Trying to find a cheap, value punt in the Brexit markets has not been straightforward. So far as the main result market is concerned, the scope for short-term trading has been limited and taking a confident view involves a sizeable risk, in a referendum which is far from clear cut.

Remain, for example, has traded between 1.45 and 1.16 since the campaigns began in earnest. So if you fancy the favourite winning – as I do – a big stake is required. However as I’ve argued before, the potential for differential turnout tempers confidence and leads me to think Remain will underperform the polls.

Imagine, for example, that we’d got involved last week, when Remain seemed to be pulling ahead and shortening on betting markets, at say [1.25] (80%), before it peaked at 86%. A couple of contrary polls later, and that rating has dropped to 75%. What felt like a solid trade, anticipating an obvious market move, would suddenly be a negative position. Not my kind of betting – too much risk for too little reward.

Again to repeat my sole existing trade, a preferred market is on Remain’s vote share and I certainly wouldn’t deter anyone from taking the 3.55 now about 50 – 55%. However the recent odds movement presents an excellent value opportunity to hedge between the main result market and the handicap giving Leave a 2.5% start.

Whereas Remain can be backed at 1.34, Leave is 2.9 to get over 47.5%. Backing both for proportionate stakes equates to odds of [11.0] about Remain getting between 50 and 52.5%.  See the calculations below.

Back Remain 90 units @ 1.34

Back Leave Over 47.5% 41.6 units @ 2.9

So, if Remain gets more than 52.5%, we make 30.6 units on the first bet, but lose 41.6 units on the latter, meaning an 11 unit overall loss. Likewise if Leave wins, we lose 90 units on the former, and win 79 on the latter, losing 11 units overall. (These precise stakes are to yield an equal return and should be altered accordingly with price changes).

But if Leave gets between 47.5 and 49.99% – accounting for half of the preferred band that I reckon represents a good value 28% chance – we will win both bets, yielding 110 units profit from an initial risk of 11 units. In short, a 10 to 1 chance or just 9%. While these odds are around, this simply must represent value and would still do so even if the combined odds shortened to 7 or 8 to 1.