MODELLING & MANAGING FULL/RESIDUAL SUPPLY GAS CONTRACTS IN CONTINENTAL MARKETS Padova, 10 Maggio 2013 Stefano Fiorenzani ALEPH - Director & Founder – ALEPH CONSULTING ENERGY – RISK – FINANCE
What is Aleph? • Aleph is a project created by Stefano Fiorenzani in order to provide Energy Companies, Consulting Firms and Financial Institutions specialized services in the area of Energy Trading and Risk Management. • Aleph ’s particular focus is on training and consulting quantitative methods applied to Energy Finance. • Aleph ’s strength is the professional and theoretical experience of its founder and team. • Aleph is in contact with other business entities able to integrate its services with pure IT, Legal, Accounting, Fiscal and Commercial aspects. 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 2
Full/Residual supply gas contracts • Full supply contracts (or Full Serve/Full Requirements) are typically structured to satisfy supply needs of big/medium/small size energy consumers. FS are highly structured deals. Only the general framework is common for different markets. • Many different contract features can characterize different markets and/or specific sectors (eg. Industrial large consumers, Distributors, Municipalities, SME etc). In FS contracts the writer (seller) has the obligation to fully satisfy the consumption need • (demand) of the buyer for the payment of a monetary price defined per unit of consumed volume (€/MWh). • Contract price may be a fixed price or an indexed one. Indexed price contracts may be subject to fixing prior to delivery. • Contract’s buyer should provide indications about future consumption needs based on consumption history. Nomination rules may apply in order to segregate volume flexibility from pure unbalancing (a different pricing may apply). 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 3
Full/Residual supply gas contracts Payoff Structure: � 𝑄 𝑢 − 𝐺 𝑢 𝐺𝐺 = � 𝑀 𝑢 + Δ𝑀 𝑢 𝑄 𝑢 + 𝑞𝑞𝑞𝑞𝑞𝑞𝑞 − 𝐶𝐺 𝑢 𝑢∈Ψ � 𝑄 𝑢 − 𝐺 𝑢 𝐺𝐺 = � 𝑀 𝑢 + Δ𝑀 𝑢 𝑄 𝑢 + 𝑞𝑞𝑞𝑞𝑞𝑞𝑞 − 𝐶𝐺 𝑢 𝑢∈Ψ spot price unbalancing penalty delivery period d-1 nominated contractual price unbalanced load unbalancing price (forecastable d-1) load 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 4
Full/Residual supply gas contracts Residual supply contracts are typically • structured to partially satisfy supply needs of big size energy consumers which have already contracted a part of its supply. The seller of RS contracts are required to • complement and balance (proportionally or RS globally) the consumption demand of the buyer net of a previously contracted supply which could be a ToP band, a ToP Profile, an ToP Band other RS contract. RS Contractual Scheme FS The RS contract’s buyer shall provide • (DP) the seller full evidence of previously contracted supplies. Supplier Consumer RS Supplier • RS contracts can have different contractual schemes according to counterparty nature (shipper/non ToP ToP shipper) and distribution rules. (Hub) (Hub) 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 5
Full/Residual supply gas contracts Payoff Structure: � − 𝐺𝑀 𝑢 ) 𝑄 𝑢 − 𝐺 𝑢 𝑆𝐺 = � ( 𝑀 𝑢 𝑢∈Ψ + Δ𝑀 𝑢 𝑄 𝑢 + 𝑞𝑞𝑞𝑞𝑞𝑞𝑞 − 𝐶𝐺 𝑢 Volumetric flexibility and balancing service apply to the whole Load supplied by consumption => proportional impact alternative supplier of risk premium and balancing service is much higher than in FS contracts. 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 6
Full/Residual supply gas contracts Physical Constraints (potentially present): Capacity Constraints => working at daily level (Max/Min MWh/d), mirroring the physical capacity of the transportation facility. Max Capacity Max Energy Min Energy Min Capacity Daily Consumption Cumulated Consumption Energy Constraints => working at yearly/season/quarter level (Max/Min MWh/period), aiming to constraint volumetric flexibilities. If they are present they should be supported by appropriate penalty structure (no pure indications). Make up /Carry Forward => linked to energy constraints, pretty infrequent in non Hub delivered contracts (swings). 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 7
