It would be a mistake to pay too much attention to the figures used by analysts in relation to the risks of Brexit. ING, the Dutch bank, was the latest into the fray, estimating yesterday that Irish GDP could be a cumulative 1.1 per cent lower by the end of 2017 if Britain did vote to leave.

There are so many assumptions behind these models – and so many other factors that could come into the equation – that it is best to look on them as offering a guide on the direction things might go, rather than any kind of exact estimate.

Nonetheless, the underlying message from all the analysis is clear: this could cost us. Potentially, we might gain from attracting investment that would otherwise have gone to the UK – a recent National Treasury Management Agency (NTMA) report referred to estimates of a €6 billion gain in foreign direct investment – but we would lose on trade, initially from a likely sterling fall. Beyond that, things would depend on what new regime was put in place. As the NTMA said, smaller firms with less ability to deal with bureaucracy and any complex rules that might emerge would be in the firing line.

However, the hardest thing of all to fit into an economic model is the impact on sentiment. If the polls suggest in the run-up to the vote that a Brexit is likely, there could be big delays in investment and spending decisions in Britain, which will knock on here. Nobody knows how Britain would renegotiate its trading relationships with Europe after Brexit, but it is sure to be a long, tricky process.

It is simply impossible to do an economic model that catches these confidence factors with any reliability. So far, the markets and most businesses reckon that the majority of the one in five or so voters who currently say they are undecided will, as happened in the Scottish referendum, choose to stay. If this view changes, market and investment conditions could turn volatile, even before a vote is cast.