ETS, toxic debt and the

Australian power market

Paul Simshauser

Overview

• The financing of power stations in

Australia
• Template power station acquisition
• Sizing “Project Finance” for a power
station & financial engineering
• Carbon pricing in 2004: …you
generators should have seen this
coming…
• ETS in 2009: toxic debt
• Implications and policy options

1
 Financing power stations

• Australia currently has about

46,000MW of large-scale power station
capacity
Generation Number Installed Energy Market CO2 Replacement Replacement Average Total Useful Remaining Depreciated
Technology of sites Capacity Produced share Footprint Cost Value Fleet Age Life Useful Life Value
(#) (MW) (GWh) (%) (t/MWh) ($/kW) ($m) (Yrs) (Yrs) (Yrs) ($m)
Brown Coal 8 7,335 55,506 24.8 1.1 - 1.55 2,750 20,171 28.1 50 21.9 8,842
Black Coal 21 21,994 128,386 57.3 0.8 - 1.0 2,250 49,487 23.8 50 26.2 25,968
Natural Gas 38 7,146 14,451 6.5 0.5 - 0.7 1,100 7,861 17.8 30 12.2 3,209
CCGT 10 2,303 11,812 5.3 0.40 1,550 3,570 6.9 30 23.1 2,747
Hydro 43 7,609 13,726 6.1 0.00 2,500 19,023 37.2 100 62.8 11,953
TOTAL 120 46,387 223,881 100.0 1.00 avg 2,564 100,111 24.9 54 29.2 52,719

• 224TWh, fleet average age of 25 years

and sunk investment of c.$53 billion
spread across 120 sites

Financing power stations

• Historically, capacity was added by

State Electricity Commissions with
financed backed or issued by State
Govts or their Central Borrowing
Agencies
• Monopoly pricing capability meant that
capital was not critically scarce
because of virtual certainty of recovery
(which is why we deregulated!)
• All this changed, critically, in the early
1990s due to fiscal imbalances in
Victoria

2
 Financing power stations

• Between 1993 and now, the private sector

has gone from 0% to 39% plant stock
ownership:
PRIVATE Number Installed Energy Market CO2 Replacement Replacement Average Total Remaining Depreciated
SECTOR of sites Capacity Produced share Footprint Cost Value Fleet Age Useful Life Useful Life Value
(#) (MW) (GWh) (%) (t/MWh) ($/kW) ($m) (Yrs) (Yrs) (Yrs) ($m)
Brown Coal 8 7,335 55,506 24.8 1.1 - 1.55 2,750 20,171 28.1 50 21.9 8,842
Black Coal 5 3,669 19,114 8.5 0.8 - 1.0 2,250 8,254 17.1 50 32.9 5,438
Natural Gas 26 5,652 11,416 5.1 0.5 - 0.7 1,100 6,218 18.0 30 12.0 2,493
CCGT 6 1,463 6,858 3.1 0.4 1,550 2,268 6.9 30 23.1 1,743
Hydro 5 526 565 0.3 0.0 2,500 1,315 43.7 100 56.3 740
TOTAL 50 18,645 93,459 41.7 1.06 avg 2,050 38,226 22 44 22.1 19,257

• The private sector has invested about $19

billion, and empirically, on a basis consistent
with MM’s Proposition II. The fit is even
better when their simplifying assumptions of
no taxes and no capital market imperfections
are lifted. That is: “max(debt_capital)”

Financing power stations

• Power sector is the 3rd largest borrower after

Govt & the Financial Services Sector itself.
• Debt is maximised in a power project capital
structure via a Project Finance.
– A single asset lending (mostly)
– Long expected useful life (25yrs+)
– Debt is long-dated money (12yr term and notional
repayment of 25yrs)
– Structured as a limited resource loan
– Various debt covenants
• Technical plant parameters, reserve accounts, MAC
clauses, review events… and
• Financial ratios: DSCR, LLCR & Gearing

