Real trades 3 - Vodafone over one week

• Underlying: Vodafone

• Time period: Seven days

If that didn't complicate things enough, it also emerged that CFDs and spread-bets are so similar that any comparison boils down to a choice between one particular spread-bet/CFD provider and another (nearly all companies that offer the one will offer the other these days).

With that in mind, I decided to stake exactly the same amount of money per point with the final trade, a week's trade on Vodafone. Having been conservative in my stop-losses, I ramped up the risk a little, betting £10 per point with both the CFD and the spread-bet with a stop-loss 20 points lower (Vodafone at 117p).

For the warrant trade, I chose S399, a Vodafone call warrant with a strike price of 140p (not far off the money, with Vodafone trading at 137p) and set to expire in September, giving me plenty of time for the market to recover should I be wrong about the stock's short-term direction.

Knowing that the general election was coming up, I also decided to leave the trades open over the night of the election itself, to see what effect market volatility might have (I closed all three trades on the morning of Friday 6 May).

Part of me wanted Vodafone to lose money. So far, I had made money on all six trades. That's not a realistic success rate. Losing money on the Vodafone trade would have been a chance to compare the derivatives when the markets catch you out.

In the end, Vodafone rose. I thought about closing the trades on Thursday evening, with the underlying shares at 139p, but stuck to my original plan to see what the markets would do once the election results were known.

As it happened, Labour's modest majority spooked the market and, on the day, both the Footsie and Vodafone fell. Still, when I closed all three positions, Vodafone was steady at 138p.

During the previous week, market volatility had been above average, a fact that helps to explain why the warrant did so well. From a 0.7 per cent rise in the underlying, the Vodafone call warrant rose from 6.1p to 6.65p (accounting for bid-offer spreads), giving me a 9 per cent return before dealing costs. That beat the performance of both the spread-bet and the CFD.

Four important distinctions emerged from this final head-to-head comparison between the spread-bet and the CFD.

First of all, the spreads worked in favour of the CFD this time around. Interactive Investor had quoted a 0.25p spread on Vodafone (the same as the market spread and IG's Vodafone warrant).

However, IG's spread was 0.5p on the opening leg of the trade and had inexplicably widened to 0.8p on the closing leg. So, while I bought at 137p and sold at 138p for the CFD, I bought at 137p and sold at 137.8p for the spread-bet. The difference meant I made 20 per cent less on the spread-bet than on the CFD (before commission).

What this underlines is the difference between trading on exchange prices and a spread-better's prices. A CFD that uses the London Stock Exchange's prices (and spreads) will guarantee you the same prices as if you were a private investor buying shares, and therefore entitled to best execution. But a spread-better running its own book can add a spread to either or both sides of the market's bid-offer spread, depending on his overall position on Vodafone.

Second, CFDs on equities can have fees that undermine their competitive spreads. Trading CFDs on forex or indices might be commission-free but, as soon as you trade equities, Interactive Investor charges a £10 minimum fee (in line with the industry average). Although I made £10 out of the one-point rise in Vodafone, I had to pay out £20 to get in and out of the position.

Third, remember when you trade indices or currencies, both spread-bets and CFDs work on a pound-per-point basis (or equivalent). However, to put together a CFD trade equivalent to my £10-per-point spread-bet, I had to buy many more CFDs. To work out how many, I calculated the number of shares in Vodafone I would have to own in order for a 1p rise to make me £10 (£10 = 1p x 1,000), therefore I needed to buy 1,000 CFDs to match a £10-a-point spread-bet).

Fourth, there were different financing costs (this arose with the overnight trade too). With the IG spread-bet, the cost is reflected in a daily roll-over fee charged on the Vodafone bet. At the end of each trading day, IG would close my position (at 137p, say) and reopen it the next morning slightly higher (at 137.026p). Over a week, that added up to seven extra spreads that I paid at £10 a point, totalling £1.87.

The financing charge for the CFD was more transparent in that the CFD provider looks at the total value of the trade (1,000 Vodafone shares x 137p) and applies a daily rate of interest based on LIBOR plus 1.5 per cent. Borrowing £1,370 at 6.25 per cent per annum worked out at £1.61 for the week.

In total, then, I lost £11.61 on the CFD Vodafone trade and made £6.13 on the spread-bet. Had I upped the stakes and bet £100 per point while buying 10,000 CFDs, I would have made roughly the same with either derivative. The tighter spread on the CFD would have compensated for the extra commission. Had I traded more than that, the maths would have favoured the CFD.

In the end, though, it was the covered warrant that did best out of this small movement in Vodafone (but only thanks to the £1 commission, which is a special offer that cannot last forever).