Download as pdf or txt
Download as pdf or txt
You are on page 1of 43

Section 1: Energy Commodities Basics

Glen Swindle
20 March 2012
c Glen Swindle: All rights reserved
1 / 43
Outline
What makes energy commodities dierent?
Pricing and Delivery
Forward Yields
Macro Perspective
Hedging and Common Structures
Themes
2 / 43
What makes energy commodities dierent?
High Volatility
Realized volatility for a price series p(n) is dened as:
_
250

n
R
2
(n)
_1
2
where R(n) = log
_
p(n)
p(n1)
_
For commodities we have used the rst traded contract (dened shortly)
GT10 volatility was proxied by the product of duration and change in yield.
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Historical Vols


SPX
EUR
GC1
GT10
WTI
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0
0.2
0.4
0.6
0.8
1
Historical Vols Normalized by WTI Vol


SPX
EUR
GC1
GT10
3 / 43
What makes energy commodities dierent?
High Volatility
Does this really matter?
Options markets exist for energy commodities that can be used to hedge
vol exposures.
Returns statistics for NYMEX crude oil (WTI) and natural gas
(NG):
First contract and the rolling 12 month (cal) strip
1 and 10 day intervals
Statistic WTI 1st Month WTI 1Y NG 1st Month NG 1Y
Std Dev (1 Day) 0.024 0.020 0.035 0.022
p
1,99
(1 Day) 0.067 0.053 0.093 0.059
Std Dev (10 Day) 0.073 0.059 0.107 0.069
p
1,99
(10 Day) 0.228 0.172 0.272 0.180
Table: Returns Statistics (2000-2010)
4 / 43
What makes energy commodities dierent?
High Volatility: Deal Pricing
Suppose you are purchasing an energy asset, for example:
A set of oil or natural gas production elds.
An ecient power generation asset.
In these (and many other cases) the value is approximately
linear in the price of the underlying commodity.
During the course of two weeks one can expect:
A 6-7% change in value
p
1,99
of nearly 20%
For the acquirer paying say a billion dollars, as well as for the
lenders supporting such an activity, a 20% change in value
would be highly problematic.
Simply converging on an acquisition price with such
underlying volatility can be challenging.
5 / 43
What makes energy commodities dierent?
High Volatility: Colateral
Whether exchange traded or OTC, hedging activities are
usually accompanied by colateral posting requirements.
High volatility in the underlying requires signicant availability
of cash and/or letters of credit (LCs) .
High volatility amplies mismatches in colateral posting terms.
Example: Retail energy companies
Provide commodities to retail end-users (who typically are not margined)
Hedge this native short position via standard futures or OTC swaps
markets (which are margined)
This mismatch in credit support can result in lethal colateral calls in
highly volatile times.
6 / 43
What makes energy commodities dierent?
High Volatility: Colateral
The following plots shows the rolling cal strip for NYMEX WTI, NG
and PJM power prices.
PJM is a power market in the eastern U.S. and the largest power market
in North America
The colateral calls against entities with long energy hedges put on
in mid-2008 were onerous.
2007 2008 2009 2010 2011
0
50
100
150
$
/
B
a
r
r
e
l
WTI
2007 2008 2009 2010 2011
0
5
10
15
$
/
M
M
B
t
u
NYMEX NG
2007 2008 2009 2010 2011
0
50
100
150
$
/
M
W
h
PJM Peak
Figure: Rolling Calendar Strips
7 / 43
What makes energy commodities dierent?
Specialness
Another relevant feature is that energy commodities are
always going special.
Special is a term borrowed from bond markets in which
particular bonds that are cheapest to deliver (CTD) into a
futures contract trade at a premium due to limited supply
0 5 10 15 20 25
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
Duration(Y)
Y
i
e
l
d
Yield Versus Duration: 15Jan2009
10Y Futures CTD
Long Bond Futures CTD
8 / 43
What makes energy commodities dierent?
Specialness
For commodities supply/demand variations are far more
extreme than in other markets resulting in breakdowns of
typical relationshipsi.e. specialness.
Specialness in commodities is either temporal or locational.
A single commodity delivered at two dierent times or
locations can behave functionally as two entirely
dierent assets.
9 / 43
What makes energy commodities dierent?
Temporal Specialness
The following gure shows a sample NYMEX NG forward
curve.
The periodicity (and non-monotone) prices are due to
seasonal variations in demand.
The winter months trade special to the summer months.
This increases the dimensionality of risk management;
Limited eectiveness using front month contracts to hedge winter
exposures.
2010 2011 2012 2013 2014 2015 2016
5.5
6
6.5
7
7.5
8
F
o
r
w
a
r
d

