Comments on: Market coupling does not lower prices!

By: Georg Zachmann

Georg Zachmann — Wed, 03 Oct 2012 16:38:06 +0000

@Jon Griffiths: thank you for your comments. I agree that the methodology is not super-sophisticated and we are working on some formal testing. This being said I believe the polynomial approximation is quite convincing (see Figure 3). In fact, the point-clouds for the Franco-German example 2008 confirm that at some high loads average prices are lower (see http://www.flickr.com/photos/88023334@N06/8050827316). For 2011, when formal market coupling has been in place between Germany and France, the picture does not look so different (www.flickr.com/photos/88023334@N06/8050821613).
Finally, using bid-offer curves would be splendid. Could you tell me where I might get access to those?

By: Georg Zachmann

Georg Zachmann — Wed, 03 Oct 2012 16:37:11 +0000

@JONAS: thank you for this thought. I actually informally tested this hypothesis in response to a similar comment I got (http://www.bruegel.org/nc/blog/detail/article/898-market-coupling-does-not-lower-prices). I now believe that your point is right. In the given example (Franco-German 2010) much of the missing effect of market coupling on prices can be due to renewables driving down prices in Germany and at the same time causing export congestion. In fact, when controlling for German renewables production (by subtracting them from the total load), price differentials systematically coincide with higher average prices (www.bruegel.org/uploads/RTEmagicC_120982_p1.jpg).

Still, I think that this is not the end of the story and more research is needed:
1) as I demonstrate in Figure 6 it is theoretically possible to have higher average prices through market coupling.
2) the effect seems to be present for other market combinations than the French-German and other points in time, in which renewables play a less dominant role.
3) even when controlling for renewables in the Franco-German example I am tentatively getting a large area (from 50 to 150 GWh) for which the price convergence does not coincide with lower prices

By: Jon Griffiths

Jon Griffiths — Wed, 03 Oct 2012 11:53:51 +0000

An interesting analysis.
I find the 5% definition as well as the polynomial approximations a little curious, and would also be keen to see an analysis based on data for periods where coupling was in force.

To investigate the effects of cable capacity constraints, why not work directly with the bid-offer curves from the hourly auctions in the various markets? This way you could choose your own capacity.

By: Jonas

Jonas — Tue, 02 Oct 2012 09:20:06 +0000

Thanks for this very interesting analysis – and for sharing these preliminary results with us. Getting a real discussion going about the promises and perils of electricity market integration is certainly needed to make it work in practice. I have some trouble with our headline though, since “market coupling does not lower prices” seems to imply some sort of causality, which you do not demonstrate. As you acknowledge yourself, market coupling is merely not correlated with lower prices.

In fact, it has me wondering whether you are not encountering an endogeneity problem. Presumably the reason why the two markets are decoupled is that there is too much congestion. So it may, for example, be the case that Germany is experiencing a lot of windy and sunny days, but cannot “share” its very low wholesale price with France as there is not enough transfer capacity between the two countries.

If there was sufficient capacity available, coupling should in most situations lead to lower prices, just as you explain at the beginning of your article. Thus, one may not want to not blame coupling, but rather the insufficient transmission capacity for your counterintuitive findings. To make a long story short, I second your call for further research!