puts
in order to get 3 % premium.
Last but not least it is possible to lower the guarantee level.
It is also possible to limit the upside using a short call
If the interest rates are higher than in our example, a participation greater then
100 % could also be possible.
[SVSP]:
Market expectation:
Rising underlying
Rising volatility
Sharply falling underlying possible
Erich Janka
FH bfi
Page 96 of 133
Characteristics:
Minimum redemption at expiry equivalent to the capital protection
Capital protection is defined as the percentage of the nominal (e. g. 100 %).
Capital protection refers to the nominal only, and not to the purchase price.
Value of the product may fall below its capital protection during lifetime.
Participation in underlying price increase above the strike.
Any payouts attributable to the underlying are used in favor of the strategy.
The structure can be represented as a combination of a zero and long at the
money calls (the strike equals the value of the underlying at inception). The
expected value of the coupons determines how many options can be bought and
therefore the level of participation in the rising underlying.
20.2
Convertible Certificate
Capital protection
Market expectation:
Sharply rising underlying
Rising volatility
Characteristics:
Erich Janka
FH bfi
Page 97 of 133
Capital protection is defined as a percentage of the nominal (e. g. 100 %)

Capital protection refers to the nominal only, and not to the purchase price
Value of the product may fall below its capital protection during the lifetime
Participation in underlying price increase above the strike (conversion
price)
Any payouts attributable to the underlying are used in favor of the strategy
Coupon payment possible
The structure can be represented as a combination of a zero and long out of
the money calls (the strike price of the options is above the initial value of the
underlying).
Since out of the money options are cheaper than at the money options, the
payment of coupons is realistic. It would be also possible that the guaranteed
value lies above 100 % or the participation in the positive returns is greater than
1.
20.3
Barrier Capital Protection Certificate
Capital protection
Market expectation:
Rising underlying
Characteristics:
Erich Janka
FH bfi
Page 98 of 133

Participation in underlying price increase above the strike up to the barrier
Possibility of rebate payment once barrier is breached
Limited profit potential
A possible decomposition:
Zero coupon bonds
Long knock out call (strike at the money, knock out level out of the money)
Long binary call (its strike equals knock out level)
Since the investor has limited upside participation, a rebate is realistic. The level
of the rebate depends on the expected coupon and the knock out level.
20.4
Capital Protection Certificate with Coupon
Capital protection
Market expectation:
Rising underlying
Characteristics:
Erich Janka
FH bfi
Page 99 of 133

Value of the product may fall below its capital protection during the lifetime
The coupon amount is dependent on the development of the underlying
Periodic coupon payment is expected
Limited profit potential
The payment of the coupons depends on the value of the underlying at the
coupon date. This product can be structured using zero coupon bonds and a
series of binary options on the underlying with varying maturities.
20.5
Discount Certificate
Yield enhancement
Market expectation:
Underlying moving sideways or slightly rising
Falling volatility
Characteristics:
Should the underlying close below the strike on expiry, the underlying
and/or a cash amount is redeemed
Erich Janka
FH bfi
Page 100 of 133
Discount Certificates enable investors to acquire the underlying at a lower

price
Corresponds to a buy-write-strategy
Smaller risk of loss than with direct investment in the underlying
Higher discounts can be achieved at greater risk if the product is based on
multiple underlyings (multi-asset)
Replication: Direct investment in the underlying or a low exercise price option
(LEPO) and a short ATM call.
20.6
Barrier Discount Certificate
Yield enhancement
Market expectation:
Falling volatility
Underlying will not breach barrier during product lifetime
Characteristics:
The maximum redemption amount (Cap) is paid out if the barrier is never
breached
Barrier Discount Certificates enable investors to acquire the underlying at
a lower price
Erich Janka
FH bfi
Page 101 of 133
Since, provided the barrier has not been breached, the nominal is repaid on
expiry, the probability of maximum repayment is higher but the discount
is smaller
If the barrier is breached the product changes into a Discount Certificate
Larger discounts or a lower barrier can be achieved at greater risk if the
product is based on multiple underlyings (multi-asset)
Limited profit potential (Cap)
Structure:
Zero Coupon bonds
Short at the money put options (on 100 % of the nominal)
Long knock out at the money put options (on 200 % of the nominal)
The long put options reduce the return in bull markets.
20.7
Reverse Convertible
Yield enhancement
Market expectation:
Falling volatility
Erich Janka
FH bfi
Page 102 of 133
Characteristics:
Should the underlying close below the strike on expiry, the underlying
and/or a cash amount is redeemed
Should the underlying close above the Strike at expiry, the nominal plus
the coupon is paid at redemption
The coupon is paid regardless of the underlying development
Larger coupons can be achieved at a greater risk if the product is based on
multiple underlyings (multi-asset)
Replication: Zero coupon bonds and a short at the money put.
Reverse convertibles and discount certificates share the same payoff diagram
and cannot be distinguished in theory (put-call parity).
In fact there can be a different because of taxes. [Blum]

