E.ON – Cleaner & better energy
Energy Trading
E.ON strategy
Investment
Less capital,
more value
Europe
Focused &
synergistic
positioning
Performance
Efficiency &
effective
organization
Outside
Europe
Targeted
expansion
From
To
Integrated across
value chain
Focus on competitive
businesses
Eurocentric
Targeted expansion
outside Europe
Selective efficiency
programs
Sustainable performance
culture
Capital intensive
Competence-based
Cleaner & better energy
Transform European utility into global, specialized energy solutions provider
2
E.ON Group strategic priorities
Performance
Æ Intensify cost & quality management
Challenging markets
Æ Simplify structures
Political interventions
Æ Execute portfolio measures
Æ Create balance sheet flexibility
Europe:
System transformation
Outside Europe:
Growth & new technologies
Growth
Æ Capture growth in renewables & decentralized energies
Æ Exploit opportunities in new markets
Markets require intensified self-help measures
3
E.ON Group key financial targets
New
Results
Old
y 2011E1 Adjusted EBITDA (€bn):
Adjusted EPS (€/share):
9.1 – 9.3
1.2 – 1.3
9-1 - 9.8
1.1 - 1.4
y 2013E Adjusted EBITDA (€bn):
Adjusted EPS (€/share):
11.6 – 12.32
1.7 – 2.02
>134
~2.44
y 2015E Adjusted EBITDA (€bn):
Adjusted EPS (€/share):
12.5 - 13.03
2.0 – 2.33
Dividends
y
y
y
y
Dividend payout policy (% adj. net income):
2011 (€/share):
2012 (€/share):
2013 (€/share):
Other
y
y
y
y
Medium-term debt factor
Investments 2011-13 (€bn):
Total disposals until 2013 (€bn):
Rating target
50 - 60
1.0
1.1
≥1.1
50 - 60
≥1.3
≥1.3
<3x
~19
~15
Solid single A
≤3x
19
~15
Solid single A
Transparent financial targets for coming years
Assumed 2015 debt factor allows ~€6bn of additional growth CAPEX
1. 2011 post €0.5bn effect of achieved disposals (€9.1bn) 2. 2013 Post €0.9bn effect of achieved disposals (€9.1bn) 3. 2015 Post ~€1.7bn effect of total disposals effect (€~15bn) 4. Pre disposals
4
Trading within E.ON group structure
Group Management
Generation1
Renewables2
Trading
Gas
Germany3
Other EU
countries3
Russia4
Support
functions5
Optimization
Proprietary Trading
Leaner and more market oriented organization
1. Incl. EBITDA of all conventional generation (previously in Market Units) 2. Incl. hydro 3. Distribution and sales; gas sales included in Germany 4. Special focus country 5. IT, Procurement, Insurance,
Consulting, Business Processes, these are not reported separately externally 6. “Outside Europe” to be reported separately after having reached the necessary size
5
Trading – Business strategy
Optimize commodities exposure and support business
Market environment
y Increasing scope and scale of integration of
power markets across EU (e.g. market coupling
between Nordic and CWE in November 2010)
y Increasing gas-to-gas competition in Europe
y Key commodities as well as LNG and CO2 traded
on global markets
Strategic priorities
y Cross-regional and cross-commodity synergies:
monetize value of flexibility in power plants, supply
contracts, gas storage
y Seek new opportunities in cross-border activities
(e.g. intra-day)
y Global commodity trading (e.g. coal & freight)
backed by European portfolio
y Origination activities to earn higher margins on nonstandard, non-commodity specific, longer term
products
Leading energy trader
6
Trading – Sustainable value contribution
Adjusted EBITDA (€ bn)
Leading energy trader
y One of the biggest and most diversified underlying
power & gas asset positions
1.2
- 0.5 – -0.7
y Market access throughout Europe to capture
synergies (e.g. reduction of credit risk)
y Global scope of trading to cover majority of E.ON’s
commodity risk position
y Strong support of European liberalization agenda
(e.g. engagement for market coupling)
Year on year increase in Trading’s volumes (2010 vs. 2009)
+34%
+30%
+30%
+19%
2010
2011
2013
2011
y Optimization result is negatively impacted by
swing in internal transfer spread
y Extrinsic value suffering from reduced volatility
y Prop trading expected to improve compared to
weak 2010
2012-2013
Power
Gas
CO2
Coal
y Less distorted optimization result
Distorting effect of transfer prices to normalize by 2013
7
E.ON – Cleaner & better energy
Discussion Material
Trading - E.ON‘s optimization and prop. trading function
Conventional
Generation
y Cross-regional optimization
