Forex seasonality: month-end flows, December liquidity and why seasonal edges are weak

Forex seasonality refers to patterns in currency markets that repeat with the calendar: flows at month-end and quarter-end, thin liquidity around Christmas and New Year, quieter summers, fiscal year-ends and holiday closures. Some of these patterns are real and well understood, but they mostly affect liquidity, spreads and volatility rather than the direction of prices. Claims that a currency pair reliably rises or falls in a given month rest on small samples and are easy to find by chance, so seasonal direction is a weak basis for trading decisions.

The quarterly calendar guides (Q1, Q2, Q3, Q4) list the dated events for 2027, including holidays and daylight-saving changes. This guide steps back and asks which calendar effects are genuine, why they happen, and what a sceptical trader can reasonably do with them.

Two kinds of seasonality

It helps to separate two claims that often get blurred together.

• Seasonality in conditions. Liquidity, spreads, volatility and the risk of gaps change predictably with the time of day, the day of the week and the time of year. This is well established and has clear causes: market participants go on holiday, exchanges close, banks shut their books.
• Seasonality in direction. The claim that a currency tends to rise or fall at a particular time of year. Some mechanisms could in principle produce this, but the evidence is usually thin and the effects, where they exist, are small, inconsistent and often traded away once they are widely known.

The first kind is useful to every trader as a risk tool. The second kind deserves a lot of scepticism.

Month-end flows and the London fix

What the fix is

Every business day, benchmark exchange rates are calculated around 16:00 London time. The best known are the WM/Reuters rates (now WM/Refinitiv, part of LSEG), commonly called "the 4pm fix" or "the London fix". Fund managers, index providers and companies use these rates to value portfolios and settle transactions, so a lot of currency business is deliberately executed at the fix.

Since February 2015 the benchmark has been calculated from trades and quotes over a five-minute window around 16:00, rather than the one-minute window used before. The change followed investigations into manipulation of the fix, which led regulators in the UK and US to fine several large banks in 2014. You do not need the history, but it explains why the fix is closely watched and why trading activity clusters around it.

Why month-end is different

On the last business day of the month, portfolio managers rebalance and adjust currency hedges to bring holdings back in line with targets, and many do so at the fix. That concentration of orders can move currencies in the final hour before 16:00 London for reasons that have nothing to do with economic news.

Worked example: a hedge rebalance

Illustrative numbers. A euro-based fund owns 1 billion US dollars of US shares and hedges all of its dollar exposure, so it is short 1 billion dollars against the euro through currency forwards.

• During the month, US shares rise 5%. The holding is now worth 1.05 billion dollars.
• The hedge still covers only 1 billion dollars, so 50 million dollars of the portfolio is unhedged.
• To restore a full hedge at month-end, the fund sells an additional 50 million dollars against the euro, often at the 4pm fix.

Multiply that by many funds and the logic suggests a tendency for dollar selling at month-end after months when US shares have outperformed, and dollar buying after months when they have underperformed. Banks publish month-end flow estimates based on exactly this kind of reasoning.

Three reasons not to treat this as a reliable trade:

1. Many other flows offset it. Bond portfolios, corporate payments, other investors with different hedge ratios, and speculative positions all hit the market at the same time.
2. It is well known. When a flow is predictable, dealers and fast traders position for it beforehand, which can move the price earlier in the day or even into the previous days, leaving little or nothing at the fix itself.
3. The size is uncertain. Nobody outside the funds knows the real hedge ratios. Estimates are frequently wrong in size and sometimes in sign.

The practical takeaway is about conditions: around 16:00 London on the last business day of the month, expect more activity and sometimes sharp, short-lived moves in the major dollar pairs. If you have a tight stop near the price at that time, be aware that it can be hit by flows unrelated to your trade idea.

Quarter-end and year-end

Quarter-ends and especially year-end add another layer. Banks report their balance sheets on those dates, and some reduce activity or adjust funding positions around them. Demand for short-term dollar funding has at times risen around year-end, which can show up in currency swap and forward pricing. Year-end also combines with the thinnest liquidity of the year, which is the next topic.