describes the situation in
Switzerland where the authorities tax income but not capital gains. Therefore
the coupon of the reverse convertible might be taxed.
Reverse convertible
Discount certificate
SMI
1 year
10 %
3%
1.05 %
8.95 %
SMI
1 year and 1 day
10 %
3%
0%
10 %
Underlying
Maturity
Coupon/Discount
Interest rate
Tax (35 %)
Net revenue
20.8
Barrier Reverse Convertible
Yield enhancement
Erich Janka
FH bfi
Page 103 of 133
Market expectation:
Falling volatility
Characteristics:
Should the barrier never be breached, the nominal plus coupon is paid at
redemption
Since, provided the barrier has not been breached, the nominal is repaid on
expiry, the probability of maximum repayment is higher but the coupon is
smaller
If the barrier is breached the product changes into a Reverse Convertible
The coupon is paid regardless of the underlying development
Larger coupon payments or lower barriers can be achieved at a greater risk
if the product is based on multiple underlyings (multi-asset)
Structure:
Zero Coupon bonds
Short at the money put options (on 100 % of the nominal)
Long knock out at the money put options (on 200 % of the nominal)
Erich Janka
FH bfi
20.9
Page 104 of 133
Express Certificate
[SVSP] Yield enhancement
Market expectation:
Characteristics:
Should the underlying trade above the Strike on the observation date, an
early redemption consisting of nominal plus an additional coupon amount
is paid
Offers the possibility of an early redemption combined with an attractive
yield opportunity
Smaller risk of loss than with direct investment in the underlying, because
the nominal is paid on redemption provided the barrier was not breached
Larger coupon payments or lower barriers can be achieved at a greater risk
There is no simple recombination of this structure and auto-call features have to
be used.
Example 18 (Express certificate (July 2008) [Blum][p.