P
r
i
c
e

(
$
/
M
M
B
t
u
)
NYMEX NG Forward curve: 25Jan2010
10 / 43
What makes energy commodities dierent?
Temporal Specialness
Spot price behavior shows short time-scale specialness.
The term spot price refers to the price for delivery of the commodity for
delivery now.
The following shows daily spot prices and spot returns for
Henry Hub natural gas.
Henry Hub is the location underpinning the NYMEX NG.
Note that daily returns in excess of 500% are not uncommon.
2000 2002 2004 2006 2008 2010
0
5
10
15
20
$
/
M
M
B
t
u
Henry Hub Spot Prices
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
3
2
1
0
1
2
3
x 10
4
%
Annualized Returns (%)
11 / 43
What makes energy commodities dierent?
Locational Specialness
The gure shows historical daily spot prices for Henry Hub
and for TETM3, a delivery location in the northeast.
The middle gure shows the spot basis price, which is the
dierence between TETM3 and the benchmark Henry Hub
prices.
The bottom gure is the price ratio.
While TETM3 is typically premium to Henry Hub due to
transportation costs, of particular note are the substantial
premia in spot prices that can arise due to high demand and
low supply on occasional days in the winter.
12 / 43
What makes energy commodities dierent?
Locational Specialness
1998 2000 2002 2004 2006 2008 2010
0
20
40
60
$
/
M
M
B
t
u
Spot Prices


Henry Hub
TETM3
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
0
10
20
30
$
/
M
M
B
t
u
Spot Basis
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
0
2
4
6
T
E
T
M
3