Single integrated view on all markets & physical assets
y Cross-commodity optimization
Renewables
Generation
Ability to realize benefits of correlation between commodities
y Risk management
Integrated portfolio view and consistent risk/hedging strategy
y Proprietary trading
Global Gas
Asset knowledge and understanding of market fundamentals
EET
Adj. EBITDA development, 2008-2010 (€ m)
1500
Germany
1100
Prop trading
700
Asset optimization
Other EU countries
300
Russia
-100
2008
2009
2010
Integration of trading expertise delivers additional value
9
Trading is E.ON’s centralized interface to the energy markets…
Upstream function1
Optimization function
Downstream function
Generation Unit 2
Retail/sales
subsidiaries
Global Gas
Trading sells product; market margin (achieved price
minus transfer price) sits in EET
Trading procures volumes for the E.ON
supply businesses
European energy markets
Power, Gas, Coal, CO2, Oil
Power/gas volumes (own & procured)
Transfer price, fuel price and volumes
… backed by a strong portfolio of assets
1. In case of Global Gas gas volumes = upstream + procured
2.. Conventional Generation and Renewables Generation
10
Sources of value creation
Optimization
Risk management
Trading creates value for E.ON
Value creation at E.ON Trading
Prop trading
Optimization
y Arbitrage
y Cross regional/border
y Cross commodity
y Timing decisions
Risk management
y Cash flow risk
y Commodity risk
y Counterparty risk
Hedging
Bulk of the value created at E.ON Trading comes from its risk management function
and its (mainly) asset backed optimization function
11
Sources of value creation
Optimization
Risk management
Cross-regional optimization
Illustration via interconnectors
Involved assets
Optimization: use of balancing markets
EET bids for
transportation
capacity on BritNed
Kingsnorth: coaland oil-fired power
plant 1.940 MW
Vilvoorde : coalfired power
plant, 556 MW
Prerequisite
y Access to transportation capacity, e.g. BritNed
Maasvlatke
(Rotterdam) :
coal-fired power
plant, 1.040 MW
Idea
y Use of interconnectors (e.g. BritNed) to assist
optimization and balancing of portfolios both for
E.ON UK and E.ON Benelux
y Balancing power can be used to cover under- or
over-supply situations in UK as well as in Benelux
Value points
y Cross-regional arbitrage
y Reduction of penalty costs for system imbalance
Reaping the value of a broad asset base via cross-regional arbitrage
12
Sources of value creation
Optimization
Risk management
Cross-commodity optimization
Arbitrage via gas-to-power optimization
Generation costs in Germany – gas vs. coal
Gas-to-power optimization
Prerequisite
y Portfolio of gas- and coal-fired power plants
y Plan to generate with a gas-fired plant
y Gas volumes from a supply contract or market
65
60
gas
55
Gen. costs [EUR/MWh]
50
Idea
y Coal becomes cheaper fuel to generate power in
that period
y Decision: Sell the gas at a higher price and
produce power with a coal-fired plant instead
45
40
35
30
coal
25
20
15
Jan-09
Mar-09
May-09
Jul-09
Sep-09
Nov-09
Jan-10
Mar-10
May-10
Jul-10
Sep-10
Nov-10
Jan-11
Value points
y Margin from selling the gas
y Margin producing power with cheaper fuel (coal)
Reaping the value of a broad asset base via cross-commodity arbitrage
13
Sources of value creation
Optimization
Risk management
Cross-commodity and cross-regional optimization
Global coal arbitrage
API 2
Prerequisite
y Combines supply opportunities in Columbia/South
Africa with demand in Europe/ India
y Time charters offer shipment-flexibility to EET
(leading to reduced transportation cost)
SalesContract
E.ON‘s coalfired power
plants
C4
SalesContract
Idea
y Increase of dark spread or sales margin because
of potential lower costs of coal supply, based on
multi-sourcing strategy
y Arbitrage between API4, C4, API2:
y e.g. buy API4 + C4, sell API2
y buy API2, sell API4 + C4
API 4
Value points
y Improve dark spread by sourcing cheaper coal
y Cost reduction through time-charter optimization
Leverage our large underlying demand to profit from global arbitrage opportunities
14
Sources of value creation
Optimization
Risk management
Cross-commodity & cross-regional optimization
Enabling arbitrage via origination
Additional big
coal-fired power
plant in region 1
CCGT Tolling
Contract in
region 1
Region 1
Gas long in
region 2
Region 2
Definition of origination