126)
]
Underlying asset
Erich Janka
SMI index
FH bfi
Page 105 of 133
Issue price
100 %
Spot as issue
6 670
Maturity
3 years
Observation dates
Annual
Early redemption condition SMI closes above the initial spot price at an observation date
Barrier
55 %
Cumulative Coupon
15 %
20.10
Tracker Certificate
Participation [SVSP]
Market expectation:
Rising underlying
Characteristics:
Participation in development of the underlying
Reflects underlying price moves 1:1 (adjusted by conversion ratio and any
related fees)
Risk comparable to direct investment in the underlying
Fees generally in the form of management fees or through the retention of
payouts attributable to the underlying during the lifetime of the product
This type of security is useful when a direct investment is not possible or inconvenient. Possible underlyings are indices or commodities.
Replication with zero bonds and futures/forwards.
Erich Janka
FH bfi
20.11
Page 106 of 133
Out-performance certificate
Rising underlying
Rising volatility
(Decreasing payouts (e. g. dividends))
Characteristics:
Disproportionate participation (outperformance) in positive performance
above the strike
Reflects underlying price moves 1:1 when below the Strike
Risk comparable to direct investment in the underlying
Replication:
Zero coupon bonds
Forwards/futures
Long at the money calls (as much as possible using the expected coupons)
Erich Janka
FH bfi
20.12
Page 107 of 133
Bonus certificate
Market expectation:
Underlying moving sideways or rising
Characteristics:
Minimum redemption is equal to the nominal provided the barrier has not
been breached
If the barrier is breached the product changes into a Tracker Certificate
Larger Bonus payments or lower barriers can be achieved at a greater risk
Replication:
zero coupon bonds
futures
Long at the money put options (referring to 100 %)
Erich Janka
FH bfi
20.13
Page 108 of 133
Bonus outperformance certificate
Market expectation:
Rising underlying
Characteristics:
Disproportionate participation (outperformance) in positive performance
above the strike
been breached
If the barrier is breached the product changes into a Outperformance
Certificate
A higher bonus payment or lower barrier can be achieved at greater risk if
the product is based on multiple underlyings (multi-asset)
Replication:
Zero coupon bonds
Futures
Long at the money call options
Erich Janka
FH bfi
20.14
Page 109 of 133
TwinWin certificate
Market expectation:
Rising or slightly falling underlying
Characteristics:
Profits possible with rising and falling underlying
Falling underlying price converts into profit up to the barrier
been breached
If the barrier is breached the product changes into a Tracker Certificate
Replication:
Zero coupon bonds
Futures
Erich Janka
FH bfi
20.15
Page 110 of 133
Reference Entity Certificate with Conditional Capital Protection
Products with reference entities [SVSP]
Market expectation:
Rising underlying
No credit event of the reference entity
Characteristics:
There are one or more reference entities underlying the product
In addition to the credit risk of the issuer, redemption is subject to the
solvency (non-occurrence of a credit event) of the reference entity
Redemption is made at least in the amount of conditional capital protection at maturity, provided that no credit event of the reference entity has
occurred
If a credit event occurs at the reference entity during the life time, the
product will be redeemed at an amount corresponding to the credit event
The product value can fall below conditional capital protection during its
lifetime, among other things due to a negative assessment of reference
issuer creditworthiness
Conditional capital protection only applies to the nominal and not the
purchase price
Participation in development of the underlying, provided a reference entity
credit event has not occurred
The product allows higher yield at greater risk
Erich Janka
FH bfi
20.16
Page 111 of 133
Reference Entity Certificate with Yield Enhancement
Market expectation:
Falling volatility of the underlying
Characteristics:
In addition to credit risk, redemption of the product is subject to the solvency (non-occurrence of a credit event) of the reference entity
The product value can fall during its lifetime, among other things due to a
negative assessment of reference entity creditworthiness
If the underlying is lower than the exercise price upon maturity, the underlying is delivered and/or a cash settlement is made, provided that no
credit event of the reference entity has occurred
If the underlying is higher than the exercise price upon maturity, the
nominal is repaid, provided that no credit event of the reference entity
has occurred
Depending on the characteristics of the product, either a coupon or a
discount to the underlying can apply
Erich Janka
FH bfi
Page 112 of 133
A coupon is paid out regardless of performance of the underlying, provided

that no credit event of the reference entity has occurred
In addition, the product can feature a barrier
More than one underlying (multi-asset) allows higher coupons, larger
discounts, or lower barriers for greater risk
Limited Profit Potential (Cap)
20.17
Reference Entity Certificate with Participation
Market expectation:
Rising underlying
Characteristics:
In addition to credit risk, redemption of the product is subject to the solvency (non-occurrence of a credit event) of the reference entity
The product value can fall during its lifetime, among other things due to a
negative assessment of reference entity creditworthiness
Erich Janka
FH bfi
Page 113 of 133
Participation in development of the underlying, provided a reference entity

credit event has not occurred
In addition, the product can feature a barrier
21
Structured Products and Barrier Options
Option
Type
Location
Crossed
Not crossed
Call
Up&Out
Above spot Worthless
Up&In
Above spot Plain vanilla call
Down&Out Below spot Worthless
Down&In
Below spot Plain vanilla call
Put
Up&Out
Above spot Worthless
Plain vanilla put
Up&In
Above spot Plain vanilla put Worthless
Down&Out Below spot Worthless
Plain vanilla put
Down&In
Below spot Plain vanilla put Worthless
Plain vanilla call

