Forex

One pair, two currencies, a single price

On this market you never buy a currency on its own. A quote is not a price in the ordinary sense, it is a ratio between two currencies. EUR/USD at 1.0850 does not say what the euro is worth in absolute terms, since no such value exists: it says how many dollars you must hand over to get one euro.

The first currency is the base currency, the second the quote currency. The base is the one you deal a fixed quantity of, the quote is what you pay with. Buying one lot of EUR/USD means buying one hundred thousand euros and selling the dollar equivalent, in a single indivisible operation.

The order of the two currencies is left neither to chance nor to your broker. The market follows a stable precedence convention: the euro first, then sterling, the Australian dollar, the New Zealand dollar, the US dollar, the Canadian dollar, the Swiss franc, and the yen last. That is why you will always see EUR/USD and never USD/EUR.

The most useful consequence fits in one sentence: a pair can rise for two opposite reasons. EUR/USD rises if the euro strengthens, and just as much if the dollar weakens while the euro has not moved at all. The chart draws no distinction between the two. Open three dollar pairs side by side: if they all move in the direction that matches a weaker dollar, the news is American, and the move you are watching does not belong to the euro.

Selling is as natural as buying

On a share, selling short is a second-class operation: you have to borrow the stock, pay for that loan, accept that it can be recalled, and some regulators restrict it in turbulent periods. Many beginners come away with a vague sense that selling is less legitimate, or riskier, than buying.

None of that exists here, for a structural reason: any position on a pair is already a purchase and a sale. Selling EUR/USD means buying dollars against euros. It is not the opposite of a purchase, it is a purchase written the other way round, and there is nothing to borrow because there is nothing to deliver that you have not simultaneously acquired.

The second consequence is less visible. A share has a mechanical reason to rise over thirty years: the company reinvests, the economy grows, inflation lifts nominal prices. A ratio between two currencies has no such engine, therefore no equivalent of buy and hold. It does carry a drift, and that drift is worth naming rather than mistaking for a trend: when two currencies live for years with widely separated interest rates and inflation, the one being eroded loses ground against the other, slowly and durably. That drift is not held for free, you pay it or you collect it every night, and this lesson comes back to it further down.

⚠️ Symmetry of direction is not symmetry of cost. On the same pair, holding a position overnight can pay you one way and cost you the other, and your broker's mark-up is taken in both. It is the cost line beginners discover last.

The pip, the lot, and the arithmetic of a point

Two units describe any position on this market. The pip measures movement, the lot measures size, and the product of the two gives what a point is worth on your account. The word comes from “percentage in point”, the smallest increment a quote used to move by when quotes were given over the telephone, and it outlived the extra decimal added since.

The pip is the fourth decimal of the price, that is 0.0001, on almost every pair. The exception concerns pairs quoted in yen, where it is the second decimal, 0.01, because a yen is worth roughly a hundred times less than a euro and you do not quote two orders of magnitude to the same precision. Most brokers display one extra decimal, the pipette: a price with five decimals is not ten times more accurate, it is ten times finer.

The standard lot is one hundred thousand units of the base currency, the mini ten thousand, the micro one thousand. The arithmetic then fits on one line: one hundred thousand multiplied by 0.0001 gives ten units of the quote currency. A standard lot on EUR/USD is therefore worth ten dollars a pip, whatever the level of the price, and on a yen pair one hundred thousand multiplied by 0.01 gives one thousand yen. You can redo that calculation in your head tonight, and it depends on no market data at all.

⚠️ Here is where beginners go wrong: when the quote currency is not the currency of your account, pip value is not constant, since it has to be converted at the rate of the moment. Two positions of identical size on two different pairs therefore do not risk the same amount. Think in amount risked, never in lots.

Majors, crosses and exotics

Three families. Majors pair the US dollar with the large convertible currencies: euro, yen, sterling, Swiss franc, Canadian dollar, Australian dollar, New Zealand dollar. Crosses put two of those large currencies face to face without going through the dollar. Exotics set a large currency against that of a smaller economy, or one less open to capital.

