1 The effect of over-allocation and price uncertainty on investments under the EU ETSFrank Venmans Frank Venmans | Université de Mons | Grantham Institute-LSE
2 Introduction and relevanceEU ETS= Cap and Trade covering 45% of European GHG emissions ETS lauded for its cost-effectiveness: equal marginal abatement costs across firms Only if rational inclusion of ETS-incentives in investment decisions Over-allocation vs under-allocation Until 2012: electricity under-allocated, rest over-allocated From 2013 for carbon leakage exposed sectors: benchmarked allocation (emissions of 10% most efficient plants). Price uncertainty > 25 € in 2008, 5,8 € yesterday Recent decision to install a reserve Frank Venmans | Université de Mons | Grantham Institute-LSE
3 «Standard » and behavioural economicsStandard theory Coase theorem predicts that over-allocation and under-allocation create the same incentive to invest: When over-allocated an abatement investment yields extra sales opportunities When under-allocated an abatement investment results in less purchases of permits. Pro-cyclical gains are risky, pro-cyclical costs are risk-hedging Option value Behavioural economics Add insight from psychology to describe decision heuristics in a more accurate way Policy more efficient when behavioural insight on reaction on incentives are added Reference-dependent preferences => Loss aversion => Endowment effect Narrow framing Frank Venmans | Université de Mons | Grantham Institute-LSE
4 Sample & methodology 19 semi-directive interviews of 1h15 to 2h3 past investments, 2 potential future investments Focus on barriers and triggers to investment 16 (out of 18) Belgian producers of construction materials in EU ETS: Bricks (9) Cement (2) Lime (2) Other (3) Sector represents 15% of CO2 emissions of EU ETS Frank Venmans | Université de Mons | Grantham Institute-LSE
5 Results Over-allocation vs Under-allocationEffect of price uncertainty Frank Venmans | Université de Mons | Grantham Institute-LSE
6 3 ways of expressing lower incentive to invest when over-allocatedAs a direct question: Only 2 managers perceived over-allocation to create the same incentive as under-allocation (~Coase) 1 perceived under-allocation to create a lower incentive (~option value) 11 who perceived under-allocation to create a stronger incentive Under-allocation was the reason why managers perceived the ETS to be a greater motivation for future investments compared to past investments Under-allocation was their reason why all gas-fuelled companies omitted carbon gains in their payback times carbon advantage increase profitability by 12% Frank Venmans | Université de Mons | Grantham Institute-LSE
7 Rationale behind perceptionsManagers explain higher motivation of under-allocations by reference-dependence (cost higher value than gains) 3 reference points: Situation without ETS Initial allocation-endowment effect Competitors Frank Venmans | Université de Mons | Grantham Institute-LSE
8 1st Reference: without ETS“In a grandfathered ETS system, growth is absolutely sanctioned. Every marginal production unit you produce above your historical capacity, you will not get extra credits. One believes that people think in mean values, but no, we think in marginal values, of course. So you get to the point where you produce until you have produced your freely allocated level. And then every extra ton that you produce is entirely submitted to the carbon cost.” Classical view: grandfathered emission rights are a lump sum transfer without marginal effect. When under-allocated the ETS creates a loss, which attracts more attention than a gain. Frank Venmans | Université de Mons | Grantham Institute-LSE
