FR 8 Metrics

Risk Management in Shipping
Modeling, Measuring, & Managing Freight Market Uncertainty
Presented at:
National Technical University of Athens

School of Naval Architecture & Marine Engineering
29 May 2003
Presentation Outline
 Introduction
 About this presentation
 About FreightMetrics
 About Risk Management

 Defining risk
 The risk management process
 Scope of risk management
 Modern applications of risk management
 Measuring Market Risk

 The traditional approach to market risk measurement
 The Value-at-Risk (VaR) approach
www.freightmetrics.com 2
Presentation Outline
 Measuring Market Risk in Shipping

 Justification for risk management in shipping
 Market risk measurement vs. market forecasting
 Identifying the impact of freight market risk on fleet cash flow
 Developing a framework for measuring freight market risk
 Measuring Market Risk in Shipping Using the Fr8MetricsTM Methodology

 Main methodological features
 How does Fr8MetricsTM work?
 Benefits of the Fr8MetricsTM methodology
 Potential users and managerial applications
 Software implementation
 Managing Freight Market Risk

 Altering the risk profile using managerial decisions
 Altering the risk profile using freight derivatives
Introduction
About This Presentation
 Our objective
 Shipping is a business activity exposed to a wide variety of risks.
 In this presentation we are concerned with the measurement of one particular form of risk –
namely freight market risk, or the risk of loss arising from unexpected changes in freight rates.
 Our motivation
 Risk management is a notion that exists in financial markets for decades, having experienced
significant technological and modeling advances over the years.
 Shipping has proved rather slow in adopting modern risk management techniques and best
practices from other industries.
 Our motivation is to present a modern framework for measuring freight market risk, using the
paradigm of other market-sensitive industries.
Introduction
About FreightMetrics
 What FreightMetrics is…

 A provider of consulting services and software solutions for measuring and managing freight
market risk.
 Working closely with Shipping Banks, Shipowners, and Freight Traders, in order to quantify their
exposure to freight market risk in terms of cash-flow sensitivity.
 Our approach lies in transferring best practices and modern methodologies from the area of
financial risk management to shipping.
 What FreightMetrics is NOT…

 Shipbroker.
 Forecasting agency.
 Market news vendor.
 Financial intermediary.
For more information about FreightMetrics, visit our website at www.freightmetrics.com
About Risk Management
Defining Risk
 Definition of Risk
 We define (financial) risk as the prospect of financial loss due to unforeseen changes in
underlying “risk factors”. These risk factors are the key drivers affecting portfolio value and
financial results. Such risk factors are equity prices, interest rates, exchange rates, commodity
prices, freight rates, etc.
 Types of Risks
 Business: The risk of loss due to unforeseen changes in demand, technology,
competition, etc., affecting the fundamentals of a business activity.
 Market: The risk of loss arising from unexpected changes in market prices
or market rates.
 Credit: The risk of loss arising from the failure of a counterparty to make a
promised payment.
 Operational: The risk of loss arising from the failures of internal systems or the
people who operate in them.
 Other types: Legal, Liquidity, etc.
The Risk Management Process
 The Risk Management process

 There is a wide misconception amongst practitioners, especially within the shipping industry, who
consider risk management as synonymous to hedging. This is an oversimplification and does not
reflect the true dimension of risk management.
 In fact, risk management is a process that involves three separate steps:
1. Risk Modeling: Before any attempt to take decisions on risk considerations, we must identify the
underlying risk factors, understand their behavior, and try to model their dynamics. This is the basic
foundation on which the other phases of the risk management cycle are built.
2. Risk Measurement: After identifying and modeling the underlying risk factors, we must
determine their significance and quantify their influence on portfolio value and financial results.
3. Risk Management : Having identified and measured our risks, we are then able to take informed
decisions on whether to reduce our exposure or alter our risk profile based on our risk preferences –
hedging is one such alternative course of action.
Scope of Risk Management
 Risk Management ≠ Hedging

