For once not a government handout - orders of magnitude more please.
Andrew don't be so sceptical, this is a good idea and needs to be massively more than £50m.
Bottom line is VCs are more effective than government quangos at selecting and supporting the successful start-ups this country needs. Co-investing is a force multiplier whereas subsidies are a black hole. More detailed explanation below for the freakonomics readers among you.
VCs may lose money on most investments, but make it exceptionally big on a few. Studies differ dramatically and suffer from timing and survivor bias, but even in the USA, the most competitive VC market out there, the average returns for VC investment funds is meaningfully above stock market return, in the order of 15% (e.g. http://faculty.chicagobooth.edu/steven.kaplan/research/kaplanlerner.pdf) . This is explainable from an investor behavioural bias point of view, but tragic for both the UK economy and for company founders seeking funding. It means that there is a whole spectrum of investments with on average good solid positive returns that are not being made due to lack of capital and secondly that founders are getting low valuations for their companies (disincentivising people to dream up and action ideas that have the positive returns). The only beneficiaries are the Dragons and other VC backers who are rich enough to disregard the fear of loss for the comfort of statistical averaging.
So should the government invest in VCs? From a cost perspective UK Gov can raise debt at c. 1-2%, and take a sufficiently long and steady view that many of the volatility risks of investing are not seen. So certainly they can afford to invest, but including the impact on long term tax take, is it a better total return investment than infrastructure, or NHS/teachers salaries? IMHO the reason that VC investment makes sense for governments whereas equities much less so (though there is a case to make for for sovereign wealth funds) is that in addition to the potential financial return on investment, £1 spent via VC will be rapidly injected into the economy via startup spending (e.g on staff costs generating tax income, reduced dole costs, and second order effects as staff spend their salaries). Invest £1 into VC fund, generate c.20p tax/savings, and you can break even as a government even when the VC fund makes a small loss and returns you only c.90p ( 20p+90p = £1 + c.10 years funding costs at 1%).
Now you can argue that the government is already doing this to some degree - investing into startups with Enterprise Capital Funds. Except here's the rub. Who do you trust to invest in the right companies? Your local government quango appointee, or the fund with a prove track record earning 15% returns and helping the startups in more ways than just cash. I believe governments should just piggyback on the highly successful VC knowledge and incentivisation structure and link the fee structure they pay VCs to long hold times and proportion of UK based staff in the funded companies. The appropriate investment levels by the way are measured in the billions, not in the £50 millions. The appropriate quantum is dictated by the point at which average VC returns are reduced to slightly below equity market levels of 5-10% (which is the point at which government should be economically indifferent to buying stock in listed UK PLCs and hence supporting the IPO market which indirectly feeds back to VC funding). A good start would be to heavily trim this 'Technology Strategy Board' and redirect the £390m.