y Physical or financial commodity transactions of a “non standard
nature” - due to their scale, tenor or structure
y Transactions are of a longer term than traded curves, linked to physical
assets or provide new optimization opportunities to the portfolio
Prerequisite
y Additional coal-fired power plant in region 1
the company’s asset portfolio in that region
will be dominated by coal-fired units
Idea
y By entering a CCGT tolling agreement the fuel
mix for region 1 can be enhanced
y Additionally the company’s long exposure to gas
in region 2 is reduced by delivering volumes to
the counterparty in region 1
y Generation portfolio improves without CAPEX
Value points
y Spark spread of the tolling agreement
y Portfolio-optimization value
Enhance portfolio value with origination structures instead of outright asset ownership
15
Sources of value creation
Optimization
Risk management
Hedging rationale
Ensure more
stable earnings
y Hedging outright power risk strongly reduces y-o-y volatility in cash flows
Increase planning
y Cash flow visibility needed to support capex planning
certainty
Reduce cost of
capital
y For a given leverage hedging reduces the cost of capital
Trading’s function as a risk manager is value enhancing
16
Sources of value creation
Optimization
Risk management
Hedging at E.ON is a key tool for risk management…
y Characterized by high intrinsic value and high value at risk (unhedged)
Hedging nuclear
& hydro plants
y High intrinsic value is function of low variable cost of assets
y Value captured and risk managed by hedging on forward markets
depending on price view and risk appetite
y Characterized by relatively high share of extrinsic value
Hedging flexible
plants
y Power plants represent real options. In case of flexible units (gas, coal)
optionality has a real value extrinsic value
y Dynamic hedging strategies to capture intrinsic as well as extrinsic value
… but also for value creation
17
Sources of value creation
Optimization
Risk management
Dispatch optionality - key characteristic of flexible plants
A gas-fired unit
A nuclear unit
Hourly power price vs. generation cost (€/MWh)
Hourly power price vs. generation cost (€/MWh)
Variable costs
(nuclear fuel + fuel tax)
Variable costs
(gas + CO2 costs)
Mon
Tue
Wed
Thu
Fri
Sat
Sun
Tue
Wed
Thu
Tue
Wed
Thu
Fri
Sat
Sun
Fri
Sat
Sun
Planned hourly output (MW)
Planned hourly output (MW)
Mon
Mon
Fri
Sat
Sun
Mon
Tue
Wed
Thu
A power plant should run as long as the profit margin against variable costs is positive
18
Sources of value creation
Optimization
Risk management
Power plants are real options
Pay-off of a power plant for a single hour
Merit order of German power plants
(Order of power plants on the basis of variable costs)
Profit margin
(fix costs not considered)
Electricity price
[€/MWh]
Hard coal
Oil
Lignite
Nuclear
Natural gas
Run-of-river
& renewables
Pay-off of nuclear
plant
Pay-off of
gas-fired plant
Power plant
not operating
Power plant
operating
0
0
10
20
30
40
50
60
70
80
90
100
Capacity [GW]
y On the electricity market, the price is set hourly
driven by the variable costs of the marginal power
plant [i.e. the last plant required to meet demand]
Variable costs
(nuclear fuel + fuel tax)
Variable costs
(gas + CO2 costs)
Electricity
price
y A power plant runs and earns a positive profit
margin if the power price is above its variable
costs
Power plants can be considered as European call options
19
Sources of value creation
Optimization
Risk management
Additional value is inherent in flexible power plants
Profit margin distribution
Price distribution
25
Spread price
€/MWh
25
20
20
15
Profit margin
€/MWh
15
Price path scenario 1
Profit margin path scenarios
10
10
Expected value
5
Expected value
extrinsic
5
intrinsic
Path scenario 2
0
0
t
t
-5
Today (forward market)
Spread forward price 5€
Path scenario 3
At maturity (delivery)
expected spread price 5€
(without market view)
y Intrinsic value is equal to the actual value of selling
the underlying as forwards
-5
Negative spread,
unit will not run thus no loss
Today (forward market)
Intrinsic value is 5€
At maturity (delivery)
expected profit > 5€
(additional extrinsic value)
y The expected profit at maturity is higher than the
observed intrinsic value in forward market. The
difference is the extrinsic value.