 As already mentioned, risk management is not synonymous to hedging.
Hedging is just one alternative for the active management of risk.
 Moreover, risk management does not necessarily imply risk reduction. In fact, the objective of
risk management is NOT to reduce risk, but – more importantly –
to quantify and control risk.
 Most of the times, the objective is not to eliminate risk, but rather to alter our risk profile
according to the prevailing market conditions, our risk preferences, and potential regulatory or
contractual requirements.
 Risks are embedded in any business activity. For a shipowner, the decision to invest in a vessel
signifies his belief that freight rates will go up, earning him a return on his investment that is
higher than the “risk-free” interest rate. However, there is no “free lunch” in the economy; his
decision to invest creates at the same time a natural exposure to freight rates, accepting the risk
that freight rates may in fact go down. Risks are simply unavoidable in any profit-taking activity.
Scope of Risk Management
 Uncertainty vs. Variability 1

“Variability is a phenomenon in the physical world to be measured, analysed and where
appropriate explained. By contrast, uncertainty is an aspect of knowledge.”
Sir David Cox
 Risk management is only useful for the mere fact that we cannot predict the future. There are two
components of our inability to be able to precisely predict what the future holds: these are
variability and uncertainty.
 Variability is the effect of chance and is a function of the system. It is not reducible through either
study or further measurement, but may be reduced through changing the physical system.
 Uncertainty is the assessor’s lack of knowledge (level of ignorance) about the parameters that
characterize the physical system that is being modeled. It is sometimes reducible through further
study, or through consulting more experts.
 Risk management can do very little to reduce variability (markets will continue to fluctuate no
matter how advanced risk management gets), but can be very effective in reducing uncertainty for
those involved in risk-taking decisions.
1 Adapted from the book Risk Analysis by David Vose, Chapter 2: “Quantitative Risk Analysis, Uncertainty and Variability”
Modern Applications of Risk Management
 Modern applications of Risk Management

 Exposure measurement and reporting
 Market risk (since early 90s)
 Credit risk (since late 90s)
 Operational risk (new area)
 Economic capital estimation
 Allocation of capital
 Risk-based pricing
 Risk limits
 Risk-adjusted performance evaluation
Modern Applications of Risk Management
 Example: Risk-Adjusted Performance Evaluation

 Consider two traders who are evaluated on the basis of their realized profits at some future date.
Trader B ended up with higher profits compared to Trader A. Does this mean he is more skilled
than Trader A? Does he deserve a higher bonus? What about the risk incurred by each trader
through their trading strategy?
Measuring Market Risk
The Traditional Approach to Risk Measurement
 The Mean-Variance framework

 Under the Mean-Variance framework, we model financial risk in terms of the mean and variance
(or standard deviation, the square root of variance) of the Profit/Loss (P&L) or the returns of our
portfolio.
 The Mean-Variance framework often makes the assumption that returns obey a normal
distribution (strictly speaking, the mean-variance framework does not require normality, but it is
easier to understand its statistics).
 Portfolio Theory
 The origin of portfolio theory can be traced back to the work of Markowitz (1952) which earned
him the Nobel prize.
 Portfolio theory starts with the premise that investors choose between portfolios on the basis of
maximizing expected return for any given portfolio standard deviation or minimizing standard
deviation for any given expected return.
 One of the key insights of portfolio theory is that the risk of any individual asset is measured by
the extent to which that asset contributes to overall portfolio risk which depends on the correlation
of its return with the returns to the other assets in the portfolio (a result known as diversification
effect).
 Portfolio theory typically makes the assumption of normally distributed returns.
The Value-at-Risk (VaR) Approach
 The origin and development of VaR

