AI in Energy Trading and Electricity Markets Quiz

Answers and explanations appear as you take the quiz.

1. Why do locational marginal prices differ from one grid node to another?
   • Each utility sets its own retail rate
   • Generators choose different currencies
   • Transmission congestion and losses make power costlier to deliver to some locations
   • Nodes closer to the equator use more air conditioning
2. What is the main purpose of intraday and real-time or balancing markets?
   • To correct for forecast errors after the day-ahead market clears
   • To set long-term fuel contracts
   • To replace the day-ahead market entirely
   • To decide where new power plants are built
3. Why is deciding when to discharge a battery a tricky trading problem?
   • Batteries cannot sell ancillary services
   • Battery output cannot be measured
   • Discharging is banned during high prices
   • The battery holds limited energy, so discharging now means giving up a possibly higher price later
4. What does quantile regression provide for price forecasting?
   • A single exact price for each interval
   • Percentile estimates such as the 10th, 50th and 90th, which describe the range of uncertainty
   • A list of market participants
   • The battery's round-trip efficiency
5. How can a model with lower mean absolute error earn less money than a less accurate one?
   • Lower error always means more profit, so it cannot
   • Markets pay only for daytime forecasts
   • It can get the timing of the daily peak wrong, and that is where most of the revenue is
   • Mean absolute error only measures demand
6. What mistake do properly built backtests avoid?
   • Using information that was not available at bid time, such as data published after gate closure
   • Using historical market data at all
   • Including battery efficiency losses
   • Testing on more than one year of data
7. Which EU agency oversees the REMIT rules against energy market manipulation, together with national regulators?
   • FERC
   • ERCOT
   • PJM
   • ACER
8. What does the guide say about predicting rare scarcity price spikes?
   • Neural networks now predict them perfectly
   • No model reliably predicts them, so good systems manage the risk instead
   • They are caused only by fuel prices, so they are easy to forecast
   • They never happen in markets with batteries