Extrinsic value is essentially the value of not needing to run the plant when it would make a loss
20
Sources of value creation
Optimization
Risk management
Size of extrinsic value influenced by several factors 1/2
1 - moneyness of options
Value
Total option value
intrinsic value
extrinsic value
Electricity price
Fuel + CO2 price
Out of the money
At the money
In the money
Hours of power plants that are nearly „at the money“ have higher extrinsic value
21
Sources of value creation
Optimization
Risk management
Size of extrinsic value influenced by several factors 2/2
3 - time to maturity
2 - volatility
Less volatile price
t0
Short time left before delivery Long time left before delivery
more volatile price
t1
t0
t1
t0
t1
t0
t1
y Higher volatility increases the extrinsic value
y Longer time to delivery increases extrinsic value
Reasoning
y The more volatile the price, the larger the
probability that an option change from “in the
money” to “out of the money” or vice versa.
Reasoning
y The more time left before maturity, the larger the
probability that an option change from “in the
money” to “out of the money” or vice versa.
22
Sources of value creation
Optimization
Risk management
Hedging focus of different types of generation assets
Different types of assets require different hedging focuses …
Profit margin
Hedging focus of
nuclear plant
Secure intrinsic value
Hedging focus of
gas-fired plant
Capture extrinsic value
Pay-off of a power plant
Total value of a power plant
electricity price range
… for which different hedging methods are utilized
Intrinsic value
Extrinsic value
Influencing factors
Power prices (nuclear, hydro)
or spread prices (coal-, gas-fired)
Moneyness, volatility, time
to maturity
Methods to capture value
Straightforward: hedge at
forward markets
Complex: dynamic forward
hedging, delta-hedging, etc.
Criteria for decision making
Price view, risk appetite
Volatility view, risk appetite,
hedge costs
23
Sources of value creation
Optimization
Risk management
How E.ON Trading captures intrinsic value
Forward hedging
100%
Playing field
75%
Delivery
(intra-year
optimisation)
Liquidity constraint
50%
Hedging strategy
Risk appetite constraint
25%
0%
80
2007
Handover
2008
2009
2010
y Upper boundary given by
available market liquidity
y Lower boundary given by Group
risk bearing capacity and risk
appetite
y Optimized hedge path with the
aim to maximize risk-adjusted
return
Forward price Cal-10
70
Achieved price: 68€
Average spot
price
in 2010: 44€
60
50
40
30
2007
2008
2009
2010
Capturing intrinsic value is hedging of the natural long position under risk/return principles
24
Sources of value creation
Optimization
Risk management
How to capture extrinsic value in flexible power plants
Example 1: Dynamic forward hedging
Time
Spread price
Planned generations
Hegdes
Locked in profit
t1
10
generate
sell spread (sell power, buy coal & CO2)
+10
t2
-2
not generate
unwind hedge (buy power, sell coal & CO2)
+2
t3
4
generate
sell hedge (sell power, buy coal & CO2)
+4
total
+16
y Through forward hedging and
rebalance of hedges according to
actual economic generation, a
part of extrinsic value can be
captured in forward market.
On a long term average the extrinsic value can be captured.
However, it is not guaranteed that the theoretical value can be captured in each time.
Example 2: Delta hedging
y Delta-hedging is a dynamic hedging strategy aimed at conserving the full value of a power plant,
without taking a price view.
y By buying or selling the spread, the total position (power plant + spreads bought/sold) can be made
delta-neutral, i.e. the value of the position does not change for small changes of the value of the
underlying. If delta-neutrality is monitored and updated regularly, the full value of the power plant is
conserved.
The extrinsic value can be captured each time. However a trade-off has to be made between high
transaction costs and the certainty of capturing extrinsic value.
