Exotic pairs and the spread you did not expect
Somebody quoting you a price has to hold what you hand them until it can be passed on. Further down the currency list the holding takes longer, and the waiting is the thing you are charged for.
A pair is exotic when getting out of it takes time
The word suggests distance and distance has nothing to do with it. What puts a currency at that end of the menu is how long a firm expects to sit on your position before it can hand it to somebody who wants it. At the top of the list that is a matter of seconds at any hour a desk anywhere is open; further down it stretches to however long the opposite need takes to appear, and the currency has no idea what hour it is in London. So a quote on a thin pair prices a wait of unknown length, and every penny the market travels while the firm carries what you handed it belongs to the firm. Width is what it charges for accepting a duration nobody can state in advance, and almost everything that looks arbitrary further down the list follows from that.
Two exotic quotes, and no way to tell what size each is good for
A quoted spread is a price for a quantity, and the quantity almost never appears beside it.
On the heaviest pairs the omission is harmless, since retail size sits well inside whatever the firm had in mind and any sensible assumption gives the same answer. Further down, the assumption starts doing real work. A price good for a modest parcel and a price good for several times that are different products printed in identical type.
So two exotic quotes set side by side are not straightforwardly a comparison. The narrower may simply be quoting a smaller amount, and finding out involves dealing rather than reading, a strange thing to say about a figure whose purpose is to be compared before anybody deals. Read a table of exotic spreads as a list of firms willing to show a price at all, ordered by a number whose terms differ from row to row.
Two institutions read one fact and never spoke
The FCA caps retail leverage at 30:1 on the major currency pairs and 20:1 on non-major ones, the rung it gives gold and the major indices. A dealing desk states its view of a thin market by widening the price it shows; the rulebook states its view by lowering the multiple it permits. Neither consulted the other, and both read one property: how dependably the thing can be got out of when everybody wants out at once. The readings differ only in resolution, width moving continuously with the hour, the news and the size you ask for, while the rung is a single step. Cross the boundary and a given deposit carries two thirds of the exposure it carried before, whether the market underneath is nearly as deep as a major’s or nothing like it.
Four charges with one cause between them
The costs attached to a thin pair arrive looking like four items and behave like one. The price of crossing the quote, what the later part of a large order costs against the first, the deposit the rules oblige you to post, and the amount charged to hold it overnight. Four parties set them, and each answers the same question about how readily the thing can be got out of.
Which is why shopping an exotic on the quoted spread alone disappoints. Move to a firm showing a tighter number and you have improved one item of four; the other three were never that firm’s to set. The rung belongs to the regulator, the funding leg to a rate market, and the rest of a large order to whoever is still at a desk.
| Charge | What it pays for | Where the market is deep | Where it is thin |
| The quoted spread | Dealing now instead of waiting | Passed on in seconds, at a price obtainable | Held until a taker turns up |
| The rest of a large order | The part the first quote will not hold | Little, since retail size fits inside it | Whether anything stands behind it |
| Margin posted | Exposure allowed per pound of deposit | The 30:1 rung, for major pairs | The 20:1 rung, whatever the real depth |
| The overnight roll | Funding the position for a night | A forward market as busy as the spot one | A forward market emptier than the spot one |
The two leverage rungs are the FCA’s published retail caps. Everything to the right of them is reasoning about how these markets are assembled.
The measuring went where the dealing is
Nothing above rests on an exotic-pair price that anybody paid. No such reading is on the record here, not one observation, so every claim that these prices run wider is inference from how the markets are put together rather than a finding off a statement.
The reason is where measurement gets spent, not whether it works. Dealing on a funded account settles a cost question outright for the instrument dealt, and the instrument dealt is the one on the front page of every platform, because that is what most readers buy and what a firm builds its pricing to be judged on. The site behind this page is plain about its method: 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 what it can say about exotic pairs at UK forex brokers therefore has to be assembled from how those markets work. Sampling the busy end of a menu is sampling the part readers use, which is the right call and exactly why the quiet end stays unmeasured by anybody.
The night charge is set in the emptiest market
Holding a leveraged currency position overnight means somebody is funding it, and the funding is priced off the gap between two countries’ policy rates. Between two large economies that gap is modest either way. Put a currency with a very different policy rate on one side and it becomes the largest recurring figure on the position, its direction a fact about two central banks rather than about your idea.
Thinness reaches this too. The roll is priced somewhere other than the market on your screen, in the forward market for that currency, which carries its own participants and can be far emptier than the spot one. In a few currencies the forward is restricted, settled in another currency altogether, or both. So the running cost is decided in the least populated market the trade touches, and no published figure covers it.
What thinness is, once you stop calling it a defect
Reading a wide price as sharp practice leads to bad decisions, and the reason is in the first section. Compress the width by instruction and the honest response is to stop showing the pair, since nothing in the instruction shortens the wait the width was covering. Thin markets tend to lose their retail access that way rather than have it repriced, which is worse for anybody who wanted to deal in one.
The complaint that survives is about visibility rather than fairness. A charge varying with conditions nobody outside the firm can see cannot be budgeted for, and the figure showing when you looked belongs to whatever prompted you to look. Judge it against the distance the pair travels: something moving three times as far absorbs a wider crossing and still leaves more of the move with you.
The instinct on meeting a wider price is to hunt for a firm quoting it tighter, and on the deepest pairs it is sound: the differences are real and easy to compare. Further down it misleads. Two firms quoting the same exotic draw on much the same short list of wholesale providers, so the gap between them is smaller than the gap you accepted when you chose the instrument.
That turns the decision back into an instrument choice rather than a provider choice. Most of what an exotic costs is a property of the currency and arrives with it; a firm can shave a portion, no more. Choosing to trade one is choosing to pay for a market where fewer people will stand, which is coherent, and better chosen on purpose with all four charges in view than discovered on the second statement.
…