3
 Financing power stations
• With a project finance, investors can obtain
60-80% from a syndicate or club
– Volatility of Cash Flows (merchant v PPA)
– Regulatory regime, plant technology, economic life,
system dd-ss, barriers to entry, position in
aggregate ss function
Power Plant Plant Plant Total Debt Debt Refi Refi Lead Project Finance Banks (i.e. excluding
Capacity Age Debt closed margins Amount Date syndication banks)
(MW) (Yrs) ($M) (Yr) (bps) ($M) (Yr)
Hazelwood 1,600 39 1,207 2001 155-185 445 2010 BA, RBS, SocGen, ANZ*
Tarong North 450 6 363 2002 100-160 162 2006 BOTM, ANZ*, Mizuho, Fortis,
Millmerran 852 6 1,025 2002 110-160 467 2012 ANZ*, Banca Intesa, Calyon, Fortis, HSBC, KBC, Mizuho, RBS,
SMBC, UOB, WestLB
Loy Yang A~ 2,120 22 2,650 2004 140-185 313 2010 ABN, BOTM, Calyon, ANZ*, Mizuho, NAB*, RBS, Sumitomo,
WestLB, Westpac*
Yallourn 1,480 30 2,500 2005 75-85 650 2009 NAB*, CBA*, JP Morgan (NB. Corporate-style facility)
Loy Yang B`` 1,000 13 1,100 2006 50-80 620 2012 BOTM, BNP, CBA, ANZ, NAB
Callide C 900 8 390 2007 90 2012 BNP, BOSI, BA, Fortis, NAB*, Mizuho
Transfield^ 180 30 800 2008 115-120 800 2011 Westpac*, ANZ*, CBA*, RBS
Bluewaters I & II 430 0 950 2008 115-145 250 2014 ANZ*, NAB*, WestLB, SocGen
BBP# 780 23 2,700 2008 190-210 1,600 2011 ANZ*, BNP, CBA*, Dexia, NAB*, Natixis, SocGen, HVB,
WestLB, Suncorp*, BOSI
Total 9,792 21 13,685 120-140 5,397 25 MLA Banks: 5 Australian, 20 Foreign
`` Loy Yang B also had approximately $200m of Subordinated Debt within its capital structure. *Domestic Bank
# BBP has $400m in Mezzanine Debt in its capital structure. ^ Transfield facility refinanced in 2008.
~ Loy Yang A also has Senior Debt provided by a large number of CPI Bond Holders.
Source: Reuters BasisPoint.

Financing power stations

• $13,700m in senior debt ($11.2b PF)

and about $5,400m due for refinancing
over the next 3-4 years
• This is a problem – all debt has been
sized on a BAU scenario, not on an ETS
scenario
• To see how ETS would impact a PF’ed
power station, we must turn to the
acquisition and debt sizing

4
The valuation of power station assets

• Assume we purchase a 1000MW brown

coal power station in 2004:
Inflation Taxation
- CPI (%) 2.75 - Tax rate (%) 30.00
- Elec Prices (%) 2.06 - Useful life (Yrs) 30

Plant Costs & Prices Debt Sizing Parameters

- Plant size (MW) 1,000 - DSCR (times) 1.8 to 2.2
- Acquisition Price ($M) 1,900 - LLCR (times) 1.8 to 2.2
- Electricity Price ($/MWh) 36.00 - Gearing (%) 67.5
- CCGT 2004$ ($/MWh) 44.00 - Lockup (times) 1.35
- CCGT 2009$ ($/MWh) 51.00 - Default (times) 1.10
- Heat rate (kJ/kWh) 13,300 Facilities
- Unit fuel ($/GJ) 0.30 - Tranche 1 (Bullet) (Yrs) 5
- O&M costs ($M) 36.7 - Tranche 1 Refi (Yrs) 20
- Capex ($M) 5.0 - Tranche 2 (Amort.) (Yrs) 12
- CO2 footprint (t/MWh) 1.32 - Tranche 2 Refi (Yrs) 13
- Remnant life (Yrs) 40 - Notional amortisation (Yrs) 25

Cost of Capital Debt Costs

- Post Tax Equity (%) 15.00 - 5 year swap (%) 6.01
- Pre Tax Equity (%) 21.00 - 12 year swap (%) 6.14
- Pre Tax Debt (%) 7.42 - 5yr Margins/BBSW (%) 1.20
- Pre Tax WACC (%) 11.80 - 12yr Margins/BBSW (%) 1.40
- Refinance (Headline) (%) 7.50

The valuation of power station assets

• EBITDA of $199m, pre-tax WACC of

11.80%, Post-tax Ke 15% (+50bps)
Enterprise
Valuation Metrics Valuation
EBITDA Multiple 9.5x 1,893,227
DCF Pre-Tax, Ungeared 11.8% 1,895,841
DCF Post-Tax Equity 15.0% 1,915,455
$/kW Multiple $1,900 $ 1,900,000
Average: 1,901,131