Worthless
Plain vanilla call
Worthless
Barrier options are usually cheaper then plain vanillas because they have
to cross the barrier to get some value or become worthless when doing so.
For in options:
the longer the maturity, the more its price is similar to a plain vanilla
the shorter the distance to the barrier, the more the price is similar to
a plain vanilla
For out options:
The longer the maturity, the cheaper the option (tending to zero)
The nearer the distance to the barrier, the lower its price (tending to
zero)
According to [Blum,
p. 123] the vast majority of structured products have Asianstyle barriers because many investors just dont pay attention and focus on other
factors of the product, such as a high coupon or a better participation.
Erich Janka
FH bfi

Product
Page 114 of 133

factor
Shark note
Barrier
Rebate
Barrier reverse convertible Coupon
on EuroStoxx50 index
Barrier
Bonus certificate
Bonus
on EuroStoxx50 index
Barrier
Am. barrier (original)
Eur. barrier
131.5 %
7.5 %
10.4 %
75.0 %
9.0 %
65.0 %
123.0 %
7.5 %
8.6 %
75.0 %
2.5 %
65.0 %
[Bouz] Draw chart: Delta of an up-and-out call: Flatwide mountainnarrower

valleyvery small hillflat
Draw chart: Vega of an up-and-out call: Flatsmall mountaindeep valleytiny
hillflat
22
Option Strategies
[CME2]
22.1
Long Risk Reversal
(Squash, Combos)
Erich Janka
FH bfi
22.2
Page 115 of 133
Short Risk Reversal
(Squash, Combos)
22.3
Bull Spread
Erich Janka
FH bfi
22.4
Bear Spread
22.5
Long Butterfly
Erich Janka
Page 116 of 133
FH bfi
22.6
Short Butterfly
22.7
Long Condor
22.8
Short Condor
22.9
Long Straddle
Erich Janka
Page 117 of 133
FH bfi
22.10
Short Straddle
22.11
Long Strangle
Erich Janka
Page 118 of 133
FH bfi
22.12
Short Strangle
22.13
Ratio Call Spread
Erich Janka
Page 119 of 133
FH bfi
22.14
Ratio Put Spread
22.15
Call Ratio Backspread
Erich Janka
Page 120 of 133
FH bfi
22.16
Put Ratio Backspread
22.17
Calender Spread
Page 121 of 133
[Bouz, p. 99]
23
Securities Lending
[Mall]
Securities lending
temporary transfer of securities from the lender to the borrower
usually on a collateralized basis
the securities must be returned either on demand or at the and of any
agreed term.
The securities are returned with some fees and also the returns (dividends).
It can eventually
improve market liquidity
Erich Janka
FH bfi
Page 122 of 133
lead to more efficient settlement

give tighter dealer prices
reduce the cost of capital.
Motivation for lender:
Fees
Lower depot costs
The lender would be subject to withholding tax on dividends, the borrower
not.
Motivation for the borrower:
Covering a short position
Convertible bond arbitrage: buying a convertible bond and selling the
underlying equity short.
Pairs trading: Buying a security and shorten another one with similar
characteristics.
Merger arbitrage: selling short the equities of a company making a
takeover bid and buying the potential acquisition company.
Long index futures vs. shortening some of its constituent securities.
Increase in share capital: A company plans to issue new stocks but doesnt
want the exact timing being public then it can borrow its own old shares.
24
Repurchase Agreements (REPOs)
[FT2] A repo is a sell and buy back agreement: A security is sold and at the same
time a buy at a later date to a specified price is fixed. The counterpart enters a
reverse-repo.
One party transfers the security and the other cash. All the risks and benefits of
the security stay with the seller.
The party providing the cash the security can be considered as collateral and
therefore the charged interests can be lower. The only risk it faces is when both
the security and the counterpart defaults during the lifetime of the repo.
Erich Janka
FH bfi
Page 123 of 133
Example 19 (Repo)
Bank A makes the following repo:
Security: 10 million Bund
Actual quote: 101.5 + 2.55 = 104.05 (dirty price)
Repo-yield: 3 %
Maturity: 30 days
Initial cash transaction:
10 000 000
104.05
= 10 405 000
100
Cash transaction at maturity:

