1

Situation

I am modeling an auction-based electricity market. Lets assume they call an auction for 500MWh, and participants can submit bids of at least 1MWh for a given price of $x$ EUR/MWh. Each participant may submit as many bids at as many prices as they want. There are no max and min (altough below some prices it would not make sense to submit a bid) prices. The auction is sealed-bid (participants dont know what others are bidding), and the auction is paid-as-bid, that is, every participant receives their submitted price. These auctions are done every few days, and will be done for many years. Each auction may achieve different prices.

Now when an auction is over, $n$ bids are accepted such that the total volume contributed by these bids adds up to the required 500MWh. These $n$ bids have different prices, however we can compute the average price $\bar x$ that was achieved.

Question

Now assume we know the distribution of this average price $\bar x$. For example lets say its a normal distribution with mean 100€ and std 50€. Over 1000 auctions, the average price will thus be close to 100€.

With this knowledge, can I assume that there is some bidding strategy that allows me to bid in this auction and on average achieve a price of 100€ with a reasonable volume (e.g. up to 10% of the volume?). To phrase it differently, would it be possible for me to bid in a way that my bids dont greatly affect the average price achieved? Or at least affect them in a way that is predictable, e.g. my average achieved over many runs would be 100€ minus some constant or relative value.

Notes

  • The reason I am asking this is because I am modeling this market, and the only data I have are the average price per MW achieved on it per auction. From this I can formulate a distribution and get an idea of what the price in this auction should look like. So if this were a stock market, I could very easily model the distribution of prices I can achieve. I am wondering if I can do similar assumptions with these sealed-bid pay-as-bid auctions.
  • I am new to this area and if there are any issues with my question please leave a comment I am happy to elaborate
  • Is this the best stackexchange for such as question? Or should I rather try the quantitative finance or mathematics stack exchanges?
  • Lastly, I would greatly appreciate any recommendations of literature (books or papers) in this area or describing this or a similar problem. I could not find much by myself.
charelf
  • 111
  • 2
  • 1
  • Now assume we know the distribution of this average price $\bar{x}$. This is a bit unclear. Are you assuming this is the disribution of the average of all bids, no matter how you yourself will bid, or the distribution of the average of the other bids?
    • – Giskard Mar 04 '22 at 04:44
  • 1
  • As this auction is repeated regularly, it seems unrealistic to assume that if you start to win with your 50MW all the time other actors will not adapt their own strategy.
    • – Giskard Mar 04 '22 at 04:47