December and early January

From roughly the middle of December until the first full week of January, the forex market is noticeably thinner. Many institutional traders take holidays, banks wind down risk-taking before year-end, and Christmas Day and New Year's Day are closures for most of the interbank market. The Q4 calendar lists the 2027 holiday dates.

What thin liquidity means in practice:

• Wider spreads. Fewer liquidity providers quoting means wider prices, particularly outside the main session hours. Why spreads widen covers the mechanics.
• Bigger moves on small flows. A trade that would barely register in March can push the price noticeably in late December.
• Sudden air pockets. Some of the sharpest moves in recent forex history have happened in thin holiday markets. On 3 January 2019, during a Japanese holiday and in the early hours of Asian trading, the yen jumped sharply against several currencies within minutes before partly recovering. The Reserve Bank of Australia, among others, pointed to the lack of holiday liquidity and to automatic orders triggered as the move gathered pace; some news was cited too, but the size and speed of the move were out of proportion to it.

None of this tells you the direction of a move. It tells you that the range of possible outcomes is wider, and that stops are more likely to be filled with slippage. The gap risk guide explains why a stop is not a firm exit price in those conditions.

Other calendar effects worth knowing

Summer

August is typically quieter in European markets, as many traders take holidays. Thinner liquidity can mean either listless ranges or outsized reactions to unexpected news; the quiet does not make the market safer. The Q3 calendar discusses the late-August Jackson Hole symposium, which has sometimes broken the summer calm.

Japanese fiscal year-end and Golden Week

Japan's fiscal year ends on 31 March, and market commentary often discusses repatriation by Japanese companies and investors in the weeks before. The evidence that this produces a consistent yen move is mixed, and in some years the effect has been overwhelmed by everything else. Japan's Golden Week holidays in late April and early May thin out Tokyo trading. The yen explainer covers what drives the currency more reliably.

Chinese New Year

The Lunar New Year holiday reduces activity in Asian markets for several days, which can affect liquidity in Asian session trading and in currencies tied to Chinese trade.

Daylight-saving gaps

Twice a year, US and European clocks change on different weekends, so for a few weeks the time difference between New York and London is four hours rather than five. Every US release and the New York close shift by an hour for Europeans. It is a purely mechanical effect, but it catches traders out every year. The market hours and sessions guide and the market hours tool help.

Intraday and weekly rhythms

The same principle applies on shorter cycles. Spreads are typically widest around the daily rollover at 17:00 New York time and on Monday mornings in Asia; volatility tends to peak in the London and New York overlap. These are seasonality too, just on a daily and weekly calendar, and they are far more reliable than any monthly price pattern.

Why seasonal direction is weak evidence

Seasonal charts are popular because they look convincing: a bar chart of average returns by month, with one month standing out. Here is why they deserve scepticism.

Small samples

Each month appears once a year. Ten years of data gives you ten Decembers. That is a tiny sample for something as noisy as currency returns.

Worked example: the coin-flip test

Suppose a seasonal table says a pair rose in 7 of the last 10 Decembers. How surprising is that if December has no real effect at all, so each year is a 50/50 coin flip?

• The probability of 7 or more "up" years out of 10 by pure chance is 176 out of 1,024, about 17%.
• The probability of a pattern at least that strong in either direction (7 or more ups, or 7 or more downs) is about double, around 34%.

So roughly one in three months will look like a "seasonal tendency" on ten years of data even if no such tendency exists.

Many pairs, many months

Now consider how these tables are produced. Someone looks at, say, 7 pairs across 12 months: 84 combinations. If none of them has any real seasonal effect, you would still expect around 14 combinations to show 7 or more up years out of 10 by chance alone (84 × 17%), and roughly as many again showing 7 or more down years. The patterns that get published are the ones that stood out, which is exactly what chance would produce. Statisticians call this the multiple comparisons problem; traders call it data mining.

