A solar farm with a contract to sell electricity is almost the lowest risk investment a pension fund can make. Only index-linked government bonds beat its reliability. As long as electricity prices are contractually secure, PV resembles nothing so much as an annuity, a guaranteed flow of money that arrives every month for 25 years. Gradually, the financial markets are realising what a superb asset PV can be, particularly for pension funds needing to match long streams of liabilities. As a result, the cost of capital for solar farms is falling. It needs to fall much further. It cannot be stressed enough how important this is. This week’s IEA report, which finally ended the Agency’s long-standing contempt for PV and large scale solar thermal technologies, put it bluntly. At today’s prices and construction costs, PV produces electricity in good global locations at around 6 US cents per kilowatt hour if interest on the capital is assumed to be 0%. Assume the cost of capital is 9% per year and the cost more than doubles to over 12 cents. At this level, interest payments are more than half the cost of solar PV.

In a sunny Cornish field you could add 15% to these costs. There, PV will generate at about 7 cents at 0% cost of capital and 14 cents at 9%. The implications of this are striking. Push the cost of financing solar down close to zero and PV is already competitive with grid electricity in the UK. (7 cents is not much over 4 pence per kilowatt hour). At 9% you still need heavy subsidy to match fossil fuels.

What is the return on capital that investors demand today for PV? DECC’s latest estimate is 5.3% in real terms, or 7.3% nominal if inflation of 2% is added. This is the figure also used by other government bodies such as the Committee on Climate Change. It’s worth pointing out that this is higher than regulators assume for electricity generation using fossil fuels. Ofgem says the cost of capital for the Big Six generators is around 6% real. The difference is inexplicable. Ask yourself which investment you would rather own. A £100 share in a fossil fuel operator facing unknown carbon taxes, fluctuating fuel costs, sharply varying electricity prices and divestment campaigns growing in strength by the day, or £100 in a PV farm guaranteed a stable price for its output?

At the current interest assumed by DECC, the IEA’s figures suggest a Cornish cost for PV electricity of around 9 pence per kilowatt hour for the raw cost of power pumped into the local grid. Transmission charges will add to this figure, meaning the full cost is probably around 10.5 pence, about double the wholesale price of electricity. [1]

Travel from Cornwall to central Germany, which gets about the same amount of sun, and things are very different. There, says the Fraunhofer Institute, the cost of capital (expressed in real terms) is 2.8%, or 4.8% with inflation added in. Nothing else has changed but solar electricity immediately costs about 20% less in Germany than it does in the UK.

Why should this be? Very long dated government bonds (gilts) do yield about 1.2% more in the UK than Germany. So investors deciding whether to put pension fund money into gilts or PV will demand that PV gives them more than in Germany. Nevertheless there’s still a striking gap between the negative (i.e less than zero 0%) yields on inflation-protected government bond yields in the UK and the 5.3% real cost of capital assumed by DECC for solar PV.

I guess that the 5.3% estimate is actually too high. The small number of publicised sales of stakes in solar farms do suggest a lower figure. Lancashire County Council pension fund put £12m into the bonds of Westmill Solar Farm at a real interest rate of 3.5% in 2013. Just launched today,, Oxford’s latest £1.5m fund-raising effort to put PV on all its local schools opened for subscription offering 5% nominal returns, over 2% less than the government estimate.

It needs to go further, and not just for the sake of emissions reduction. The UK and other countries are facing increasing problems from the impact of low interest rates on pension liabilities. This week the UK universities pension fund, the largest in the UK, announced a strike ballot over plans to cut entitlements. The principal cause of its deficit is low or negative real interest rates. It has many billions of pounds now looking for a safe home that yields slightly more than the negative returns available on inflation protected gilts.

Solar PV, done at the large scale pioneered in Germany, could provide exactly the right form of asset for this fund. A sensible industrial strategy would be putting 5 gigawatts of PV on to brownfield land in the south west and funding it with pension fund money at 2 or 3% real return. I know that this won’t happen but it really is a much better idea than fracking most of Sussex and Lancashire.

(Boffins wanting to experiment with numbers on the importance of the cost of capital to renewable technologies may like to play with the US government online calculator at http://www.nrel.gov/analysis/tech_lcoe.html . Remember that you can use your own currency and don’t need to convert into dollars since the calculator is using percentages rather than absolute amounts).

[1] DECC says UK PV costs around 12 pence per kilowatt hour at a real discount rate of 5.3%. I believe it uses figures for construction cost of large scale solar farms that are too high, explaining the difference between my numbers and the Department’s.