30
10 405 000 1 + 0.03
= 10 431 012
360
25
25.1
Basel II, Basel III

Basel II
[OeNB2] Banks havelike any industryface specific business risks. They have
a important position in the economy as intermediates since they act as lenders
and borrowers. They are crucial for funding and the costs of financing.
Giving loans bears the risk of credit risk: the borrower fails to his obligations.
Banks are aware of this risk and charge different rates depending on the estimated risk of their debtors.
Due to competition, banks tend to reduce their rates or even take risks they should
not. On the other hand the authorities want to limit the capital ratios. In order to
get an as level playing field as possible there is demand for an internationally
coordinated regulatory for capital requirements.
Erich Janka
FH bfi
Page 124 of 133
Pillar 1: Minimum Capital capital requirements

Pillar 2: Supervisory review process
Pillar 3: Market discipline
Pillar 1 arises from
credit risk,
market risk and
operational risk.
Each of these risks has to be calculated using an approach which is
suitable and
sufficient
for the individual bank.
For each risk category two methods are possible (combinations are possible):
standardized approach
internal method
The standard approach gives quite high risk values which is an incentive to
continue development of risk management in order to get better values.
Erich Janka
FH bfi

25.1.1
Page 125 of 133
Credit Risk
Besides the standard approach, internal rating-bases (IRB) can be used. For the
standard approach, banks are required to hold a minimum capital standard of
8 % of the volume of risk-weighted assets (outstanding loans). Depending on
the quality of the credit exposure the weights are 0 %, 20 %, 50 %, 100 % or even
150 %.
Under the IRB approaches, the capital requirements are calculated with respect
to the five asses classes:
sovereigns,
banks,
corporates,
retail customers,
equity.
The in-hose rating class is determined via the
probability of default (PD) with a one-year time horizon
loss given default (LGD) as a percentage of the
outstanding claim at the time of default (exposure at default, EAD)
25.1.2
Market and Operational Risk
Operational risks are hard to capture and the methods are not as well developed
as for the other risks whose foci are much narrower. They are growing with the
increasing complexity of the business processes. According to the OeNB, banks
set aside about one fifth of their economic capital for operational risk. Most
major losses were at least in part due to operational risks.
The market risk is considered to be quite straightforward.
The second pillar (supervisory review process) requires the authorities to carry
out quantitative reviews in order to
assess the quality of the methods a bank uses,
Erich Janka
FH bfi
Page 126 of 133
check if the capital resources are adequate

monitor the constant improvements
The third pillar (market discipline) are mainly rules for the publication of information such as the banks risk profile or the adequacy of capital items.
25.2
Basel III
25.2.1
Pillar 1
Banks must have a greater amount of common equity: Raised to 4.5 % of risk
weighted assets (after deductions).
A capital conservation buffer is introduced: Additional 2.5 % risk weighted assets
comprising common equity.
A countercyclical buffer is introduced: Comprising common equity of 02.5 % of
risk weighted assets. The value is set by the authorities.
Introduction of a stressed VaR for the trading book. Also liquidity must be taken
into account.
A non-risk-based leverage ratio that includes off-balance sheet exposures will
serve as backstop to the risk-based capital requirement.
25.2.2
Liquidity
Introduction of a liquidity coverage ratio (LCR) in order to have sufficient high

quality liquid assets in order to withstand a 30-day stressed funding scenario
specified by the supervisors.
The net stable funding ratio (NFRS) is a longer-term structural ratio designed to
address liquidity mismatches. It covers the entire balance sheet giving the banks
incentives to use stable sources of funding.
The supervisors will monitor the liquidity ratios in order to identify risk trends
at both the bank and system-wide level.
25.2.3
Systemically Important Financial Institutions (SIFIs)
SIFIs must have higher loss absorbency capacity because they impose a higher
risk to the financial system.
Erich Janka
FH bfi

