Last reviewed on August 27, 2026.
"CHF to INR forecast, 7 days." "Will the pound go up next month?" These are among the most-searched questions in foreign exchange, and they have an uncomfortable answer: over horizons of days to months, exchange rates are close to unforecastable, and the sites that publish confident numbers are not doing what their headline suggests.
This site does not publish rate predictions. This guide explains why, what the things that look like forecasts actually are — forward rates, analyst targets, technical projections — and what you can usefully do instead when you have a real transfer to time.
What the research actually says
The finding is old and unusually robust. In 1983 Richard Meese and Kenneth Rogoff tested the leading economic models of exchange rates and found that none of them beat a random walk — the assumption that tomorrow's rate is simply today's rate — at horizons up to a year, even when the models were fed the actual future values of their own inputs. Four decades of follow-up work has qualified the result at long horizons but has not overturned it at short ones.
The intuition is straightforward. An exchange rate is a price set in the deepest market in the world, trading around the clock. Anything that is already known — an expected rate decision, a published inflation trend, a widely forecast growth number — is in the price already. What moves the rate is the part nobody expected. And the unexpected part is, by construction, not forecastable.
The practical version. Over a week, the best available estimate of a currency pair's future rate is roughly its current rate. Any prediction that differs materially from today's number is making a claim the market itself is not making.
Forward rates are not predictions
The most commonly misread number in FX is the forward rate — the rate at which you can contract today to exchange currency at a fixed date in the future. It is often presented as the market's forecast. It is not.
A forward rate is fixed by arbitrage from the two currencies' interest rates, a relationship called covered interest parity. If rupee deposits pay more than franc deposits, the forward CHF/INR rate must show the rupee weaker — otherwise you could borrow francs, convert, deposit in rupees, and lock in a risk-free profit through the forward. The forward simply cancels out the interest differential.
So when a forward curve shows a currency "expected" to decline, it is usually telling you that the currency has higher interest rates, not that anyone expects it to fall. Historically, the currencies forwards imply will weaken have tended, on average, to do slightly better than the forward implied — the observation the carry trade is built on.
What analyst targets and "AI forecasts" represent
Bank research desks do publish exchange rate targets, typically at three, six, and twelve months. Read them as scenario markers rather than point predictions: they express a house view about interest rate paths and growth, they are revised frequently as data arrives, and the banks publishing them will tell you plainly that the dispersion around them is wide.
The forecasts you find on consumer currency sites are usually a different thing again. Most are produced by extrapolating recent price action — a moving average, a trend line, sometimes a small model fitted to past prices. They produce a smooth-looking number to several decimal places for a date weeks away, which conveys a precision nothing in the underlying method supports. A useful test: check whether the site publishes a track record of its past forecasts against what happened. Almost none do.
What is genuinely predictable
Some things about exchange rates are far more knowable than the level:
- The recent range. Where a pair has traded over the last week and month is a fact, and it tells you whether today's rate is unusually high, unusually low, or ordinary. Every pair page here carries 7-day and 30-day tables and a 12-month chart for exactly this purpose.
- Volatility clustering. Calm periods tend to follow calm periods and turbulent ones follow turbulence. You cannot predict the direction, but you can often tell whether the next week is likely to be eventful.
- The event calendar. Central bank meetings, inflation releases, and elections are scheduled in advance. You cannot know the outcome, but you can know when the risk of a jump is elevated — and choose not to leave a conversion to a day when a rate decision lands.
- Your own costs. The spread and fee a provider charges is knowable today, and on a typical retail transfer it is larger than the rate movement you would expect over the following week. This is the part worth optimising.
The asymmetry that matters. Shopping around for a provider might save one to three percent, reliably and immediately. Timing a transfer on a seven-day view might gain or lose a similar amount, unpredictably. One of those is a decision; the other is a coin flip.
Managed and pegged currencies are a special case
Not every pair is a free float. Where a currency is pegged — the Saudi riyal and the UAE dirham against the dollar, for instance — the rate genuinely is highly predictable, because a central bank is committed to holding it. The SAR to USD page shows what that looks like: a nearly flat 30-day table.
The caveat is that pegs and managed bands change in steps rather than drifts. A currency can hold a level for years and then move a long way in a single day when the arrangement is adjusted. Predictability under a peg is not the same as safety, and the recent history of the Egyptian pound illustrates the point.
A practical approach to timing a conversion
- Check the recent range first. Look at the 30-day table on the relevant pair page. If today's figure sits mid-range, there is no timing signal to act on.
- Compare total delivered amounts, not rates. Ask each provider how much of the target currency lands in the account, after every fee.
- Check the calendar for the days around your transfer and avoid leaving a large conversion sitting over a scheduled rate decision if you have flexibility.
- Split large amounts across dates if the sum is significant and the deadline is not tight. Averaging in does not improve the expected rate, but it narrows the range of outcomes.
- For a business exposure, hedge rather than forecast. A forward contract removes the uncertainty at a known cost. That is a different activity from trying to guess the direction.
Where to go next
To read the rates themselves accurately, start with how currency exchange rates are quoted and direct vs indirect quotes. The cost that actually is under your control is covered in mid-market rate vs the rate your bank gives you. How much a pair is likely to move in the first place follows partly from its category — see major, minor, and exotic currency pairs. For a live reference rate and recent history on a specific pair, try CHF to INR, USD to EUR, or pick any pair from the home page.