When should I convert the money I receive?
Convert while the market is open, on a schedule you decide in advance, and only hold the currency for a reason you can state without forecasting. Beyond that there is no answer that applies to every month, because the question mixes two different problems. One of them has a reliable answer and the other does not.
Fees and markups are costs. You pay them every time, they are knowable in advance, and reducing them is arithmetic. Timing is exposure. It can move in your favour or against you, nobody knows which in advance, and the only thing you can reliably change is how much it varies.
Treating the second like the first is what leads people to sit on money waiting for a better rate. This article is about what can actually be decided.
Does it cost more to convert on a weekend?
Usually, and this part is a cost rather than an exposure.
The global FX market runs from Sunday evening to Friday evening, New York time, and is closed in between. A rate you see on a Saturday may be based on Friday's closing rate rather than a live one, because there is no live one. Many providers price weekend conversions from that close with additional margin, since the market can gap when it reopens on news that broke while it was shut. How much margin, and whether they add any at all, varies by provider.
Liquidity can thin around the edges of the trading week, particularly at the Sunday open, and spreads can widen when fewer participants are active. Public holidays in the relevant financial centres can have a similar effect.
This is one of the few timing decisions with a predictable direction. If you control when the conversion happens, converting while the markets relevant to your currency pair are active generally costs less than converting when they are not. Foreign exchange is a decentralised over-the-counter market with activity concentrated in a handful of financial centres, and spreads tend to be tighter in the hours when those centres overlap and wider when few of them are open.
The cost appears as a slightly worse rate rather than as a separate charge, which means it scales with the amount being converted. How much it is worth to you therefore depends on how much you convert and how often. You can measure it the same way you measure everything else here: compare what you received against the mid-market rate at that moment, and do it once inside market hours and once outside.
One thing to be clear about: waiting until Monday may avoid an additional markup, but it does not eliminate exchange-rate exposure. You are holding the currency across the gap, which is the same position described in the next section. You are trading a known cost for an unknown movement, and on a short horizon that is usually a reasonable trade rather than a free one.
If you do not control it, because your client's payment converts automatically on arrival, then this is not a decision you have. It becomes one only if you can receive in the original currency and convert separately.
Should I hold the money and convert later?
Only if you have a reason that is not a prediction.
Holding foreign currency is a position. If you receive in dollars and your costs are in another currency, every day you hold is a day the rate can move either way. That is not automatically bad, and it is not free either.
Holding is not free either, though the cost is easy to miss because no fee appears anywhere.
There are sound reasons to hold:
- You have expenses in that currency, so converting and converting back would cost you two spreads
- You are waiting out a weekend or holiday, as above
- Your provider charges a fixed fee per conversion and you are batching to pay it fewer times
There is one common reason that is not sound: believing the rate is about to improve. Unless you have a matching need for that currency, holding it leaves you exposed to a future exchange-rate move you cannot predict reliably. Whatever expectation led you to wait is already reflected, to some extent, in the current market price.
Is it better to convert everything at once or in parts?
Splitting reduces the variability of the rate you end up with. It also multiplies any fixed fee. The right answer is where those two meet.
If you convert once, your effective rate for the month is whatever the rate was on that one day. If you split the same amount across several conversions, your effective rate becomes the average of those days, and the average has lower variability than any single observation.
The standard deviation of that average falls with the square root of the number of conversions:
| Conversions per month | Variability of the average, relative to one conversion |
|---|---|
| 1 | 100% |
| 2 | 71% |
| 4 | 50% |
| 12 | 29% |
This is a simplified statistical illustration, not a forecast of exchange-rate movements.
Two things it does not capture. It assumes each conversion is an independent draw, and exchange rates are not independent from one day to the next, so the real reduction is smaller than the arithmetic suggests. And it reduces variability in both directions: you give up the good months along with the bad ones. Splitting is a variance decision, not a profit one.
Now the cost side. If your provider charges a $5 flat fee per conversion, converting weekly instead of monthly costs $15 extra a month, or $180 a year. On $3,000 a month that is 0.5% of the amount, which is a meaningful share of a total cost that might be 1.5% in the first place.
So the question becomes concrete: is halving your rate variability worth 0.5% of the amount? That depends on how much the pair actually moves and on whether an unlucky month would cause you a problem. Both are things you can look up and answer for yourself, which is more than can be said for the question of where the rate is going.
If your provider charges no fixed fee, the direct cost of splitting is small, because a percentage markup applies to the same total either way. It is not quite free: some providers price smaller amounts less favourably than larger ones, and each conversion is another thing to remember. Work out your in-transit cost first, because it tells you whether a fixed fee is part of your structure at all.
What can actually be known about a currency pair?
Its recent behaviour, not its direction.
These are observable and useful:
- How much the pair typically moves over a week or a month. This tells you whether the timing question is worth any effort at all. For a stable pair it may not be.
- Where the current rate sits relative to its own recent range. This is description, not a signal. A rate at the low end of its range is not more likely to rise.
- When the market is closed in the relevant sessions, which is the one thing with a predictable cost.
These are not knowable, by anyone:
- Where the rate will be next week
- Whether today is a good day to convert
- Whether waiting will help
Anything presenting the first list as evidence for the second is doing something the first list cannot support. That distinction matters more than it sounds, because most currency content blurs it, and the blur is what convinces people to wait.
So what should I actually do?
In order of how reliably each one helps:
- Convert during market hours rather than over a weekend or holiday, whenever you control the timing. Predictable, usually small, and avoidable.
- Reduce the cost side first. A 1.5 percentage point difference in in-transit cost is larger than most timing decisions and it applies every single month. Timing is worth optimising after that, not before.
- Decide on a conversion pattern and hold to it. Monthly, fortnightly, weekly, whatever your fee structure supports. The pattern matters less than not making the decision fresh each time based on where the rate happens to be.
- Only hold currency for a reason you can state that does not involve a forecast. Expenses in that currency, a fee you are batching, a weekend you are waiting out.
The last one is the discipline that matters. Most of the money lost to timing is not lost on a bad conversion. It is lost while waiting for a better one.
How do I keep track of this?
Add one column to the record you are already keeping.
If you are logging what arrived against what was sent, add the date of the conversion. After six months you will be able to see whether your timing pattern helped, hurt, or made no difference, which is the only honest way to evaluate it.
Most people never find out, because they change approach every month based on the last outcome. A pattern held long enough to measure is worth more than a better pattern abandoned early.
Sources: Bank for International Settlements, Triennial Central Bank Survey for the structure and scale of global FX markets; FOREX.com and Dukascopy for market hours; ExchangeRates.com for weekend rate behaviour, liquidity and spreads.
This article describes how conversion timing affects cost and variance. It is not financial advice, does not forecast currency movements, and cannot account for your individual circumstances. The fee figures are illustrative. The variability table is the output of a simplified statistical model that assumes independent observations, which exchange rates are not; treat it as a way of thinking about the trade-off rather than a prediction of results.