/

H
H
Spot Ratio
13 / 43
What makes energy commodities dierent?
Specialness
These phenomena described apply at even shorter time scales
in power: daily, hourly and even sub-hourly.
A commodity for delivery at a particular time and location can
exhibit dramatically dierent price dynamics from the same
commodity deliverable at a dierent time and location, even
when the times and locations are seemingly close.
The set of tradables (swaps and options) available to a
portfolio manager is often far smaller than the number of
ways that a commodities portfolio can go special.
A major theme of this course is to elaborate on gap between
risks that can be hedged and risks that are often embedded in
commodities businesses.
14 / 43
Pricing and Delivery
Basic Terms
The pricing of a commodities trade requires specication of:
The underlying commodity
When and where will the commodity delivered/referenced
The price to be paid for the commodity
The notional quantity
The mechanics of delivery/settlement
Credit / Margining:
OTC contracts: specic margining provisions in ISDAs or related energy
specic credit docs.
Futures: Exchange traded with daily margining / mark-to-market.
For futures settlement has occurred implicitly through daily
margining.
15 / 43
Pricing and Delivery
Examples
WTI Crude CME/NYMEX (Futures)
Notional 1000 barrels delivered anytime in the contract month
Specic grades of crude (adjusted for value) delivered at Cushing, OK
Natural Gas CME/NYMEX (Futures)
Notional 10,000 MMBtus delivered ratably over the contract month
Delivery location: Henry Hub, LA
Gas Daily Swap (OTC)
Buyer will pay seller $5.20 per MMBtu for 10,000 MMBtus per day of
natural gas in Dec 2011
Seller will pay buyer the average Gas Daily Index at Henry Hub for the
delivery month.
Settlement is 10 business days after the last ow date.
16 / 43
Pricing and Delivery
Physical Versus Financial
Physical transactions (forwards and some futures) involve
delivery of the commodity at a specied location.
Financial transactions (swaps) involve cash settlement based
upon a benchmark price index prevailing at the time of
analogous physical delivery.
In a physical transaction no reference to an underlying price index is
required.
Delivery is often assumed to be ratable (uniform volume) over a specied
interval.
The third example is a nancial transaction.
17 / 43
Pricing and Delivery
Notional
The notional quantity of a transaction is often dened in
terms of ow rate (per day or per hour) as opposed to a total
notional.
A market standard for delivery quantity is often referred to as
a lot.
For natural gas a lot is 10,000 MMBtus.
For crude oil a lot is 1000 barrels.
In the third example:
The total notional would be 310,000 MMBtus.
This would be articulated as one lot a day or more succinctly as
one-a-day on a trading desk.
18 / 43
Pricing and Delivery
Notation
We will denote the forward price observed at time t for
delivery at time T by F(t, T).
In the case of a delivery interval, this will be replaced by
F (t, T, T + S) where [T, T + S] denes delivery interval over
which ratable (uniform) delivery of the commodity is assumed.
In cases where the delivery interval is a contract month m we
will abbreviate notation along the lines of F
m
(t) or F (t, T
m
).
In the third example F (0, T
m
) = $5.20 .
19 / 43
Pricing and Delivery
Notation
In practice the delivery interval is treated as a discrete set of
delivery days for the purpose of pricing and operations.
Note that:
F
m
(t) =
1
T
m+1
T
m
_
T
m+1
T
m
F(t, T)dT (1)
There is no discounting here as settlement usually occurs on
the same date in the following month.
20 / 43
Pricing and Delivery
Spot Prices
The oating price in the third example is often referred to a
spot price, which is the price for immediate delivery.
A spot price is in almost all situations technically a forward
price with a delivery time very close to the present.
Formally the spot prices is represented by F(t, t)
In practice, the price is usually established slightly before the
delivery time, rendering the distinction between spot and
forward somewhat arbitrary.
In the case of natural gas, trading for delivery on day d occurs on day
d 1 which is when the index print is established.
For power the spot price can be set a day before, hour before or
immediately at delivery.
For coal in which logistics and shipment are an issue, spot can refer to
a time-lag between trade date and delivery measured in weeks or months.
21 / 43
Pricing and Delivery
Basic Facts about Forwards
The unit value of a long position in a forward contract at time
t struck at time t = 0 is given by:
V (t, F (t, T)) = d (t, ) N [F(t, T) F(0, T)] (2)
where denotes the settlement time, N the notional and d()
the discount factor.
Deltas for forwards are discounted notionals:

V
F(t, T)
= d (t, ) N (3)
For a futures contract the unit is the undiscounted notional
N due to daily margining.
22 / 43
Pricing and Delivery
Basic Facts about Forwards
Swaps and forwards often trade as strips.
The term strip refers to a set of adjacent months.
Except at short tenors, commodities usually trade as strips.
Seasonal strips are a collection of commodity specic adjacent months.
Calendar strips to the months in a calendar year. Cal12, for example,
refers to the delivery period consisting of the twelve months comprising
the year 2012.
Strips typically trade at a single xed price, even though
individual contract prices can vary signicantly.
23 / 43
Pricing and Delivery
Basic Facts about Forwards
The fair-value of the strip m {M
1
, . . . , M
2
} must satisfy:
M
2

m=M
1
N
m
[F
m
K] d (t,
m
) = 0
where:

m
are the settlement times
N
m
denotes the monthly notionals (which in general are dierent due to
day count.
Therefore the xed price for the strip is:
K =

M
2
m=M
1
N
m
F
m
d (t,
m
)