 In the late 70s and 80s, a number of major financial institutions started working on internal
models to measure and aggregate risks across the institution as a whole.
 The best known of these models is the RiskMetrics model developed by JP Morgan. According to
industry legend, this model is said to have originated when the chairman of JP Morgan, Dennis
Weatherstone, asked his staff to give him a daily one-page report – the famous “4:15 report” –
indicating risk and potential losses over the next 24 hours, across the bank’s entire trading
portfolio.
 The report was ready by around 1990 and the measure used was Value-at-Risk (VaR), or the
maximum likely loss over the next trading day. VaR was estimated from a system based on
standard portfolio theory, using estimates of the standard deviations and correlations between the
returns of different traded instruments.
 In early 1994, JP Morgan set up the RiskMetrics unit to make its data and basic methodology
available to outside parties. This bold move attracted a lot of attention and raised awareness of
VaR techniques and risk management systems.
 The subsequent adoption of VaR systems was very rapid, first among securities houses and
investment banks, and then among commercial banks, other financial institutions and non-
financial corporates.
 Today, VaR is widely used in almost every market-sensitive industry (with the exception perhaps
of shipping!) and has even gained recognition from regulatory authorities.
 VaR in practice
 VaR Basics:
 VaR on a portfolio is the maximum loss we might expect over a given holding or horizon
period, at a given level of confidence (probability).
 VaR is less restrictive on the choice of the distribution of returns and the focus is on the
tail of that distribution – the worst p percent of outcomes.
 VaR in practice
 Estimating VaR: The various methodologies for estimating VaR actually differ on their
particular technique for constructing the distribution of possible portfolio values from which VaR
is inferred. The most common methodologies are:
 Analytical methods (Variance/Covariance)
 Historical simulation
 Monte-Carlo simulation
 Attractions of VaR:
 VaR is a single, summary, statistical measure of possible portfolio losses, providing a
common and consistent measure of risk across different positions and risk factors.
 It takes account of the correlations between different risk factors.
 It is fairly straightforward to understand, even for non-technical people.
 VaR variants: Following the same logic, other “at risk” measures have been proposed to
quantify risk in various settings: Cash Flow at Risk (CaR), Earnings at Risk (EaR), etc.
Measuring Market Risk in Shipping
Justification for Risk Management in Shipping
 Business and market risk in shipping: the two faces of the same coin
 Most industries can distinguish between business risks and market risks. These industries have to
worry about business risks and try to hedge away market risks which may have an adverse side-
effect on financial results. For example, an auto manufacturer has to worry about business risks
such as technology, competition, production, R&D, but may also have an exposure to FX risk,
which may hamper exports, or interest rate exposure which may increase debt service on floating
rate obligations.
 Other industries cannot distinguish between business risks and market risks. The most pronounced
example is maybe that of financial institutions. A significant part of the business of financial
institutions is to take direct exposure in the world’s equity, interest rate, currency, and commodity
markets.
 Shipping can be said to belong to the industries that cannot distinguish between business risks and
market risks. Financial results in shipping are directly affected by movements in the world’s
freight rate markets.
Shipowners are in effect in the business of managing shipping risk affecting a portfolio of
physical assets, rather than simply managing a fleet of vessels.
 High market volatility

 Freight rates have historically been very volatile. The impact of unforeseen geo-political events
and the slow speed of adjusting supply to demand have often resulted in dramatic fluctuations in
the level of freight rates.
Example of Freight Market Fluctuations BFI
(Baltic Freight Index, BCI Route 2, BPI Route 1) BCIr2
Standardised scaling (base=100 @1/3/99) BPIr1
300
250
200
150
100
50
0
1/4/1985 1/4/1987 1/4/1989 1/4/1991 1/4/1993 1/4/1995 1/4/1997 1/4/1999 1/4/2001 1/4/2003
 Industry inefficiencies
 Capital needs vs. sources of funds:
 Shipping is a capital intensive industry with significant funding needs for fleet expansion
and replacement purposes. Yet, it has very limited opportunities to diversify its sources
of funding, as most of its financing comes in the form of bank debt.
 Asset – Liability (mis)matching

 Asset economic life >> term of debt financing
 Variable (uncertain) revenues to meet fixed debt claims
 Pro-cyclical lending practices

 Many banks tend to be influenced by the general sentiment of the market and ignore the
cyclical nature of the business. Thus, they appear more willing to lend when the market
(and vessel prices) is high, despite the fact that the market will eventually revert back to
lower levels. In contrast, they appear rather hesitant to extend credit at a period of low
freight rates, although they are likely to rise to more sustainable levels.
 Lessons from the High Yield disaster of the late 90s

 In the late 90s, many shipping companies decided to tap the public debt markets with high-yield
bonds. Historically, it was the first massive attempt of the industry to diversify away from bank
debt, using an alternative source of external funding.
 Unfortunately, nearly all of these high yield issues subsequently defaulted, mainly due to
insufficient cash flow generation.
 This indicated poor risk assessment and a lack of appropriate tools to evaluate shipping market
risks.
 Example:
Source: Moody’s (2002)
 Lessons from the High Yield disaster of the late 90s

 Examples:
Source: Moody’s (2002)
Market Risk Measurement vs. Market Forecasting
 Types of maritime forecasting

 Structural econometric models: Model freight rates as a dependent variable, driven by a number
of independent variables, usually representing macro-economic factors that influence shipping
demand, e.g. GDP growth, oil prices, industrial output, etc.
 Time series models: Model freight rates using the structure (serial correlation) in the past history
of the data itself. Future freight rates are determined based on lagged values of their own history
and do not exploit or infer causality with other economic variables.
 Difficulties in maritime forecasting