Assumed Acquisition Price: 1,900,000

Running Equity Yield 12%

• LYB was acquired for slightly below $2b

in 2004, so the metrics stack up well

5
 Sizing a PF

• In the previous table, the equity IRR (at

15.50%) was calculated with an implied
$1,240m PF in place
– DSCR 1.8 to 2.2 (lockup 1.35x, default 1.1x)
– LLCR 1.8 to 2.2
– Gearing:  67.5%
– Debt structured in two tranches (25yr notional term)
• Tranche 1: 5 yr bullet, 6.01BBSY + 120bps
• Tranche 2: 12yr amort, 6.14BBSY + 140bps

Trache 1 Tranche 2 Aggregate

Debt sizing covenant Bullet Amortising Term Debt
($M) ($M) ($M)
DSCR at 1.8x 434 806 1,240
LLCR at 1.8x 498 926 1,424
Gearing at 67.5% 449 834 1,283

Min Result: DSCR 1,240

A few observations:
• Sizing is being driven by CFs in year 6 on
Sizing a PF DSCR calcs, notice the step-up in ‘p’.

Expected annual earnings

•Opex is 27% and capital costs & returns 73%
($ '000) •Equity IRR is 15.5% or 50bps over
800,000
Equity Debt sizing parameters
- Gearing: 65.4%
Taxation
700,000 - LLCR: 2.07 times
Interest Payments - DSCR: 1.85 times
Debt Redemption
600,000
Carbon
Capex
500,000
O&M

400,000 Fuel

300,000

200,000 CADS is 1.8x

CADS is 1.85x

100,000

-
04

05

06

07

08

09

10

11

12

13

14

15

16

17

18

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Calendar Year

6
 Financial engineering
• Thus is the ‘bank case’. The equity case will
look a little different. Higher prices, different
refi assumptions, because of variances in
forward expectations
Moving pre-tax valuation ($)

2,500,000

2,000,000

1,500,000

1,000,000

500,000

-
04

06

08

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40

42
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Calendar Year

Financial engineering
Expected annual earnings
($ '000)
800,000
Equity Debt sizing parameters
- Gearing: 65.4%
Taxation
700,000 - LLCR: 2.07 times
Interest Payments - DSCR: 1.85 times
Debt Redemption
600,000 Gains from refi
Carbon
Capex
500,000
O&M

400,000 Fuel
Step-up in ‘p’
300,000

200,000 CADS is 1.8x

CADS is 1.85x

100,000

-
04

05

06

07

08

09

10

11

12

13

14

15

16

17

18

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Calendar Year

7
Carbon pricing in 2004: you generators…

• The subject of continuous revision… you

generators should have seen this coming:
– IPCC in 1993, Federal Govt in 1999, 2004 and 2007
(PMT), States in 1996 (NETT)
– But formal position until mid-2007 was, not until the
technology exists, & not on our own (read: not
before 2012)
– “The fact that all equity and debt market
participants used BAU economics is the practical
evidence of how govt policy was interpreted”
– And then on 3 June 2007 the PM made his
statement, all generators that existed prior to this
date will qualify for ‘compensation’

Carbon pricing in 2004

• I believe that industry accepted ETS would be

in force between 2012-2015
• But I also believe that:
– Not likely pre-2012
– CO2 prices do not have a long history at high levels
($5/t in Millmerran 1999, $10/t in ESAA 2003, $12/t
avg of 100 peer-reviewed studies in IPCC 2007)
– 95% of permits allocated for free to EU15 Gencos
until quite recently, and for 8 years (2005-2012)
– Urgency of the task is relatively new, post Stern &
in our case, Garnaut.
• So how would our $1240m debt be resized in
2004 given what we knew then? ($10/t, 2012
start, CCGT at $44)

8
 Carbon pricing in 2004
As a scenario, debt would not be resized. But as base case, down just $57m on $1240m
Expected annual earnings
Asset revalued down by $140m. Equivalent to pe  $1/MWh & -25bps on CPI
($ '000)
800,000
Equity Gearing: 70%
LLCR: 1.91 times
Taxation
700,000 DSCR: 1.80 times
Interest Payments
Debt Redemption
600,000
Carbon
Capex
500,000
O&M

400,000 Fuel

300,000 CADS is 2.0x

200,000

100,000

-
04

05

06

07

08

09

10

11

12

13

14

15

16

17

18

19

20
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Calendar Year

ETS in 2009: creating toxic debt

• 2009 is a very different environment

– Task of abatement more urgent
– Quantity cuts much greater
– e(p) CO2 is materially higher
• To analyse the effect, this study uses
three key variables for revenue;
– NEM CO2 intensity from Simshauser, Doan
and Lacey (2007) and CRAI (2007)
– pe of electricity commodity at $51/MWh and
– CO2 at $20/t plus $10/t shift in 2013, and
cost of carry escalation at CPI+4%