M
2
m=M
1
N
m
d (t,
m
)
24 / 43
Pricing and Delivery
Options
Mechanics varies by commodity.
Common Themes
Options mechanics tend to mirror conventions for futures and
swaps.
Expiration can result in either nancial settlement or physical
positions.
Expiry is usually close to the contract month.
MxN options markets where expiry can be M units of time before
delivery at N are not traded.
Multiple Time Scales
Typically markets support options that exercise into monthly
exposure or into annual (cal strip) exposure.
For power daily options are commonly traded.
25 / 43
Forward Yields
Backwardation and Contango
Backwardation: Forward price decreases with tenor
(associated with supply stress).
Contango: Forward price increases with tenor (associated with
supply excess).
The following gures shows snapshots of forward curves for
WTI and NG.
Note the variations in regime, including mixed states of
contango and backwardation
26 / 43
Forward Yields
Snapshots
WTI forward curve at a variety of dates:
Note the range of prices as well as the changes in the monotonicity
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
40
50
60
70
80
90
100
$
/
B
B
l
WTI Forward Curves


05Jan2006
05Jan2007
07Jan2008
05Jan2009
05Jan2010
05Jan2011
27 / 43
Forward Yields
Snapshots
NG forward curve at a variety of dates:
Note:
The seasonality superimposed on macro trends.
The breakdown from the WTI price levels in recent years.
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
4
5
6
7
8
9
10
11
12
$
/
M
M
B
t
u
NG Forward Curves


05Jan2006
05Jan2007
07Jan2008
05Jan2009
05Jan2010
05Jan2011
28 / 43
Forward Yields
The Carry Formalism
Forward curves can be viewed as yield curves.
Forward yield:
y(t, T, T + S) =
1
S
log
_
F(t, T + S)
F(t, T)
_
The forward yield annualized rate implied by borrowing to buy
the commodity at time T and sell it at time T + S.
Negative forward yields imply that market participants are
willing to pay a premium for earlier delivery
This is eectively lending at negative rates.
This happens when supply is contrained.
29 / 43
Forward Yields
The Carry Formalism
Yields often exhibit extreme values.
The following is the WTI forward curve and forward yield for
S = one month in early Jan2009.
2009 2010 2011
35
40
45
50
55
60
65
70
F
o
r
w
a
r
d

P
r
i
c
e

(
$
/
B
a
r
r
e
l
)
WTI Forward curve: 15Jan2009
2009 2010 2011
0
50
100
150
200
250
Y
i
e
l
d

(
%
)
WTI Forward Yields: 15Jan2009
30 / 43
Forward Yields
The Carry Formalism
Seasonality yields negative forward yields consistently for
seasonal commodities.
2010 2011 2012 2013
5.5
6
6.5
7
7.5
F
o
r
w
a
r
d

P
r
i
c
e

(
$
/
M
M
B
t
u
)
NYMEX NG Forward curve: 25Jan2010
2010 2011 2012 2013
150
100
50
0
50
100
Y
i
e
l
d

(
%
)
NYMEX NG Forward Yields: 25Jan2010
31 / 43
Forward Yields
The Carry Formalism
For purely nancial assets the presence of decreasing forward
curves presents an apparent arbitrage opportunity.
Why not short the commodity at the high prices and
repurchase at the low prices?
The answer is that you can, but that the lender of the
commodity, being fully aware of the term structure will charge
accordingly.
In practice this deal structure, referred to as a park-and-loan
usually involves repo in the easy direction, namely buying in
the cheaper months and storing to the expensive months.
How does one address this specialness?
32 / 43
Forward Yields
The Carry Formalism
Case 1: Investment commodity with no storage costs:
F(t, T) = F(t, t)e
r (t,T)(Tt)
or more generally:
F(t, T) = F(t, S)e
r (t,S,T)(TS)
33 / 43
Forward Yields
The Carry Formalism
Case 2: Investment commodity with storage costs:
F(t, T) = F(t, t)e
[r (t,T)+q(t,T)](Tt)
where q(t, T) denotes the instantaneous cost of storage.
Given that q(t, T) 0 this would result in contango being
observed almost universally; a statement at odds with the
facts.
The cost of storage is not exogenous.
Storage owners will charge what the market will bear
The cost of storage is in reality a function of forwards and vols
as opposed to an input.
34 / 43
Forward Yields
The Carry Formalism
Case 3: For a consumption commodity all we can state with
certainty is that:
F(t, T) F(t, t)e
[r (t,T)+q(t,T)](Tt)
.
Rationale: One can always buy the commodity at the spot
price and ensure storage to delivery at T.
Convenience Yield: Solely for the comfort of an seeing
equality, this is often rewritten as:
F(t, T) = F(t, t)e
[r (t,T)+q(t,T)(t,T)](Tt)
.
All that can be ascertained from market data is q , which
makes the above representation more form over substance.
35 / 43
Forward Yields
The Carry Formalism
Recall the forward yields for WTI shown shortly after
inception of the credit crisis:
2009 2010 2011
0
50
100
150
200
250
Y
i
e
l
d