 Demand for shipping is characterized as derived
demand, meaning that it depends on the demand of
the commodities that are shipped by sea.
 Econometric models are prone to specification
problems. Model fit can always improve by
including more explanatory variables, which may
introduce multicollinearity problems.
 It is possible that a small error in demand
estimation may lead to a gross mis-estimation of
freight rates (compare D1→D2 vs. D2→D3).
Market Risk Measurement vs. Market Forecasting
 Differences in scope
 Scope: Prepare for future vs. Predict the future
 Motivation: Prevent unexpected losses vs. Make a profit (beat the market)
 Horizon: Long-term vs. Short-term
 Emphasis: Tail of the distribution vs. Mean of the distribution
 Differences in methodology
 Measurement does not presuppose causality relations between economic variables.
 Forecasting models have potentially infinite specifications (depending on choice of explanatory
variables).
 Measurement focuses on producing the complete picture of potential outcomes (entire
distribution) rather than producing a point estimate (the mean of the distribution).
So, do we discard forecasting? NO, it can serve a useful complementary role, especially in
revealing causality relations between economic variables. Forecasting may also assist in certain
chartering or trading decisions in the short-run, where it is most effective.
Impact of Freight Rate Volatility on Cash Flow
 Identifying the impact of freight rate volatility on fleet cash flow

 Fluctuations in freight rates directly affect fleet cash flow.
 Cash flow performance is the topmost concern in shipping.
 Ship financing belongs to the family of “Project Financing” (other forms of project financing
include airlines, infrastructure, real estate, etc.). There is a principle in project financing:
repayment MUST come from the operating cash flows of the financed asset.
 So, what really matters in measuring freight market risk is the impact of freight rate variability on
cash flow performance.
 Case study
 We performed a simple exercise (historical simulation) of the cash flow generation of a handysize
dry bulker over two separate time periods (period A: Jan-1980 to Dec-1986, and period B: Jan-
1987 to Dec-1993).
 We used the same set of assumptions in both cases (see the following slide), except for using the
actual freight rates for each period and the actual second-hand value of the financed vessel at the
start of the each period.
 This exercise not only illustrates the impact of freight rate volatility on cash flow, but also
emphasizes the impact of shipping cycles and the importance of proper timing in maritime
decision-making.
 Case study: impact of freight rate volatility on fleet cash flow

 Historical simulation assumptions and actual freight rate data (source: Clarksons)
1-yr time charter rate for Handysize (30000 dwt)

Period A: Jan80-Dec86 Period B: Jan87-Dec93
12000
PERIOD
ASSUMPTIONS A (80-86) B (87-93)
Operating Rev. days 25 25
10000
Opex days 30 30
Opex ($/day) 3,500 3,500
Loan Vessel value 10,000,000 3,800,000 8000
Loan amount 7,000,000 2,660,000
Advance ratio 70% 70%
($/day)
Installments 84 84 6000
Balloon 30% 30%
Loan interest 5% 5%
Start. Liquidity Cash balance 0 0 4000
2000
Jan-88
Jan-89
Jan-95
Jan-96
Jan-76
Jan-77
Jan-78
Jan-79
Jan-80
Jan-81
Jan-82
Jan-83
Jan-84
Jan-85
Jan-86
Jan-87
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-97
Jan-98
Jan-99

 Chart of monthly net cash flow (after debt service, but excluding balloon payment):
Impact of freight rate volatility on fleet cash flow A (80-86)

B (87-93)
150,000
100,000
Monthly Net Cash Flow ($)
50,000
0
4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84
-50,000
-100,000
-150,000
Month

 Chart of accumulated liquidity (excluding balloon payment):
Impact of freight rate volatility on fleet cash flow A (80-86)

B (87-93)
5,000,000
4,000,000
3,000,000
Accumulated Liquidity ($)
2,000,000
1,000,000
0
4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84
-1,000,000
-2,000,000
-3,000,000
-4,000,000
-5,000,000
Month
A Framework for Measuring Freight Market Risk
 Basic Assumptions and Objectives