9
 ETS in 2009: creating toxic debt

Price of CO2 Headline Electricity Price ($/MWh) NEM CO2 Intensity

(A$/t €/t) Deutsche EU ETS CO2 Price (Nominal A$) (CO2/MWh)
120.00 Deutsche EU ETS CO2 Price (Nominal €) 1.00
CO2 Price (Nominal A$)
NEM CO2 Coefficient (t/MWh)
CO2 Price (2009 A$)
NEM CO2 Intensity (RHS)
$97.22
100.00 $94.79
$92.39 0.80
$90.03
$87.71
$83.69 $85.12
Headline Electricity Prices $81.72

80.00
$72.00
$70.33
$68.66 0.60

60.00 $54.23
CO2 Price (Nominal A$) $50.80
Deutsche EU ETS CO2 Price in $A/t $47.59
$44.58 0.40
Deutsche EU ETS CO2 Price in €/t $41.76
$39.12
40.00 $36.65
$34.33

$21.35 $22.79
$20.00 0.20
20.00

- -
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
A$1.00 = €0.5215 Year

ETS in 2009: creating toxic debt

• The ramifications for our 1000MW brown

coal power station are very material
– DSCR will not be breached, so the usual
triggers of financial distress are not present
– The triggering actually occurs via the
Directors of the firm under the Corporations
Act 2001(Cth)
• S.295, 296 and 297 re financial statements
• S.344, 1308 and 1309 in Chapter 2M
• Driven by AASB136 Impairment of Assets
– The application triggers to covenants
• Gearing ratio (103% vs 67.5%
• LLCR (0.91x vs. 1.1x default)

10
ETS in 2009: creating toxic debt

• At this point, the asset is written down by

$900m or 44%; from $2040m (pv valuation) to
$1140m
• Orinarily owners would need to step in and
cure the default, but as Harvin & Dell noted in
2003:
• It is naïve to expect that project sponsors will
contribute additional capital to avoid the default,
especially if there has been deterioration in the long-
term credit quality of the project… Examples include
difficulty in complying with evolving environmental
policy and the introduction of more efficient
technologies. It is conceivable that technology
deficiencies can be resolved with capital infusion.
More likely, however, the reduction in annual cash
flow is permanent… Equity investors have walked
away from projects, essentially turning over the keys
to debt-holders…

ETS in 2009: creating toxic debt

• At this point, the banks will establish a

workout committee with a view to sell the
asset or place it in administration
– Between 2002, there were 57 Project Financings.
No Bank has ever lost principle in Australia.
– They did in UK on Drax, with consequences for the
industry
– In our case, it is a ‘cents in the dollar’ exercise for
the project banks
– 97cents on our base forecast or 60cents if you
switch to the Deutsche f’cast of CO2.
– Equity is in lockup (default) immediately and the
plant is forecast to be bankrupt in 2022.

11
 ETS in 2009: creating toxic debt
•First 6 years outperformed bank base case by $30m or 50bps
•ETS leads to a $944m hit to assets; 44% writedown, (97c in the $)
Expected annual earnings •EU ETS price leads to a 65% writedown (60c in the $)
Equity
($ '000) Taxation Gearing: 103%
800,000 Interest Payments LLCR: 0.91 times
Scheduled Debt Redemption DSCR: 1.80 times
Accelerated Redemption
700,000 Carbon
Capex
O&M
600,000 Fuel
Bankers Base Case
500,000

400,000 CADS is 2.1x

300,000

200,000

100,000

pt
04

05

06

07

08

09

10

11

12

13

14

15

16

17

18

19

20

Ba 1
2

ru
20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

nk
Calendar Year

Policy Implications

• Without stating the obvious, this is seriously

bad on sovereign risk grounds
• $15b collective hit on $35b aggregate coal
plant stock; 43% loss on average
• Policy makers know losses are inevitable,
hence ESAS
• But it is unlikely to provide 100% of the ‘asset
hit’. Equity participants will exit and be
replaced seamlessly, because that’s what
microeconomic theory tells us (in a world
where firms are passive variables in a GEM
and do not supply an essential service)
• Policy changes occur all them time without
adjustment assistance, although rarely with
such intensity

