One account reached two ways. Ask what it costs and the answer is no difference, which is true and settles less than it sounds.
The bill is written at the account, not at the screen
Price on a spread betting account is set by the terms you signed, the instrument, the size and the minute you deal in, and not one of those is a property of the window in front of you. Open one account on a desktop terminal and on a handset, deal the same stake in the same minute, and quote, overnight charge and commission come out of one arrangement.
What no difference actually settles
So the honest answer to whether the app costs more is almost certainly no, and the answer is smaller than it looks. It settles the price list and says nothing about what you can do with it. The two builds are not two price lists, they are two sets of things you can ask for, and the second difference never arrives in money.
The comparison nobody is organised to make
Worth saying at the front, before anything reads as a finding. No published cost comparison of a handset build against a desktop build sits behind this piece, not from the site linked below and not from anywhere else. What follows is an argument about what to check, and reading it as a report of measurements is the mistake it exists to warn about.
The gap is structural rather than an oversight. Firms compare themselves with competitors, and reviewers compare one firm with another. This axis runs inside a single provider, between two things sold as one product, and nobody has reason to travel along it. A firm publishing that its phone build will not construct what its desktop build constructs would be arguing against half its own offering on the record.
An order type is a sentence the software lets you finish
An instruction to a broker has slots in it. What, which way, how much, at what level, how long it stays live, and what happens to the position afterwards. The build supplies the grammar, and fewer slots make a shorter sentence. The shorter sentence is perfectly valid and asks for less.
A phone build is designed by subtraction
Desktop builds accumulate slots, because there is room on the screen and years of requests behind them. A phone build starts from a thumb and a small pane, so the design work is subtraction. It is usually done well, by people aiming at the instructions most customers send, and nobody performing it has reason to tell you which slot went.
What you do instead when the sentence will not finish
A missing instruction rarely stops anyone. It converts into a manual procedure.
The position goes on and the protective level is added a few seconds behind it. The half close becomes a full close followed by a smaller reopening. The pair of exits that should have cancelled each other becomes you, watching.
Every substitute works, in that the outcome stays reachable by hand. What each one does is move a step out of the software and into you. So a shorter instruction set is a transfer of work rather than a reduction in it, and there is no mechanism by which such a transfer could change the price: the tariff was agreed at the account, and it has no field for who did the arranging or how many attempts it took them.
| What you want to send | Who settles whether you can | The substitute you perform | What the substitute takes off you |
| Protective level attached as you open | The build | Open first, attach second | Seconds with nothing on the position |
| Closing part of a position | The build, within the firm’s minimum stake | Close the lot, reopen smaller | An extra crossing of the spread |
| Two exits, either cancelling the other | The build | Watch, and cancel by hand | Attention, and the exit you missed |
| Moving the level on a resting order | The build | Cancel it, place another | The interval when nothing rests |
| An expiry on an order awaiting a fill | The build offers it, the firm fills it | Leave it live and remember | A trade landing after you stopped wanting it |
| Dealing in what the shorter list omits | The firm’s product range, not the build | Reach the desktop, or leave it | Nothing, which is why this row is here |
Six instructions, who settles whether you can send them, and what people do by hand instead. The second row is the exception the rest of this argument has to live with: an extra crossing of the spread does arrive on a statement, itemised.
A fee changes by announcement and a build changes overnight
A change to charges is an event with a dated page behind it and usually a notice period. A change to what the software will construct arrives as a download while you sleep. The public record of that second change is a release note in an app store, composed to sell the update rather than to inventory what moved.
Which is where funding accounts stops helping here. A finding about an instruction set belongs to a version, and a testing programme is organised around firms, filing what it learns under a provider’s name. A price held that way stands until the provider says otherwise. A capability held that way expires the moment a release lands, nothing about the firm having changed, and the filing gives no sign it has happened.
The case for the handset is better than the sceptical reading allows
The concession, and not a grudging one. For a great many spread betting accounts the instruction needed is buy this and later sell this, and both builds finish that sentence without strain. The build you can reach from a coat pocket is the one that gets used, and a position you can manage beats a richer menu two hours away.
There is a genuine cost on the other side that feature comparisons leave out. A capability you hold only at a desk is one you do not hold at the moment most exits are wanted, which is rarely a moment spent at a desk. Anyone arguing the desktop case on the length of its menu owes the phone side that entry.
The two questions have different owners
Notice who is answerable for each half. What it costs is set out in a document the firm is obliged to get right and keep current, and you can hold it against your own statement to see whether the two agree.
What the build will construct is answered by a help article, a gallery of screenshots and, at best, whoever picks up the chat window, none of them under that first obligation.
Settling it means holding both builds open on one funded account and trying the awkward order in each. Doing it across a market would take one person funding accounts at several firms and working each from two devices for a month, and no research programme is built that way, The Investors Centre included, though it gets nearer than a feature list can: it opens and funds live accounts with its own money to test UK trading platforms rather than compiling rankings from providers’ published fee schedules, and its comparison of the firms offering spread betting on a handset is built from accounts it paid to open.
Send the awkward order before it matters
A check costing nothing beyond the deal you were doing anyway. Pick the most involved instruction you have ever wanted to send: a level attached at the moment of opening, a partial close, a pair of exits that answer to each other. Then, at a stake whose outcome you do not mind, try to construct it in the build you use most, and afterwards in the other. Do not read about it. Attempt it.
Three results, and each of them tells you something
It goes through in both, and the question is closed. It goes through in one, and you know which screen the awkward moments have to be met on. It goes through in neither, in which case the instruction was never on offer at this firm, and that is a fact about your broker rather than your phone.
…