 It is possible to develop a framework for measuring freight market risk using the VaR
paradigm.
 The risk factors in this framework consist of freight rates.
 Freight rates are assumed to follow a random walk and are modeled using appropriate
stochastic processes.
 The stochastic processes that describe the evolution of freight rates are able to replicate certain
known characteristics of freight rate dynamics (cyclicality, seasonality, random shocks).
 Monte-Carlo simulation is used to generate future freight rate scenarios, in accordance with the
underlying stochastic process for each risk factor.
 The key measure of risk is fleet cash flow.
 For each freight rate scenario, we re-compute future cash flow, using an appropriate cash flow
model which takes into account debt repayment and other cost items (e.g. drydocking costs,
special surveys, etc.).
 Thus we construct the entire distribution of future cash flow, from which we can make VaR-
type inferences based on a specified confidence level.
 Modeling the stochastic behavior of freight rates

 Any time series data can be thought of as being generated by a particular stochastic or random
process, the “true” data-generating process (DGP). A concrete set of data, such as a historical
data-series on a freight rate, can be regarded as a particular realization (i.e. a sample) of the
underlying true DGP. The distinction between the stochastic process and its realization is akin to
the distinction between population and sample in cross-sectional data. Just as we use sample data
to draw inferences about a population, in time series we use the realization to draw inferences
about the underlying stochastic process.
 Selecting a stochastic process

 The modeling of any price variable begins with the choice of a particular stochastic process which
captures the characteristics of asset price dynamics. In order to make this selection, we are guided
by theoretical considerations, such as the theory concerning the operation of freight equilibrating
mechanisms, as well as by empirical analysis of historical data (e.g. mean reversion, fat tails,
autocorrelation, volatility clustering, etc.)
 There is a large number of alternative stochastic processes that can be tested to capture the
dynamics of freight rates. Below we provide a few indicative (simplistic) models:
 Geometric Brownian Motion (GBM): dr (t )  r (t )dt  r (t )dz (t )
 Ornstein-Uhlenbeck (O-U) process: dr (t )  a (   r (t ))dt  dz (t )
 Jump-Diffusion (O-U with Jumps): dr (t )  a(   r (t ))dt  dz (t )  Jd (h)
 Estimating model parameters

 As discussed previously, a concrete set of historical data can be regarded as a particular
realization (i.e. a sample) of the underlying “true” DGP. The objective is to find a theoretical DGP
that provides the best fit for the actual data. This is accomplished by estimating the parameters of
each theoretical DGP and comparing the various models in terms of some measure of goodness-
of-fit.
 Many stochastic processes admit exact discretization or numerical approximations (using the
Euler or higher-order methods) which allow the testing of the underlying processes using standard
econometric techniques. Both the GBM and O-U processes admit such discretizations which lead
to time-series specifications of a linear autoregressive form.
 For example, the GBM model can be estimated by running the following regression:
r (t )  r (t  1)  c   (t )
 The parameters of an O-U process can be estimated using discrete-time data by running the
regression: r (t )  r (t  1)  c  br (t  1)   (t )
log(1  bˆ)
and then calculating   cˆ / bˆ a   log(1  bˆ)   ˆ 
(1  bˆ) 2  1
 More advanced models require other techniques, e.g. Maximum Likelihood Estimation
 Simulating the stochastic evolution of freight rates

 Having discretized the stochastic process and estimated its parameters, we proceed with iterative
sampling from the probability distribution(s) used in our model, in order to generate future freight
rate scenarios.
 This technique is known as Monte-Carlo simulation and includes several steps:
 Random number generation (pseudo-random or low-discrepancy numbers)
 Transformation of independent random number into correlated random numbers
 Variance reduction methods (to improve the accuracy for a given number of runs)
 Example: As discussed in the previous slide, the O-U process can be interpreted as the
continuous-time version of a first-order autoregressive process in discrete time. Specifically, the
O-U process is the limiting case as Δt  0 of the following AR(1) process:
r (t )  r (t  1)   (1  e  a )  (e  a  1) r (t  1)   (t )
where ε(t) is normally distributed with mean zero and standard deviation σε
 Thus, we can simulate an O-U process, by drawing random numbers from a normal distribution
with mean zero and standard deviation σε and generating r(t) as follows:
r (t )   (1  e  a )  e  a r (t  1)   (t )
 Building the distribution of future fleet cash flows