9 2nd reference: initial allocationThe endowment effect creates a lower Willingness To Accept to sell allocated permits than the Willingness To Pay for extra permits. => Extra permits from energy efficiency are not cashed “Cash is King” When over-allocated, the ETS is seen as an environmental regulation with which they comply and thus to which no attention is required. “When you save gas, you emit less CO2 and those rights will remain unused and at a given moment they can be sold?” “I suppose that is the case, but we didn’t consider it at all, because the carbon aspect is dealt with by somebody who is closer to the accounting department. It is disconnected from the one who is dealing with production... I think that for us this is somewhat too abstract.” Frank Venmans | Université de Mons | Grantham Institute-LSE
10 3rd reference: competitors“In plant X the situation is very critical, due to the quarry. Because we are above the benchmark. So at any price, what can we do, because it will cost us money... A fundamental element in this are the production costs. The one that will survive tomorrow, because there is a terrible competition, will be the one that has the lowest costs. We are on a commodity market. We have new competitors that arrive...” Frank Venmans | Université de Mons | Grantham Institute-LSE
11 3rd reference: competitorsAccording to standard economic theory, both over-allocation and under-allocation have the same effect on the NPV of a project. The NPV equals the extra profit or reduced price. Compared to extra-European competitors over-allocation = competitive advantage under-allocation = competitive disadvantage. Managers are more motivated to reduce a competitive disadvantage than to increase a competitive advantage. Compared to intra-European competitors even more puzzling because allocation method doesn’t affect competitiveness When over-allocated, all producers gain (at least in the short run) A competitive disadvantage (low energy efficiency) is less urgent to absorb Frank Venmans | Université de Mons | Grantham Institute-LSE
12 Results Over-allocation vs Underallocation Effect of price uncertaintyFrank Venmans | Université de Mons | Grantham Institute-LSE
13 Standard financial theoryPro-cyclical gains are risky (Counter-cyclical gains are risk-hedging) (CAPM, Fama 1977) Pro-cyclical costs are risk-hedging (Counter-cyclical costs are risky) (Ariel 1998, Armitage 2005) Carbon costs are pro-cyclical By design Observed: correlation 2nd period 0.27 => volatility increases the risk-hedging potential of carbon costs => volatility decreases the incentive to invest because It increases riskyness of carbon gains when over-allocated it reduces riskyness of avoided carbon costs when under-allocated CAPM can be written as: 𝑟 𝑖 = 𝑟 𝑓 + 𝐸 𝑡 𝑅 𝑀 − 𝑅 𝐹 𝜎 𝑅 𝑀 𝜌 𝑟 𝑖 , 𝑟 𝑀 𝜎 𝑟 𝑖 => Procyclical gains are risky (Countercyclical gains are risk-hedging) 𝜌 𝑟 𝑖, 𝑅 𝑀 has the same sign for positive and negative cash flows =>Procyclical costs are risk-hedging (Countercyclical costs are risky) Risk-hedging character of procyclical costs increase with volatilityConsider