Full/Residual supply gas contracts Sell Back Options (potentially present): The main difference between volumetric flexibility sold to a wholesale trader (with a swing • contract typically) or to a final consumer (industrial or retail) is related to the delivery point and hence to the opportunity to re-sell gas in the market instead of consuming it. • This opportunity present in swings and not typically in FS allows contract owner to optimize and extract value from contact’s flex otherwise impossible to extract. • In particular cases also in FS contracts sell back options can be present: Within portfolio construction schemes + RS contracts => opportunity to buy and sell back portfolio building blocks (only) Portfolio Construction Framework before delivery. Complete sell back opportunity up to d-2 The wider is the sell back opportunity (in this case FS become almost a pure the higher is the contract’s value. swing despite the delivery point). 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 8
Framing the pricing problem FS contracts are non-perfectly hedgiable products, hence the pricing problem should be tackled in different steps in order to obtain a coherent even if subjective valuation: 1. Construct (typically by simulations) the complete conditional distribution of contract’s payoff structure (conditional distribution of the “naked position”); 2. Determine the contract price which sets the conditional expectation of contract’s payoff to zero (“break even price”); 3. Determine the optimal “static” or “dynamic” hedging strategy which minimize overall risk (at deal level) and construct the complete conditional distribution of “hedged position” (remember that we usually work under the assumption that hedging strategies with linear instruments does not contribute to overall expected payoff); 4. Determine the subjective minimum risk remuneration (“risk premium”), eventually disentangled in different risk components such as pure volumetric risk, unbalancing risk, liquidity risk etc. 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 9
Framing the pricing problem FS Payoff Simulation � 𝑄 𝑢 − 𝐺 𝑢 𝐺𝐺 = � 𝑀 𝑢 + Δ𝑀 𝑢 𝑄 𝑢 + 𝑞𝑞𝑞𝑞𝑞𝑞𝑞 − 𝐶𝐺 𝑢 𝑢∈Ψ • The main challenge is determining and implementing a realistic and robust joint dynamics of the triplet of stochastic processes [L(t), S(t), P(t)] (we can also reduce the triplet to a couple price/ load [L(t), Prices(t)] with Prices(t)= [S(t), P(t)] ). • The designated model should be able to correctly represents individual dynamics but also joint behavior (dependence structure). • The designated model should be consistent with observed forward price structures and with historically measured load paths (or forecasted ones). • Two alternative modeling approaches: semi fundamental models, reduced form models. 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 10
Framing the pricing problem Un-hedged Payoff Distributions The more load and price are correlated in their movements the more FS payoff structure becomes non-linear wrt price ( form y=-c*x to y=c*x-x^2 ) 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 11
Framing the pricing problem Un-hedged Payoff Distributions Load-Price correlation induce non linearity in the payoff structure as can be reflected also in distribution variance and asymmetry. -5 Payoff Distribution x 10 1.2 1 0.8 0.6 0.4 0.2 0 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 5 x 10 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 12
Framing the pricing problem Optimal hedging in a static and dynamic framework FS contracts cannot be perfectly hedged (daily granularity, volume risk), nevertheless the better is the hedge the lower is the payoff risk (left tail of payoff distribution). When we face the pricing problem we should necessarily consider hedge possibilities and effectiveness (we can’t price the contract based on payoff expectations!) Static Hedging Problem (hedging with respect to static payoff expectation) � 𝑄 𝑢 − 𝐺 𝑢 𝐺𝐺 = 𝑛𝑞𝑛 𝑄 0 , 𝜄 0 , 𝑢 𝐹 𝑉 � 𝑀 𝑢 + Δ𝑀 𝑢 𝑄 𝑢 + 𝑞𝑞𝑞𝑞𝑞𝑞𝑞 − 𝐶𝐺 𝑢 + 𝜄 0 , 𝑢 𝐺 0, 𝑞 − 𝐺 𝑢 𝑢∈Ψ determine optimal contract pricing (P0) and hedging ( θ (t)) which maximize expected utility ?? - determine your utility function (risk charged expected profit?) - avoid trivial problems - consider the market (liquidity, maturity, bid-ask spreads…etc) 11/05/2013 Stefano Fiorenzani - Aleph Consulting - 13
Recommend
More recommend
