Uniform-price back-test (illustrative)

If IEX, PXIL and HPX day-ahead bids had cleared together, 1 Jan 2026 to 31 Mar 2026.

Assumptions

  1. ILLUSTRATIVE ONLY. Not a forecast, not a reconstruction of any exchange's order book.
  2. Window: IEX, PXIL and HPX DAM, 1 Jan to 31 Mar 2026, the months for which the Atlas holds all three exchanges' daily figures (CERC Market Monitoring Cell monthly reports).
  3. Aggregate bid curves are not published in a form the Atlas holds, for any exchange. Each exchange's demand curve is a straight line from its total purchase bid at Rs 0 to its cleared volume at its clearing price, then to zero at the Rs 10,000/MWh ceiling; its supply curve runs from zero at Rs 0 to its cleared volume at its clearing price, then to its total sell bid at the ceiling. In a block that cleared at the ceiling, demand at the ceiling is the cleared volume. These curves reproduce each exchange's own clearing exactly; their shape away from the clearing point is assumed.
  4. IEX: per-block purchase bid, sell bid, cleared volume and clearing price (All India), as published.
  5. PXIL and HPX: only daily cleared volume and volume-weighted price are held. Each day's volume is spread over the 96 blocks in proportion to IEX's cleared volume; each block's price is IEX's block price scaled so the day's volume-weighted average equals the exchange's published daily price; bids are set at IEX's block bid-to-cleared ratios.
  6. Market splitting: no transmission congestion data is used. IEX's own area prices for its 13 bid areas carry the transmission constraints IEX saw; each area's coupled price is the coupled uniform price plus that area's IEX premium over the IEX All India price in the block, kept within Rs 0 to Rs 10,000.
  7. No cross-exchange arbitrage, bid re-optimisation or change in participation is modelled: the same bids clear together. Real coupling would change behaviour.

What the data does not hold

  • No PXIL or HPX block-level bids, volumes or prices: daily figures only (see the shaping rule above).
  • No aggregate demand or supply curves for any exchange: the curve shape is assumed.
  • No CERC Market Monitoring Cell daily data after 31 Mar 2026 in the Atlas, so the window stops there.
  • No corridor-level transmission congestion data: IEX's 13 area prices stand in for it.
  • In the window PXIL cleared DAM volume on 46 of 90 days and HPX on 4; on the other days only IEX takes part.

Results (illustrative)

  • 90 days, 8,640 blocks. The coupled price sits ₹3/MWh above IEX's on average, block by block (largest gap ₹68/MWh): IEX carried almost all the volume.
  • On the 46 days PXIL cleared DAM volume, the coupled daily average was ₹5,703/MWh below PXIL's.
  • On the 4 days HPX cleared DAM volume, it was ₹5,779/MWh below HPX's.
  • Blocks at the ₹10,000/MWh ceiling: IEX 709, coupled 709. Blocks in which IEX split into areas: 43.
Monthly averages of the daily volume-weighted DAM prices, illustrative
MonthIEXCoupledPXIL (days)HPX days
January 2026₹3,852/MWh₹3,855/MWh₹10,000/MWh (18)1
February 2026₹3,547/MWh₹3,552/MWh₹10,000/MWh (13)1
March 2026₹3,925/MWh₹3,927/MWh₹9,571/MWh (15)2

Code and data

The code, the data snapshot and a container are public at github.com/India-Energy-Atlas/market-coupling-backtest (opens in a new tab) (Apache-2.0), commit 3c9204b2 (opens in a new tab). One command reproduces every figure on this page: docker compose run --rm backtest. Back to the Market coupling tracker.