What genuinely separates them is measured in three quantities. The spread, paid on every round trip. Depth, the quantity the market absorbs without the price moving, which decides your slippage on the day you exit in a hurry. Hours, finally, because a currency is only truly dealt while its country is awake. A cross also inherits the nervousness of both its legs, one for each of the two currencies it is made of.

⚠️ On an exotic, the spread alone is sometimes enough to cancel a correct analysis. A forty-five-pip toll on a round trip, on a market that covers four hundred pips in a day, takes eleven per cent of the available move, against roughly two per cent on a major. Add quoting gaps in quiet hours, and the very same method produces an unrecognisable result.

That ranking is not a hierarchy of prestige, it follows from a measured fact. The triennial survey the Bank for International Settlements has published since 1986, whose results are public, finds the US dollar on one side of close to nine currency trades out of ten, a proportion that has not moved in twenty years. A pair containing the dollar therefore sits in the densest flow on the market, and that flow is what pays for your tight spread. A cross, meanwhile, is still often manufactured by your liquidity provider out of two dollar legs: you settle two spreads instead of one, which explains the step between the first and the second family far better than the standing of the currencies involved.

Three sessions, and the overlap that carries the day

The currency market has no bell. It opens on Sunday evening with the first Asian centres and closes on Friday evening with New York, without interruption, because when one centre goes to sleep another has already woken. The rhythm exists nonetheless, it is hourly instead of daily.

Three sessions carry most of the activity. Tokyo covers the European night, London the morning and early afternoon, New York the afternoon and evening. London is by a wide margin the largest foreign exchange centre in the world, a fact you can verify rather than believe: the same triennial survey breaks turnover down centre by centre, and London has handled more than a third of world volume for decades, more than New York, Singapore and Hong Kong combined.

The moment that counts is none of those sessions taken alone, it is their overlap. During the few hours when London and New York work together, the two largest concentrations of participants are present at the same time: spreads tighten, depth increases, and a break stands a chance of being defended by somebody. The same setup, drawn three hours later, is drawn in an empty room.

⚠️ Twice a year your landmarks shift by an hour, and not on the same day on both sides of the Atlantic. For two or three weeks, the overlap no longer falls at the hour you had learnt, and many traders spend those weeks wondering why the market is limp at the exact moment it should be sharp.

Leverage, and why it does so much damage here

A major pair generally covers less than one per cent in an ordinary day. On a thousand dollars of capital that is a handful of dollars, which interests nobody. Leverage exists to make those moves meaningful, and that is exactly where the trap closes. At 1:30, a thousand dollars carries thirty thousand dollars of position: a one per cent move in price becomes thirty per cent of your capital.

⚠️ Leverage does not raise your risk by itself. Position size does, and leverage merely makes it possible. Available leverage is a ceiling, not a setting: it says what your broker is willing to let you carry, not what you should carry. A trader sizing from their stop can have 1:500 available and still risk one per cent. A trader taking one lot because it is a round number will have the same ceiling and a problem.

The costliest confusion concerns margin. Margin is what your broker locks up to let you carry the position, not what you risk: two unrelated numbers. When equity falls below a percentage of required margin, the broker closes your positions itself, and it picks that moment for you.

Getting out of the trap is mechanical. Decide the amount you accept to lose, measure the distance to your stop in pips, then divide the first by the second multiplied by the pip value. Size comes out of that calculation, and leverage appears nowhere in it: that absence is precisely the sign the calculation is the right one.

The swap, or the price of a night

A spot currency deal does not settle instantly: on almost every pair it settles two business days out, a convention written T+2. A handful of pairs settle in a single business day instead, among them the Canadian dollar against the US dollar, aligned on the North American securities calendar. A position still open at rollover time therefore changes settlement date, and that shift has a price, called the swap.

Its origin is an interest rate, not a disguised commission. Holding a pair makes you the lender of one currency and the borrower of the other: you receive the interest of the first, you pay the interest of the second, and the balance of the two, adjusted by your broker's mark-up, is credited or debited every night. That balance can therefore be positive or negative, and it flips sign with the direction of your position.

Wednesday is the exception that surprises everyone, and it is deduced rather than memorised. A position held on Wednesday settles on Friday. Rolled on Wednesday evening it becomes a Thursday position, which settles the following Monday: the value date steps over Saturday and Sunday, and the rollover charges all three days at once. That is not a penalty invented by your broker, it is the banking settlement calendar. The check is within your reach: on pairs settling in one business day, the triple rollover falls on Thursday evening rather than Wednesday, exactly where the same calculation puts it.

