Comments on Damsgaard et al. ”Exercise of Market Power in the

Comments on Damsgaard et al.
”Exercise of Market Power in the
Nordic Power Market”
Olli Kauppi
HSE, HECER
Overview
• Did power companies exercise market power in
the Nordic market in 2001-06?
• A study of price-cost margins in the Nordic
market in 2001, 2002-03 and 2006
• Water values computed using ECON BID – a
detailed market simulation model
• Comparison of simulated price and output paths
to the observed behavior
• Simulated prices generally below observed
prices -> no conclusive evidence of market
power
Main comments
•
Observed prices almost always below simulated. Explanation?
–
Actual hydro use higher than simulated during the sample periods
•
•
•
–
Costs of thermal sector overestimated
•
•
•
From which periods does this water come from?
Extend sample periods: there must be scarcity sometime
Report total hydro and thermal output
Possible cause: use of average availabilities
Test by plotting predicted prices using observed thermal generations
levels
The Four-Week-Perfect-Foresight
–
–
Leads to more storage in the Fall and to emptier reservoirs in the
Spring
Could you compute inflow expectations weekly and the other
expectations monthly?
• Presentation of the results:
– Figures 7,9,11: observed demand = total
generation + net trade
– There can’t be much difference in model
predictions and observed outputs – the only
difference is the amount of trade
– Hydro/thermal division more interesting than
country level total outputs
– Replace figures with comparison of actual and
predicted hydro output paths
– Add a figure on reservoir levels (at least for
2002-03)
• Presentation of the results II:
– averaging mark-ups over the sample periods seriously
underestimates the margins (and overestimates model accuracy)
in the periods when market power actually kicks in
– report only average positive (or negative) mark-ups, average
absolute deviations or concentrate on changes in profits
– report also some statistics on the hourly prices
• Interpretation:
– intro: ”lasting and large price-cost mark-ups would be a strong
indication of market power”
– results: 20-50% mark-ups for many months
– how is ”lasting and large” defined
– especially if combined with overestimated marginal costs
– if mark-ups can be explained by differences in expectations, then
you should experiment with the expectations
Minor comments
• Discuss the benefits of the DCR/DCS approach
• Is there a reference for the method?
• The reservoir space is confusing:
– is there a single reservoir or several reservoirs (that you keep
track of)?
– how does the distribution of water values relate to this issue?
– how is the distribution of water values reflected in area prices?
• Discounting?
• Minimum level of reservoirs?
• What kind of stochastic process do you assume for fuel
prices, nuclear availability, demand etc.?