 For each freight rate scenario produced by our simulation, we re-compute the fleet cash flow
based on some cash flow model and plot the results in a histogram. This represents the
distribution of future fleet cash flows.
 Making risk inferences from the distribution of future fleet cash flows
 The distribution of cash flow results reveals the risk profile of the fleet, in terms of the range of
possible cash flows that the fleet is able to generate in the future.
 We can “read” this distribution in order to make probabilistic inferences about the risk of our
fleet. For example:
 What is the probability that the fleet will breakeven?
 What is the maximum possible cash deficit at the 95% probability level?
 A practical example
 Assuming a simple O-U process, we modeled the 1-year Time-Charter rate for dry bulk handysize
vessels and simulated 1000 different scenarios. Below we compare the distribution of actual
monthly returns (282 observations –from Feb-76 to Jul-99) with the distribution of simulated
(random) monthly returns. From this we can compute the cash flow of the vessel and produce the
distribution of possible cash flows for next month.
90
Empirical
Simulated
80
70
60
50
40
30
20
10
0
-19% -16% -14% -11% -9% -6% -4% -1% 1% 4% 6% 9% 11% 14% 16% 19%
Measuring Market Risk with Fr8MetricsTM
Main Methodological Features
 Main methodological features

 Fr8Metrics™ is a framework for quantifying freight market risk in shipping portfolios.
Fr8Metrics™ is generally based on the Value-at-Risk (VaR) concept, but differs in the use of
proprietary stochastic models developed to simulate the evolution of freight rates. These models
are designed to replicate the unique seasonal and cyclical characteristics of shipping markets.
 The Fr8Metrics™ methodology is simulation-based rather than forecast-based. It draws on
advanced Monte-Carlo simulation techniques to generate future freight market scenarios from
which we estimate the likely distribution of various financial measures, such as cash flow,
accumulated liquidity, hull cover, NPV, etc.
 Fr8Metrics™ is able to incorporate the influence of correlations, not only across different market
segments within the shipping industry, but also between shipping and financial markets. Thus, it
is possible to capture potential diversification effects within a portfolio that combines both
shipping and financial assets.
 Fr8Metrics™ is able to support portfolios that combine both physical assets (vessels) and “paper”
assets (derivatives).
How Does Fr8MetricsTM Work?
 Step 1: Portfolio Definition

 Input details for the fleet (charter agreements, cost data, etc.), loans (repayment schedules, interest
cost, collateral vessels, etc.), and derivatives.
 Step 2: Risk Mapping

 Assign risk factors to vessels and derivatives.
 Step 3: Project Definition

 Select the portfolio(s) which will be simulated.
 Specify cash flow model.
 Specify risk metric.
 Specify simulation parameters (number of scenarios, horizon, confidence level, etc.)
 Step 4: Scenario Generation

 Generate n (=number of scenarios specified in Step 3) future realizations for each risk factor for
the time horizon specified in Step 3, in accordance with the underlying stochastic process of each
risk factor:
 Step 5: Metric Computation

 Iteratively substitute values from each of the n scenarios from Step 4 into the cash flow model
specified in Step 3, calculate the n future cash flow results, and plot them in a histogram:
 Step 6: Risk Inference

 Using the distribution of cash flow results from Step 5, find the cash flow estimate corresponding
to the desired confidence level specified in Step 3.
 Having exposed the complete risk profile of the portfolio(s) specified in Step 3, the user (banker,
shipowner, etc.) is able to take calculated, risk-informed decisions in accordance with his risk
preferences:
Benefits of the Fr8MetricsTM Methodology
 General benefits of Monte-Carlo based methods

 Flexibility to support a wide range of stochastic processes.
 Not restrictive in terms of distributional assumptions.
 Ability to incorporate correlations among risk factors.
 Ability to incorporate decision rules along the simulated paths (e.g. exercise of charter options).
 Particular benefits of the Fr8MetricsTM methodology