12
 Policy implications

• Following ETS and RET, the investment task

facing the power sector has been increased.
Between now and 2020, the median forecast
is c.$35 billion.
• Given current forecasts of wealth transfers, a
sovereign risk tag for the Australian Power
Sector is both more than a theoretical
possibility and non-trivial in every sense…
• Vast parts of the industry have drawn on
foreign equity and foreign debt, and our
reliance on this source of capital is critical on
cost competitiveness grounds…

Policy Implications
Enterprise Value of
Coal Power Stations
($ million) Number of firms
4,000
4
27%
3,500 Aggregate investment by 13 firms: $14,280 million (100%)
- Domestic investment by 4 firms: $3,875 million (27%)
2 - Foreign investment by 9 firms: $10,405 million (73%)
3,000
22%
3
1
2,500 19%
18%

2,000 3
14%

1,500

1,000

500

0
Domestic China & Hong Kong Japan UK USA
Investor Base

13
Policy Implications

• So, 13 firms invested in coal fired

assets and 9 of them (73% by value)
are foreign
• And there are only 22 generators in
Australia (18 private and 4 State Govts)
• Banks… Just 25 listed on Slide 7.
– It is not hard to see this number drop below
10 in the current (unrelated) crisis
– Memory of banks under loss-making
conditions is not like equity. The memory
runs much, much longer; bear in mind there
is nowhere near the level of diversification
(infinite equity investors vs. 25 banks).

Policy Implications

• Our 5 domestic banks are simply incapable of

writing the debt necessary to finance $35
billion in power assets
– To give this some perspective; Uranquinty. Great
project, right NEM region, right technology etc etc
– $500m Investment, $340m term facility
– 3 MLA’s – Suncorp, nabCapital & KBC
– 3 further syndication banks (all foreign)
– So one simple peaking project involved 2 local
banks and 4 foreign banks (importantly, pre-
financial crisis. It would take more banks now.)
– And the banks have other homes. In 2008 despite
the financial crisis, M&A activity topped A$115
billion and the forecast greenfield development
between now and 2030 is US$6,800 billion

14
Survey of Project Bankers

• Sent to 23 Project Finance and Syndication

Bankers. 16 responses received from 4
domestic & 9 foreign banks
• Survey sought historic experience and
expectations on margins, term, participation
and key drivers
• On margins: 120-140bps in 2005, currently
250-300bps, 200-250bps by 2011 after the
crisis
• On term: 12 yrs in 2005, currently 5 years
max, up to only 7 yrs in 2011
• On participation for a $1240m term facility: 4-
6 lead banks plus 4-8 syndication banks in
2005; now club deal with >9 banks; in 2011, 7-
8 lead banks plus >9 syndication banks

Survey of Project Bankers

• Market participation: 27 banks in 2005; in

2008 have reduced to just 14, and only 17
expected to be around in 2011
• In the event of a power project insolvency
due to ETS policy; 50% will exit from coal
finance; 36% from gas-fired finance and 14%
would exit entirely (including renewables)
• Key drivers: health of market and players;
stabilisation of financial markets; stability of
regulations; margins; performance of
committed facilities

15
Policy Implications
• ESAS is likely to comprise an allocation of permits,
and probably over a 5 year time frame
• This will suit some generators, but not all generators
• Given the assumptions in this paper, our 1000MW
plant would need 75% of its 9.6mtpa CO2 permits
allocated to restore covenants, and 110% to restore
equity
• The pool of funds for the ESAS is limited; and
competing with households and EITEI
• So this being the case, there are two residual policy
alternatives:
1. Quickly adjust the NEM mechanisms to inlcude a
supplementary capacity payment to coal generators
2. Do nothing, and accept a Wounded-Bull Scenario per
Simshauser & Doan (2009); which is code for price shift
from the current c.$45/MWh to c.$100/MWh (including
CO2) at the wholesale level.

Concluding remarks
• In the absence of a suitable adjustment package, it
seems we are likely to create toxic debt
• Private sector holds 39% of power station capacity
worth$19 billion. $13.3 billion in senior debt
supporting this
• CO2 only became this big a problem fairly recently, i.e.
from 2007, not before.
• In the case of our 1000MW brown coal generator with
1.32t CO2, equity is wiped out and debt recovery is 60-
97 cents in the dollar
• 20 of the 25 project banks are foreign
• 9 of the 13 equity investors are foreign
• Sovereign risk therefore seems both more than a
theoretical possibility, and given the magnitude of the
funds and potential impact on values, is non-trivial in
every sense.
• The form and quantum of ESAS is critical.

16