…a simple discussion of market making and how volatility and the lack of market makers lead to $0.01 prices
For many years I was a market maker in the commodities space, specifically in the electricity markets. Producers, consumers, hedgers, and speculators all would contact me for quotes on numerous locations and forward delivery dates, which often enough were not quoted elsewhere. What’s a good sell side market maker to do? You make a price and hope your curves are close enough that you don’t get killed. How did I formulate a market for a utility customer looking to buy, lets say for example, 2011 delivery for the Month of May peak hours into Boston? To be honest, a lot of boot strapping and prayers that correlations will hold up.
Power for delivery in Boston, is a basis to the primary regional market for New England, in this case called MassHub. MassHub is often priced as a spread to New York – Zone G (essentially the Hudson valley) or the Mid-Atlantic market known as PJM. Power in Boston is also heavily correlated to the delivered price of gas in the city known as Algonquin (which is itself based on other gas locations). The price for May 2011 is also closely associated with the price of May 2010, April 2011, and June 2011, and to a lesser extent, May of 2012. Markets are usually most liquid in the front of the curve, so everything starts there. You can easily see this by looking at trade volumes for any futures contract, and see how the volume drops considerably after the first few months of delivery. These other similar markets allow me to “bootstrap” a curve from the relative value versus other products.
From a combination of cross commodities spreads and ratios, combined with historical spread settlements, traders gain confidence that when they sell power for Boston delivery in May of 2011, they can go out, buy some PJM for May 2010, buy a NY-G May 2010/2011 time spread, and perhaps buy a similar amount of Q2’11 NYG/MassHub spread. Add in some gas basis spread to get from PJM to NY-G and perhaps a power spread to get from MassHub to Boston, and you are nearly home. Seems like a lot of trades to make a deal, but it’s these kind of market makers and spread traders that support a huge marketplace with several dozen locations across 15-20 years of delivery. Does it make sense that 2010 June natural gas delivered to Houston is the basis for 2022 power delivered to Cincinnati? Not really, but through dozens of traders spreading across regions, times, and products, the market works as a giant interconnected web of spreads all holding each other up.
At the end of the day, traders “Mark the Book”, and by that I mean they enter their estimates for market curves several years out, on a month by month basis, for often several dozen locations. Are there visible quotes on the screens or exchanges or in the OTC broker market to validate these curves? Often not. Most of the exchanges and brokers rely on yesterday’s spread differences or spread markets.
What happens during times of volatility when some of the spread traders pull their markets or widen them considerably across all of the spreads? Bid/offers become scarce. If someone has to sell, because they were tapped on the shoulder or something changed in their other positions, there may not be a “rational” market. If someone had to sell something at “market”, and there was no one willing to spread into the position, then technically there may not be a bid at all. Rationally, someone should have bid for Accenture against a spread to the futures or some other stock, but those spread offers were not there. Sounds a lot like of what happened yesterday when the NYSE was on “pause” and the other exchanges found themselves with no rational bids.
Market fundamentals, price momentum, historical cash settles, the size of the order and the nature of the counterparty all clearly contribute to any price, but at any instant, the current prices for associated markets matter the most. If the price of the first delivery month of natural gas disappeared, I would not be able to calculate any ratios of power to gas and most likely not be able to get any other power prices to spread to. Lacking that information, I might bid $0.01.