 Utilizes stochastic models specifically developed to capture freight rate dynamics.
 Reveals diversification effects across shipping assets, as well as between shipping and financial
exposures.
 Provides a framework for monitoring derivatives, developing hedging strategies and assessing
hedge effectiveness.
Potential Users and Managerial Applications
 Shipping Banks
 Determining credit terms: maximum advance ratio, liquidity covenant, loan spread
 Risk assessment: repayment risk, probability of covenant breach
 Estimating default probabilities, verifying internal risk ratings
 Promoting cross-selling, derivatives sales, hedge proposals
 Shipowners
 Investment decisions, e.g. dry bulk vs. tanker segments
 Chartering decisions, e.g. time charter vs. spot employment
 Financing decisions, e.g. high-yield bond vs. bank debt
 Hedging decisions, e.g. derivatives vs. long-term charter
 Freight Traders
 Risk assessment and monitoring
 Risk-adjusted performance evaluation
Software Implementation
 Product features
 Hierarchical portfolios
 Multi-currency environment
 Periodic updates of parameter estimates for underlying stochastic models
 3 cash flow model formats (Fleet, GAAP, Sources and Uses)
 User-defined cash flow items
 Generation of pro-forma cash flow statements
 Technology
 Windows-based
 Developed in .NET environment
 Extensive use of XML
 Databases: SQL / Access
Managing Freight Market Risk
Altering the Risk Profile with Managerial Decisions
 Risk-informed decision-making
 As mentioned previously, the objective of risk management is not necessarily to eliminate risk,
but rather to alter our risk profile according to the prevailing market conditions, our risk
preferences, and potential regulatory or contractual requirements.
 Having exposed the complete risk profile of a shipping portfolio within a VaR framework, we are
able to decide whether it suits our risk preferences or to make comparisons among alternative
business strategies.
Choice of strategy is subject to risk preference Strategy A is dominant

(Strategy B: higher expected return, but higher risk) (Strategy A: higher expected return AND lower risk)
Altering the Risk Profile with Managerial Decisions
 Asset Allocation decisions

 Expand the fleet in the dry cargo or the tanker segment?
 Buy one VLCC or two Aframaxes?
 Buy one 5-year old vessel or two 15-year old vessels?
 Order a newbuilding in Korea (cheaper, but FX risk) or in the US?
 Chartering decisions
 Trade in the spot market or lock in a 3-year time charter at a rate that is –currently- lower than
the spot rate?
 Accept a high time charter rate or a lower time charter rate with an option to renew?
 Charter-in or charter-out for the next one year?
 Funding decisions
 Finance new acquisitions through bank debt or high yield issue?
 Go for a 5-year loan with low spread or a 7-year loan with higher spread?
Altering the Risk Profile with Freight Derivatives
 Definition of derivatives
 In chemistry, a derivative is a “substance related structurally to another substance and
theoretically derivable from it (...) a substance that can be made from another substance”. 1
Derivatives in finance work on the same principle. They are financial instruments whose promised
payoffs are derived from the value of something else, generally called the underlying.
1 This definition comes from the online version of the Merriam-Webster Collegiate Dictionary. See
http://www.britannica.com/cgi-bin/dict?va=derivative
 Types of freight derivatives

 Forward Freight Agreements (FFAs): An agreement between two counterparties to settle a
freight rate for a specified quantity of cargo or type of vessel, for a certain route, and at a certain
date in the future.
 The underlying asset of the FFA contracts can be any of the routes that constitute the indices
produced by the Baltic Exchange.
 FFAs are settled in cash on the difference between the contract price and an appropriate settlement
price at expiration.
 To establish an FFA, we need to specify: route, price, duration/quantity, settlement
 Other types of derivatives: Options, Swaps, Swaptions, etc.
 The market of freight derivatives

 Historical development of freight derivatives:
 Freight derivatives existed since 1985 with the creation of the BFI (Baltic Freight Index),
a basket of individual dry cargo routes. This index served as a settlement mechanism for
freight futures listed on BIFFEX (subsequently merged with LIFFE, contracts de-listed
in April 2002).
 Since 1992, the individual shipping routes could be traded “over the counter” (i.e. not
through an exchange) in the form of FFAs.
 Current market size and players:
 Estimated annual turnover: $4.0 billion in notional value of freight.
 Types of players: Shipowners, Charterers, Trading Houses, Shipbrokers
 The role of the Baltic Exchange:
 Sets the rules and oversees the process of collecting and processing the brokers’
assessments of freight rates in more than 30 cargo routes. These prices are used for the
settlement of FFA transactions.
 Fundamentals of freight derivatives trading

 Trading process
 Price discovery through brokers
 FFA negotiation
 Counterparty clearance
 Documentation
 Basis risk (sources: correlation, time lag)
 Marking-to-Market
 Designing a hedging program

