Donald Trump’s odds of winning the US Presidential election have halved - but he still remains a long shot to walk into the Oval Office.

The revelation that the FBI has reopened its investigation into his Democratic opponent Hilary Clinton’s use of a private e-mail server while Secretary of State is the latest scandal to impact upon the race.

While Trump had been trading at around 5-1 when I analysed the betting markets last Wednesday, his price fell to as low as 2-1 over the weekend. At the time of writing several bookies went going 9-4 on a Trump victory although it wasn’t hard to find the reality TV star and businessman at 11-4.

That latter price implies that he has a 26.7 per cent chance of winning, compared to 16.7 per cent implied by his pervious odds of 5-1. It's a considerable move in his favour.

By contrast FiveThrityEight, which uses polling and other data to produce its forecasts, rates his chance at 21 per cent using a polls only model and 23 per cent using polls plus historical and economic data. So while there appears to be little value in backing Trump at 11-4, Clinton at 1-3 looks under priced.

Part of the betting move has been fuelled by a string of polls that suggest the race is tightening. The BBC’s latest polls of polls, which uses the median average of the five most recent national US polls, has Clinton at 49 per cent and Trump at 46 per cent. The Washington Post-ABC tracking poll has Clinton at 46 per cent and Trump at 45 per cent, including third party candidates.

FiveThirtyEight's Nat Silver, however, notes that Trump had been enjoying a surge in support before Friday’s news about Clinton’s e-mail server, which might have had less of an impact than some commentators suggested, perhaps because it is already on Clinton’s “rap sheet”. It is therefore open to debate whether Friday’s news, by itself, will have changed many people’s views about her.

If the prospect of Trump in the White House makes you shudder, the fact that betting patterns on the US election so closely mirror the betting patterns seen ahead of the UK’s EU referendum, with smaller punters who backed Brexit backing Trump while big money punters who backed remain backing Clinton, is worrying enough. The latest Trump surge will have exacerbated those worries.

Alex Donohue at Ladbrokes said of the latest betting: “There has actually been a small move back to Hillary this morning (1/3 from 4/11) which is due to some of the larger staking players backing her whenever her price hits a certain point - clearly they are many who believe they can extract some value from the way the prices have been manipulated by the masses of smaller-staking Trump voters.

“Certainly some of the high-rolling Hillary backers believe anything related to emails is pretty much priced in anyway, so she looks a very attractive bet at 1/3 compared to far shorter odds offered last week.”

By contrast, Graham Sharpe at William Hill, said: “We’ve seen virtually nothing of any significance for Hillary. Trump has come down and is making the running. We’re interested to see where the next move goes.”

Part of the reason Clinton is starting to look like value at 1-3 is that it is the US electoral college that elects the President and not the popular vote. The college favours Clinton.

Indications from early voting in both North Carolina and Nevada suggest that these two battleground states have been trending in her direction. The Trump surge would need to continue, and strongly, if he were to overcome her cushion in those states.

Alternatively he needs to take some of the other in play states that have looked like they are in Clinton’s column, such as Wisconsin and Michigan, assuming (as looks likely and as I have tipped) he takes Ohio and also Iowa (and Arizona).

Wisconsin, however, also has early voting (not all states do). New Hampshire is also in play as is the second district of Maine (it assigns its college votes based on the results in each of three districts).

Sporting Index’s spread on Clinton’s performance in the college has fallen to 303-315, down from 325-335. That still represents a comfortable win for Clinton. Using FiveThirtyEight’s polls plus forecasts, which are more favourable to Trump, and assuming they are correct for each state, Clinton should win 319 votes. Which makes the spread worth a modest buy at 315. I'm assuming with that recommendation that Clinton takes the mid western states she is expected to (Michigan, Minnesota, Wisconsin, Illinois), Florida, North Carolina, Maine and New Hampshire, while losing Iowa, Ohio, and Arizona.

Why would I remain a Clinton backer? The “Brexit” effect can’t be ignored, Trump rage filled supporters have much in common with Brexit backers, but none of the motley crew of Brexiteers that led the two Leave campaigns had the same level of scandal attached to them as does Trump and I’d be willing to bet that the next rumpus to emerge out of a campaign with more twists than a John Grisham legal thriller, impacts on him.