25.2.4
Page 127 of 133
Risk Adjusted Return on Risk Adjusted Capital (RARORAC)
[Brun] Consider a bank that wants a target rating of AA. For any given
portfolio there is capital requirement of equity capital, the economic capital (EC).
EC = VaR99.9 % Expcected Loss (EL)
Monte Carlo Simulation is usually used to get the loss distribution.
Incremental capital and marginal capital:
ECi = EC( P + Mx ) EC( P)
EC( P + Mx )
ECm = x
x
P . . . reference/basis credit portfolio
If x is very small with respect to the overall

portfolio value, the difference between vanishes.
Mx . . . marginal transaction with nominal amount x.
Return on Capital:
ROC =
Revenues
Allocated Capital
Risk Adjusted Return on Capital:

RARORAC =
Revenues EL
ECm
Example 20 (RARORAC)
Contract 1 (low credit risk): 100 000 000 USD, with 20 bp fees; EL1 = 1 bp;
EC1m = 30 bp
Contract 2 (high credit risk): 100 000 000 USD, with 100 bp fees; EL1 = 20 bp;
EC1m = 500 bp
The ROC of Contract 2 is higher but for the RARORAC holds:
RARORAC1 =
Erich Janka
20 1
63 %
30
RARORAC1 =
100 20
16 %
500
FH bfi
26
Page 128 of 133
Monte-Carlo-Simulation
Sources of errors and how to deal with them:
Applications for Monte-Carlo-simulations:

Integrals (Areas)
Optimization (genetic algorithms)
Product pricing
Risk calculation (e. g. VaR)
26.1
Generate Correlated Random Variates
In order to generate correlated normal random numbers, follow these three steps:
1. Generate uniformly distributed variates
2. Use them to generate uncorrelated normal variates
3. Induce correlation into them
Erich Janka
FH bfi

26.1.1
Page 129 of 133
Uniform Random Numbers
Uniformly distributed [0, 1]-variates can be generated via complementary multiply

with carry (CMWC).
26.1.2
Normally Distributed Numbers

Method
Sum of 12
Numerical inversion
Box-Muller
Polar-Method
Rejection sampling (Ziggurat)
quality
speed
complexity
+
+
+
++
+
+
Example 21 (Polar method)

1. u1 , u2 . . . uniformly distributed [0, 1]
2. q = u21 + u22
3. if q = 0 or q > 1, go back to 1.
q
2 ln(q)
4. p =
q
5. x1 = u1 p; x2 = u2 p
x1 and x2 are now two iid N (0, 1) numbers
x N (0, 1) x + N (, )
26.1.3
Correlated Random Numbers
The Cholesky-decomposition of the correlation matrix can be used to induce

correlation:
C = LL T
Xc = LX
Erich Janka
FH bfi
Page 130 of 133
C . . . Correlation matrix (n n)
L . . . Triangular matrix (n n)
X . . . Uncorrelated random numbers (n m)
Xc . . . Correlated random numbers (n m)
Example 22 (Two correlated normal numbers)

1. A, B . . . uncorrelated normal (N (0, 1)).
p
2. Ac = A, Bc = A + 1 2 B.
3. Transform using and
An alternative method uses the spectral-decomposition:

C = EE T = E E T = SS T
Xc = SX
26.2
Product Pricing
1. Simulate the values of the underlying factors at all dated necessary to

calculate the pay off
2. Calculate the pay off
3. repeat 1. and 2. many times
4. Calculate the average pay off
5. Discount it in order to get the present value
26.3
Value at Risk
. . . level of significance
1. Price the PF using the current values of the risk factors
2. Simulate changes in the risk factors
3. Price the PF with these simulated risk factors
Erich Janka
FH bfi
Page 131 of 133
4. Repeat 2. and 3. many times