(
%
)
WTI Forward Yields: 15Jan2009
36 / 43
Forward Yields
Eects of Inventory
Inventory levels and forward yields are intimately coupled.
High forward yields (contango) incentives owners of storage to
injectthis occurs when there is a surplus.
Negative forward yields (backwardation) encourages withdrawalsduring
times of scarcity.
The following gure shows OECD crude oil stocks.
2002 2004 2006 2008 2010
3700
3800
3900
4000
4100
4200
4300
4400
M
i
l
l
i
o
n
s

o
f

B
a
r
r
e
l
s
OECD Crude Oil Invetory
37 / 43
Forward Yields
Eects of Inventory
This gure shows the forward yield between the rst two cal
strips of the WTI forward curve yield versus inventory levels.
3700 3800 3900 4000 4100 4200 4300 4400
25
20
15
10
5
0
5
10
15
20
25
WTI 1st/2nd Cal Strip Carry Versus Inventory
Millions of Barrels
A
n
n
u
a
l
i
z
e
d

C
a
r
r
y

(
%
)
Backwardation (Carry<0)
Contango (Carry>0)
38 / 43
Forward Yields
Eects of Inventory: Forward Yields
Incentives: the huge credit-crisis contango resulted in a
massive increase in the use of VLCCs store oil and rened
products.
The gure shows the result outside of the Port of Singapore
during Jan2009.
39 / 43
Forward Yields
Eects of Inventory: Benchmarks
The high dimensional nature of energy commodities requires
benchmark pricing.
Prices of an array of products are referenced as a spread
(basis) to liquidity centers.
For crude oil the dominant global benchmarks are WTI and
Brent.
In recent years there has been a massive decoupling of WTI
from global crudes due to build-up of PADD2 (mid-U.S.)
crude oil inventory.
This has resulted in serious concerns about the viability of
WTI as a benchmark.
40 / 43
Forward Yields
Eects of Inventory: Benchmarks
2005 2006 2007 2008 2009 2010 2011
5
6
7
8
9
10
11
x 10
4
0
0
0
s

B
b
l
PADD2 Inventory
2005 2006 2007 2008 2009 2010 2011
5
0
5
10
15
20
U
S
D
/
B
b
l
BrentWTI 2nd Nearby
41 / 43
Forward Yields
Eects of Inventory: Benchmarks
This gure shows the scatter of the Brent/WTI basis versus
Padd2 inventory.
5 6 7 8 9 10 11
x 10
4
4
2
0
2
4
6
8
10
12
14
16
000s Bbl
U
S
D
/
B
b
l
Brent/WTI 2nd Nearby (Weekly Average) versus PADD2
42 / 43
Conclusion
Summary
High volatility impacts deal valuation, hedging and credit
exposure/capital requirements.
High dimensionality is an inherent feature of energy markets
requiring benchmark pricing/hedging and often resulting in
residual incompleteness.
Viewed as yield curves, energy forward curves can exhibit very
large yields of both signs.
Inventory eects are a signicant driver of forward yields (and
conversely).
43 / 43

You might also like