the value of a claim on a future cash flow YT between t-1 and t can be written in the form of a risk adjusted discount rate or a certainty equivalent 𝑉 𝑡−1 = 𝐸 𝑡−1 ( 𝑉 𝑡 ) 𝐸 (𝑅 𝑖1 ) = 𝐸 𝑡−1 ( 𝑉 𝑡 ) 1 1+ 𝑅 𝑓 1− 𝜙 𝑡 𝜌 𝜀 𝑡, 𝑅 𝑀𝑡 𝜎 𝜀 𝑡 𝑡−1 With 𝜀 𝑡 = 𝐸 𝑡 𝑌 𝑇 − 𝐸 𝑡−1 𝑌 𝑇 𝐸 𝑡−1 𝑌 𝑇 the expectations adjustment variable This yields for several periods: 𝑉 0 = 𝐸 0 𝑌 𝑇 𝐸 (𝑅 𝑖1 ) … 1 1+𝐸 (𝑅 𝑖𝑇 ) Fama (1977) shows that if the CAPM holds for every period, 𝑐𝑜𝑣 𝜀 𝑡 ,𝑅 𝑀𝑡 ; Rft and Øt are known or non-stochastic over time. Frank Venmans | Université de Mons | Grantham Institute-LSE
14 Option value When investment leads to a slow learning process=>allows for more flexibility under potentially high carbon prices in the future => investment allows to increased option value When there is an option to postpone the investment Allows not to invest if future information reveals investment to be unprofitable Under very low carbon prices Under very high carbon prices => lost option value. Both gained and lost option value increase with price volatility Effect of price uncertainty on incentive to invest ambiguous. Was mentioned by 4 companies, especially big and long term investments Frank Venmans | Université de Mons | Grantham Institute-LSE
15 𝑉𝑎𝑙𝑢𝑒=𝑁𝑃 𝑉 𝑤𝑖𝑡ℎ𝑜𝑢𝑡 𝐸𝑇𝑆 + 𝑡 0 ∞ 𝐸[ 𝑝 𝑡 𝑞 𝑡 ] 1+ 𝑅 𝑖 𝑡 + 𝑜𝑝𝑡𝑖𝑜𝑛 𝑣𝑎𝑙𝑢 𝑒 𝑖𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝑂𝑛𝑙𝑦 𝑖𝑓 𝑙𝑒𝑎𝑟𝑛𝑖𝑛𝑔 𝑡𝑎𝑘𝑒𝑠 𝑡𝑖𝑚𝑒 𝐷𝑟𝑖𝑣𝑒𝑛 𝑏𝑦 𝑎𝑣𝑜𝑖𝑑𝑎𝑏𝑙𝑒 𝑝𝑙𝑎𝑛𝑡 𝑐𝑙𝑜𝑠𝑢𝑟𝑒 − 𝑜𝑝𝑡𝑖𝑜𝑛 𝑣𝑎𝑙𝑢 𝑒 𝑝𝑙𝑎𝑛𝑡 𝑐𝑙𝑜𝑠𝑢𝑟𝑒 𝑂𝑛𝑙𝑦 𝑖𝑓 𝑤𝑎𝑖𝑡𝑖𝑛𝑔 𝑖𝑠 𝑝𝑜𝑠𝑠𝑖𝑏𝑙𝑒 𝐷𝑟𝑖𝑣𝑒𝑛 𝑏𝑦 𝑢𝑛𝑎𝑣𝑜𝑖𝑑𝑎𝑏𝑙𝑒 𝑝𝑙𝑎𝑛𝑡 𝑐𝑙𝑜𝑠𝑢𝑟𝑒 − 𝑜𝑝𝑡𝑖𝑜𝑛 𝑣𝑎𝑙𝑢 𝑒 𝑙𝑜𝑤 𝑝 𝑐𝑎𝑟𝑏𝑜𝑛 𝑎𝑛𝑑 𝑝 𝑒𝑛𝑒𝑟𝑔𝑦 𝑂𝑛𝑙𝑦 𝑖𝑓 𝑤𝑎𝑖𝑡𝑖𝑛𝑔 𝑖𝑠 𝑝𝑜𝑠𝑠𝑖𝑏𝑙𝑒 Frank Venmans | Université de Mons | Grantham Institute-LSE
16 Contradicting perceptions“Avoiding the uncertainty of the ETS” is an important motivation to invest “Avoiding the uncertainty of the ETS” is an important motivation to invest, even more important for future investments when under-allocated. Gained option value (learning effect) is important. “What would create highest incentive to invest: narrow price channel between 14 and 16€ or volatile carbon price with expected mean of 15€” 3 companies see volatility (riskiness) of carbon costs as incentive to invest Gained option value is important Managers perceive pro-cyclical volatility of carbon costs as a risk-increasing feature = first narrow framing 9 companies see volatility (riskiness) of carbon costs as a disincentive to invest Lost option value is important for certain investments with long payback times Managers perceive pro-cyclical volatile carbon costs as a risky gains instead of risk-hedging costs =double narrow framing Confirmed by the fact that “avoiding uncertainty of the ETS” is very important and more important in the future For subset of 9 companies also inconsistent with “Avoiding the uncertainty of the ETS” as a motivation to invest almost as important as the motivation “Expected gains of cost savings from the ETS”. Frank Venmans | Université de Mons | Grantham Institute-LSE
17 Conclusion Under-allocation reduces behavioural barriers to invest.High price uncertainty has an ambiguous effect on investment incentives. Higher incentive for investment related to slow learning processes Lower incentive for investments with long payback times Lower incentive for investments that cannot be postponed, but this effect is sometimes reduced by narrow framing. Frank Venmans | Université de Mons | Grantham Institute-LSE