25
E.ON – Cleaner & better energy
Backup Material
Glossary
Term
Description
Economic generation
Volume of power that is “in the money” for a given period in the future (based on forward
market prices).
Clean Spark Spread
Theoretical gross margin of a gas-fired power plant from selling a unit of electricity, having
bought the gas and the carbon emission certificate required to produce this unit of electricity.
Clean Dark Spread
Theoretical gross margin of a coal-fired power plant from selling a unit of electricity, having
bought the coal and the carbon emission certificate required to produce this unit of electricity.
Intrinsic Value
Part of option value that is equal to actual mark-to-market price of underlying (actual value of
selling the underlying as forwards)
Extrinsic Value
Time value of the option (total option value less Intrinsic Value)
Delta-hedging
A hedging strategy aimed at conserving the full value of an option, i.e. not only the "intrinsic"
value, but also the "extrinsic" value.
27
Split of responsibilities and risk between Trading and
generation unit
Mid term planning horizon
EET responsibility
Generation unit responsibility
Traded market access
Today
Asset optimization
Asset investments
Commodity price risk
management
Commodity price risk
management
Asset availability
Asset availability
Trading main responsibility
2013
Generation unit responsibility
2020
Trading is responsible for commodity risk management and the optimization
three years prior to delivery
28
Transfer pricing mechanism for outright power
Understanding Trading‘s optimization result using a simplified scheme with an example of Cal 2010 delivery
Handover => transfer price
2010 baseload forward 2007-2010
100
€/MWh
80
y Generation unit transfers all 2010
volumes to EET in course of 2007
y Transfer price for the transferred
volumes is set up (based on 2010
forwards in 2007)
60
40
20
0
Jan-07 Apr-07 Jul-07 Oct-07 Jan-08 Apr-08 Jul-08 Oct-08 Jan-09 Apr-09 Jul-09 Oct-09
Æ Volume x transfer price =
generation unit result in 2010
Hedging => achieved price
100
2010 baseload forward 2007-2010
€/MWh
80
60
40
20
0
Jan-07 Apr-07 Jul-07 Oct-07 Jan-08 Apr-08 Jul-08 Oct-08 Jan-09 Apr-09 Jul-09 Oct-09
y EET hedges volumes within risk limits
y Achieved price by EET evolving with
hedging (e.g. for Central Europe this
is € 68/MWh for Cal 2010)
Volume
x
achieved price
=
Group‘s
generation
revenue
Æ Volume x (achieved price – transfer
price) = Trading outright
optimization
Delta between external achieved price and internal transfer price for a given year of delivery
is reported in the Trading accounts in the optimization result in the year of delivery
29
Hedging of E.ON‘s outright generation
As of Sep 30, 2011
German, Benelux and French market
2011
~ 59 €/MWh 1
2012
~ 54 €/MWh 1
2013
~ 56 €/MWh 1
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Nordic market
2011
~ 43 €/MWh 1
2012
~ 43 €/MWh 1
2013
~ 45 €/MWh 1
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
= percentage band of generation hedged
1. Average realized price only relevant for the pure outright power position (Nuclear/Hydro) sold in the respective year
30
Simplified example: Handover of 2010/2011 baseload
volume in 2007/2008
Simplified examples - very different outcome
German baseload power price (€/MWh)
100
2007 (handover period) => transfer price
Average achieved
Price for 2010
= € 68 per MWh
80
70
• The average price of 2010 volumes was ~€55 in
2010 hedging
90
Average price in 2010 handover
period = 55 €/MWh
outright power at EET’s CE book is ~€681
60
50
40
Jan-07
105
• As of June 2010 the average achieved price for
Transfer margin
negative one at MU Central Europe
2010 handover period
Jul-07
Jan-08
Jul-08
• The average price of 2011 volumes was ~€70 in 2008
2011 hedging
95
(handover period) => transfer price
85
75
Average price in 2011 handover
period = 70€/MWh
Average achieved
Price for 2011
= € 59 per MWh
45
Jan-08
• As of June 2010 the average achieved price for
outright power in EET’s CE book is ~€591
Transfer margin
65
55
• Currently a positive transfer effect at EET and a
• Currently a negative transfer effect at EET and a
2011 handover period
Jul-08
positive one at MU Central Europe
Jan-09
Jul-09
Depending on the time of the handover transfer prices may turn out