1. Understand the distribution / dynamics of freight rates.
2. Estimate the impact of adverse freight rate movements on fleet cash flow.
3. Decide whether to hedge, depending on external and internal considerations.
4. Choose the appropriate financial instruments.
5. Determine how much to hedge.
References, Links, and Further Reading
 References
 Adland, Roar (June 2000), Theoretical Vessel Valuation and Asset Play in Bulk Shipping, Thesis submitted for the MS in Ocean
Systems Management, MIT
 Attikouris, Kyriakos (April 2000), Modeling Freight Rates, Thesis submitted for the Diploma in Mathematical Finance,
University of Oxford
 Attikouris, Kyriakos (March 1996), Time Series Applications in the Ocean Shipping Business, Project submitted for the course
Applied Time Series Analysis (MBA program), University of Rochester
 Concalves, Franklin de Oliveira (September 1992), Optimal Chartering and Investment Policies for Bulk Shipping, Thesis
submitted for the PhD in Ocean Systems Management, MIT
 Dowd, Kevin (2002), Measuring Market Risk, Wiley
 Drewry Shipping Consultants (1997), Shipping Futures and Derivatives, Briefing Report
 Moody’s Investor Services (2002), Default & Recovery Rates of European Corporate Bond Issuers, 1985-2001
 Stopford, Martin (1997), Maritime Economics, Routledge
 Vose, David (2000), Risk Analysis, Wiley
 Wilmott, Paul (1998), Derivatives, Wiley
 Magazines: Risk, Marine Money, Lloyd’s Shipping Economist
 Seminar notes: Freight Derivatives seminar, organized by the Cambridge Academy of Transport and the Baltic Exchange (25
November 2002)
 Links
 www.riskmetrics.com RiskMetrics Group
 www.gloriamundi.org GloriaMundi (the best internet source on VaR material)
 www.balticexchange.com The Baltic Exchange

FR 8 Metrics

Uploaded by

Document Information

Original Title

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

FR 8 Metrics

Uploaded by

Copyright:

Available Formats

Risk Management in Shipping

Modeling, Measuring, & Managing Freight Market Uncertainty

National Technical University of Athens

 About Risk Management

 Measuring Market Risk

 Measuring Market Risk in Shipping

 Measuring Market Risk in Shipping Using the Fr8MetricsTM Methodology

 Managing Freight Market Risk

 What FreightMetrics is…

 What FreightMetrics is NOT…

For more information about FreightMetrics, visit our website at www.freightmetrics.com

 The Risk Management process

 Risk Management ≠ Hedging

 Uncertainty vs. Variability 1

 Modern applications of Risk Management

 Example: Risk-Adjusted Performance Evaluation

 The Mean-Variance framework

 The origin and development of VaR

 High market volatility

 Asset – Liability (mis)matching

 Pro-cyclical lending practices

 Lessons from the High Yield disaster of the late 90s

Source: Moody’s (2002)

 Lessons from the High Yield disaster of the late 90s

Source: Moody’s (2002)

 Types of maritime forecasting

 Difficulties in maritime forecasting

 Identifying the impact of freight rate volatility on fleet cash flow

 Case study: impact of freight rate volatility on fleet cash flow

1-yr time charter rate for Handysize (30000 dwt)

 Case study: impact of freight rate volatility on fleet cash flow

Impact of freight rate volatility on fleet cash flow A (80-86)

 Case study: impact of freight rate volatility on fleet cash flow

Impact of freight rate volatility on fleet cash flow A (80-86)

 Basic Assumptions and Objectives

 Modeling the stochastic behavior of freight rates

 Selecting a stochastic process

 Geometric Brownian Motion (GBM): dr (t )  r (t )dt  r (t )dz (t )

 Ornstein-Uhlenbeck (O-U) process: dr (t )  a (   r (t ))dt  dz (t )

 Jump-Diffusion (O-U with Jumps): dr (t )  a(   r (t ))dt  dz (t )  Jd (h)

 Estimating model parameters

 Simulating the stochastic evolution of freight rates

 Building the distribution of future fleet cash flows

 Main methodological features

 Step 1: Portfolio Definition

 Step 2: Risk Mapping

 Step 3: Project Definition

 Step 4: Scenario Generation

 Step 5: Metric Computation

 Step 6: Risk Inference

 General benefits of Monte-Carlo based methods

 Particular benefits of the Fr8MetricsTM methodology

Choice of strategy is subject to risk preference Strategy A is dominant

 Asset Allocation decisions

 Types of freight derivatives

 The market of freight derivatives

 Fundamentals of freight derivatives trading

 Designing a hedging program

References, Links, and Further Reading

You might also like