Averages hide the distribution

An average December return can be dominated by one or two large years. A pair that rose 6% one December and fell slightly in five others can show a positive "average December". Look at the median and at how many years were positive before believing an average.

Regimes change

Even where a seasonal mechanism exists, its effect depends on the environment: interest rate levels, which investors hold which assets, how much hedging is done, and how many traders are already positioned for the pattern. A tendency that held for a decade can disappear once it is widely published.

Costs and timing

A small average seasonal return can vanish after the spread, the swap on a month-long position, and the slippage that comes with trading in thin holiday markets. And "December" is not a trade: you still need an entry, a stop, a position size and an exit.

Using seasonality sensibly

If you want to look at seasonal ideas, test them properly before using them. The backtesting guide explains the process; for seasonal ideas in particular:

1. State the mechanism first. If you cannot say why a flow should happen, the pattern is more likely to be noise.
2. Use as much data as you can, and split it: find the pattern in one period and check whether it holds in a later one you did not look at.
3. Count the comparisons. If you scanned 84 combinations to find one, judge it against what chance would produce from 84.
4. Include realistic costs, including holiday spreads and swaps.
5. Record it. Track seasonal ideas in your trading journal separately from your main approach, so you can see whether they add anything.

The more reliable use of seasonality is defensive. Build a risk calendar that marks month-end afternoons, quarter-ends, the second half of December, holiday closures and daylight-saving gaps, and decide in advance how you will adjust: smaller positions, wider stops with smaller size, or no new trades. The position size calculator helps keep the money at risk constant when you widen a stop.

Leveraged CFD trading carries a high risk of loss, and most retail accounts lose money; thin markets make losses larger and faster. See the risk warning.

Common mistakes

• Treating a seasonal chart as a forecast. It describes the past in small samples and says nothing reliable about this year.
• Ignoring the multiple comparisons problem. A pattern found by scanning many pairs and months is expected by chance.
• Confusing conditions with direction. Thin December liquidity is reliable; a December rally in a given pair is not.
• Holding tight stops through the month-end fix. Flows around 16:00 London can trigger them without any news.
• Assuming a quiet market is a safe market. Holiday and summer markets can produce some of the sharpest moves of the year.
• Forgetting costs. Wider holiday spreads and month-long swap charges can consume a small seasonal edge entirely.
• Missing the clock change. The daylight-saving gaps shift every US event by an hour for Europe.

Frequently asked questions

Is there seasonality in forex?

Yes in trading conditions: liquidity, spreads and volatility follow the calendar, with thinner markets around Christmas and New Year, in August and on holidays, and concentrated flows at month-end. Seasonality in price direction is much less reliable; the patterns usually rest on small samples and tend to be unstable.

What happens in forex at month-end?

Portfolio managers rebalance holdings and currency hedges, often at the 16:00 London benchmark fix on the last business day of the month. That can create bursts of activity and short, sharp moves in major pairs in the final hour before the fix. The direction is hard to predict because many flows offset each other and well-known flows are often anticipated.

Why is forex liquidity low in December?

Many institutional traders are on holiday, banks reduce risk-taking ahead of year-end reporting, and the main holidays close large parts of the market. With fewer participants, spreads widen and smaller orders can move prices further than usual.

Is it a good idea to trade between Christmas and New Year?

That depends on your strategy, but conditions are generally less favourable: spreads are wider, liquidity is thin and sudden moves are more likely. Many traders reduce position size or stop trading over the period, and some close positions to avoid holding them through holiday gaps.

Do seasonal patterns in currency pairs really work?

Some calendar flows have a plausible mechanism, but published seasonal tendencies in price direction are weak evidence. With only one observation per year per month, and many pairs and months to choose from, patterns of 7 out of 10 years appear frequently by chance. Any seasonal idea should be tested on data you did not use to find it, after costs.

What is the London 4pm fix?

A set of benchmark exchange rates calculated each business day from trading over a five-minute window around 16:00 London time. Fund managers and companies use it to value portfolios and settle transactions, so trading activity often concentrates around it, especially at month-end.