to be higher than average achieved prices
1. For outright power hedging please refer to slide 8
31
E.ON transfer price - Setting a price for optionality
Value
Visualization of intrinsic and extrinsic value
Time value
(Extrinsic value)
Spread
(Intrinsic value)
Extrinsic and intrinsic value
• E.ON transfer price mainly consists of two elements
• Intrinsic value: clean spread based on market
forward prices (as on previous slide)
• Extrinsic value: time value of the real option based
on changes of market data (e.g. price volatility) and
plant characteristics
Price
Marginal cost
Variable
Price/costs
Power price Forward price
Probability distribution of
future power prices (after
convolution for uncertainties)
Capacity Price
– Trading pays a price for the time value of the
capacity
– Value consists of the right (not the obligation) to
exchange fuel for electricity (make or buy)
Operating hours
of the power plant
– For a nuclear power plant the extrinsic value is
basically zero
– For the marginal plant of a system it is very high
Market price for fuel + CO2
(For gas plants: contract price)
Time
Capturing the value of a flexible generation fleet
32
Example: make or buy strategy
… for make or buy
Two simplified examples…
Example 1 - buy (in €/MWh)
t1
t2
Locked in clean dark spread
New clean dark spread (spot)
10
10
-6
Make (net result)
10
Buy (net result)
16
Example 2 – make (in
€/MWh)
Locked in clean dark spread
New clean dark spread (spot)
Make (net result)
Buy (net result)
t1
t2
10
10
6
10
• Example 1:
If the real option is out of money at delivery,
additional value can be generated above the lock-in
price in forward
• Example 2:
If the real option at a lower spread than hedged
but in the money the decision would be to deliver
the physical product as hedged
4
Extracting the maximal value from flexible power plants
33
Example: Time spread arbitrage in practice
Simplified example…
…for time spread arbitrage
• It enables E.ON to profit from the shape or trends
€/MWh
in the forward curve and exploit flexibilities in
storage & contracts:
30
Forward curve in Contango
25
20
Spot
Quarter 1
Quarter 2
Selling 1TWh in Q1 has a value of € 25m
– Nordic Hydro: the flexibility of the hydro-power is
based on the storage possibilities in the
reservoirs
– Gas storage: Trading manages several storages
around Europe with flexibility in selling gas
Selling 1TWh in Q2 has a value of € 30m
Hedging financially and postponing the physical production
from Q1 to Q2 will create a profit of €5m 1
– Take or Pay contracts: Trading manages several
contracts with flexibility in take-volumes
Creating value from the flexibility in storages and supply contracts
1. Simplified – disregarding the time value of money
34
Investor Relations
E.ON Investor Relations Contact
Sascha Bibert
Head of IR
Peter Blankenhorn
Manager
François Poullet
Manager
Marc Koebernick
Manager
Dr. Stephan Schönefuß
Manager
Aleksandr Aksenov
Manager
Carmen Schneider
Manager
Sabine Burkhardt
Executive Assistant
T +49 2 11-45 79-5 42
[email protected]
T +49 2 11-45 79-4 81
[email protected]
T +49 2 11-45 79-3 32
[email protected]
What can we do to
help you?
T +49 2 11-45 79-2 39
[email protected]
T +49 2 11-45 79-48 08
[email protected]
T +49 2 11-45 79-5 54
[email protected]
T +49 2 11-45 79-3 45
[email protected]
T +49 2 11-45 79-5 49
[email protected]
35
Investor Relations
E.ON IR and reporting calendar
Date
Event
Location
March 14, 2012
Annual Report 2011
Düsseldorf
May 3, 2012
AGM 2012
Essen
May 4, 2012
Dividend payment
May 9, 2012
Interim Report I: January – March 2012
Düsseldorf
August 13, 2012
Interim Report II: January – June 2012
Düsseldorf
November 13, 2012
Interim Report III: January – September 2013
Düsseldorf
36
This presentation may contain forward-looking statements based on current assumptions and forecasts made
by E.ON Group management and other information currently available to E.ON. Various known and unknown
risks, uncertainties and other factors could lead to material differences between the actual future results,
financial situation, development or performance of the company and the estimates given here. E.ON AG does
not intend, and does not assume any liability whatsoever, to update these forward-looking statements or to
conform them to future events or developments.
37