5. Calculate the P/L for all simulations and sort them by size
6. Drop (1 ) of the smallest values
7. Multiply the smallest remaining value with (1); this ist the VaR().
8. Scale to desired time horizon.
Erich Janka
FH bfi
Page 132 of 133
References
[Acwo] Will Acworth (2010): Annual Volume Survey 2010; Internet; Futures
Industry Magazine.
[Bier] Bernd Biermann (2002): Die Mathematik von Zinsinstrumenten; 2. Auf
lage; Munchen;
Oldenbourg.
[BIS]
No Author (n. d.): Basel Committee on Banking Supervision reforms Basel III; Internet (accessed Nov. 27th 2011); Bank for International Settlements.
[Blum]
Andreas Blumke
(2009): How to Invest in Structured Products; Chichester; Wiley.
[Bouz] Mohamd Bouzoubaa, Adel Osseiran (2010): Exotic Options and Hybrids;
Chichester; Wiley.
[Brun] Vivien Brunel (n. d.): Risk Adjusted Return On Risk Adjusted Capital
(RARORAC); Internet (accessed Dec. 14th 2011); n. p..
[Bud] Alfred Buder et al. (2011): Instrumente des Zins-, Wahrungs- und
Rohstoffmanagements; Wien; Osterreichische

Volksbanken AG.
[DB]
No Author (2008): The Global Derivatives Market, An Introduction;
Frankfurt; Deutsche Borse

AG.
[Chri] David Christie (3. 11. 2003): Accrued Interests & Yield Calculations and
Determination of Holiday Calenders; n. p.; SWX Swiss Exchange.
[CME] No Author (n. d.): CME Group; Internet; CME Group.
[CME2] No Author (2011): 25 Proven strategies for trading options on CME
Group futures; Internet (accessed Nov. engordnumber27); CME Group
[EX]
No Author (n. d.): Eurex; Internet; Eurex.
[EX1] No Author (n. d.): Deliverable Bonds and Conversion Factors; Internet
(accessed Aug. 1st 2011); Eurex.
[Fab] Frank Fabozzi (editor) 2002: The Handbook of Finacial Instruments; New
Jersey; Wiley.
[FT2] No Author(n. d.): FOREXBASICS; N. p.; Finance Trainer.
Erich Janka
FH bfi
Page 133 of 133
[FT]
No Author (2010): Kapitalmarkt-Futures; N. p.; Finance Trainer.
[Fis]
Bernd Fischer 2001: Performanceanalyse in der Praxis, 2. Auflage;
Munchen;
Oldenbourg Verlag.
[FTC] No Author (2007): Die seltsame Welt der Zinsderivate; Magazin fur
technisches Trading, 9. Jahrgang; Wien, FTC Capital GmbH.
[Glass] Paul Glasserman (2004): The PRM Handbook;
[Hagan] Patrick S. Hagan, Graeme West (2008): Methods for Constructing a
Yield; Wilmott Magazine. Curve;
[Mall] Mallesons Stephen Jaques (2005): An Introduction to Securities Lending
(Australia); Sydney; Spitalfields Advisors.
[Nevt] Salih Neftci (2000): An Introduction to the Mathematics of Financial
Derivatives, Second Edition; San Diego; Academic Press.
[OeNB] Gunther
Thonabaur (ed) (2004): Financial Instruments, Structured Products Handbook; Wien; Oesterrichische Nationalbank;
[OeNB2] No Author (n. d.): Basel II Basics; Internet (accessed Nov.27th 2011;
Oesterrichische Nationalbank.
[Para] No Author (n. d.): Day count conventions; Internet (accessed Jun. 27th
2011); Paranzasoft.
[SVSP] No Author (n. d.): http://www.svsp-verband.ch.
[Tuck] Bruce Tuckman, Pedro Porfirio (2003): Interest Rate Parity, Money Market
Basis Swaps, and Corss-Currency Basis Swaps; Lehman Brothers.
[Wilm] Paul Wilmott (2009): Frequently Asked Questions in Quantitative Finance, Second Edition; Chichester; Wiley.
[Wost]
Christoph Woster
(2005): Die Ermittlung des Conversion Factors im
Futures-Handel (Diskussionspapier Nr. 543); Bielefeld; Universitat Bielefeld.
Erich Janka
FH bfi

Product Management

Uploaded by

Copyright:

Available Formats

Product Management

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Product Management

Uploaded by

Copyright:

Available Formats

Product Management 2011