⛔ Never open a position for the sole purpose of collecting a positive swap. Two relations carry a name here, and confusing them is the most expensive mistake in the subject. Covered interest rate parity, described by John Maynard Keynes as early as 1923 in A Tract on Monetary Reform, says the interest rate gap between two currencies reappears exactly in the forward price. That one holds, because it rests on a riskless arbitrage. Uncovered parity would add that the spot rate moves on average so as to cancel that gap, which is precisely what it would take for your positive swap to be a transfer with no consequence. The data has contradicted it for forty years, so consistently that the phenomenon earned its own name: the forward premium puzzle.

The trap is therefore not that a positive swap is an illusion. It lies elsewhere, and it is worse. That payment is genuinely there, it accumulates quietly, and it pays you for a risk whose shape is documented: positions funded that way unwind all together on the day the market tightens, handing back in a few sessions what months had patiently set aside. Collecting a swap is not earning a yield, it is writing insurance, and insurance is paid for on the day of the claim.

An intraday method never pays the swap. A method that holds positions for several days must write it into its expectancy, exactly like the spread and the commission. Open your broker's swap table once, note your pairs in both directions, and you will know whether your method is allowed to sleep.

No central volume, and what that rules out

The spot currency market is an over the counter market: no exchange, no central book, no authority publishing a last price. Your broker builds its quote from the ones its liquidity providers send it, and adds its own spread. Two brokers therefore show two slightly different prices at the same instant, and neither of them is wrong.

What that structure rules out is very specific, and many methods on sale take care not to mention it. The volume your chart displays is not a quantity traded: it is a count of price changes, tick volume, seen by one broker among hundreds. It tracks activity reasonably well, which makes it all the more misleading. Every reading resting on real volume fails here: confirming a break by volume, the volume profile, the volume-weighted average price.

The same gap shows up at the extremes of the chart. There is no official high of the day, only the high your broker saw. A wick a few points deep exists at one broker and not at the next, which means a stop can be hit on one server and not on its neighbour. That is not foul play, it is the absence of an exchange.

Two honest substitutes remain, both centralised and public. The volume of currency futures first, born in Chicago in 1972 and traded on an organised market ever since: it covers only a fraction of the market, but what it shows was genuinely dealt, at a price a clearing house recorded. Then the commitments of traders report, published in the United States since 1962 and weekly since 1992, which describes aggregate positioning on those same contracts, category of participant by category. Neither replaces a book, and both are worth more than a tick counter.

Three pairs, for a long time

There are dozens of pairs available, and that abundance is one more trap. Pick three, and keep them for several months. This is not modesty: what separates a trader who improves from a trader who accumulates hours is the number of times they have watched the same thing happen.

A second reason argues for that small number. Pairs are not independent, and direction matters as much as the pair. Buy EUR/USD, GBP/USD and AUD/USD on the same day: you are not holding three positions, you are holding one short-dollar position of close to triple the size, carrying a risk your statement displays as three reassuring lines. Buy EUR/USD and USD/JPY together instead, and the dollar is no longer on the same side: the two bets partly cancel, without your having decided it or measured it. It is one of the quietest ways to empty an account, and the only guard is to count your exposure by currency rather than by line.

Build your trio in that spirit. One dollar major whose active hours fall when you are available, a second major that does not share the same second currency, and a third only once the first two have become familiar. An exotic has no business in a learning trio: its toll makes the quality of your decisions impossible to read.

The exercise, to start tonight and hold for ten business days. A sheet with three columns, one per pair. Every day, at a fixed hour, record four things: the displayed spread, the day's range in pips, the hour of the high and the hour of the low. Two minutes a day, no paid tool involved.

In two weeks that record gives you what no product sheet ever will: the hour your pair wakes up, what it costs depending on when you trade it, and whether the high of the day falls inside the overlap often enough to make waiting worthwhile. Add the swap of your three pairs in both directions, once and for all, and you hold a sheet that is true for your broker, which no outside source can promise.