The economic calendar in practice

What the economic calendar does and does not do

This lesson explains how to build a personal event filter: a set of criteria that settles, before publication, whether a given event enters your plan at all. It is written for a trader who knows the calendar and knows where to look up dates, but decides at the last minute and has no written rule.

The calendar collects publication dates and times, plus what the market expects. It does not predict the direction of price and does not tell you whether to trade. Its role is narrower and more practical: it lets you establish in advance at which moments volatility and transaction costs rise far enough that an ordinary plan stops being a plan.

The most common mistake is reading the calendar only on the morning of the publication day. By then there is no time left to measure the spread, to calculate the expected range, or to close a position that should not survive the event. You read the calendar while something can still be changed — ideally at the weekend, before the week begins.

Three criteria that decide before publication

The filter rests on three numbers. You calculate all of them from your own instrument’s history, without guessing what the release will say. The thresholds given below are illustrative figures — what matters most is that they are fixed and written into the plan, not set during the publication.

Expected_range  = median (high − low) of sessions with this type of release, last n observations
Buffer          = |entry − SL| / Expected_range
Relative_spread = spread before publication / median spread of sessions without a release

1. How the spread behaves. You are not interested in the spread expressed in points itself, but in its ratio to the typical level. A Relative_spread of 2 means you are paying twice as much as in a normal session. At 3 and above, the transaction cost removes the sense from many setups, so factor spread and slippage into the plan before you make a decision.
2. The expected range against your stop. Expected_range is the median session range for a given type of release, calculated on your instrument. Buffer tells you how many times your distance to the stop fits inside the typical range of the event. When Buffer falls below 1, the stop sits inside a range that the publication can cover within minutes.
3. Liquidity. Check the depth of the order book before and after publication, and how large a share of a given minute’s turnover your order represents. If your position is a significant fraction of turnover, you do not have liquidity, only the hope that someone will take the other side.

The order matters: first you check the spread, then the range, and finally liquidity. Criteria 1 and 2 can be calculated from history, so do that before the week begins. Criterion 3 you check on the publication day, before entering.

A weekly protocol in six steps

1. Set the time zone to match your platform. The economic calendar and the broker platform can show different times, especially around the switch between summer and winter time. The error of assuming that the calendar time is the platform time costs the most when you use leverage.
2. List the high-impact events for the coming week, with the time and the source of the reading. Do not copy numbers from this text or from social media — check the reading and the consensus at the source.
3. For each event, write down the previous reading and the current consensus together with the date you checked them. Without a fresh consensus you have no reference point for the release.
4. Calculate Buffer and Relative_spread on your instrument’s historical data. Record both numbers, even when you reject the event straight away.
5. Assign a rule to each event: A — flat before and after publication, B — entry only once conditions calm down, C — a news playbook with a pre-limited size. You choose the rule by the criteria from the previous section, not by gut feeling.
6. Write the plan into your session playbook together with the daily R limit and the list of positions that will survive the publication. That last point is the most important one: a position left open by accident is not a plan, it is exposure.

After publication you do not chase the first move. Many traders wait for the spread to calm down and for the level to be retested. On CFDs and on the Warsaw exchange, after a gap, that discipline is often worth more than haste. On forex these three criteria enter one layer of analysis together with the levels from the daily chart — the arrangement of layers is described in the lesson Forex analysis.

Worksheet: an event journal

You complete the table at two moments: before publication (columns 1–6) and after it (columns 7–8). The columns are deliberately empty — value comes from what you measure on your own instrument.

You fill in the Relative_spread and Buffer columns before the decision and the I kept to the plan column after the fact. After a dozen rows you will see which events regularly spoil your results and need a stricter rule, and which can be handled more calmly. Without that column you cannot tell discipline from chance.

When the calendar says “do not trade”

Four conditions, each in the form of a rule. Meeting any one of them means no new position — not smaller risk and not a trial entry.

1. Relative_spread is 3 or more. The cost of entering and exiting is then so high that any realistic reward-to-risk ratio calculated from the chart stops applying. If you have an open position with a tight stop, close it before publication.
2. Buffer is less than 1. The stop sits inside the typical range of the event, so holding a position through the publication is no longer a plan, it is a coin toss. Flat before the data, or a consciously recorded gap risk in the plan — there is no third option.
3. The consensus is stale or contradictory across sources. If you cannot point to one number the market expects, you have nothing to compare the release against. Skip the event instead of looking up a consensus after the fact.
4. You intend to hold a position through publication without a written rule. No entry in the playbook means the decision has not been made yet. In that situation, close the position or write the exit condition before publication, not after it.

How to match the event list to your instrument

Not every market reacts to the same releases to the same degree. A short list of watched events works better than a full calendar, because attention is limited.

• Forex and indices react most strongly to inflation readings, central bank decisions and labour market data. You will find a description of sessions and liquidity in the lesson on forex sessions.
• Warsaw-listed stocks additionally require the earnings calendar and current reports from companies on your list. That is a separate mode of work, described in the lesson on company results.
• A portfolio settled in PLN (Polish złoty) feels NBP and ECB decisions even when you trade a foreign instrument. Knowing about the Belka tax alone will not replace awareness of the exchange rate.
• Crypto is sometimes less sensitive to local data and more sensitive to global sentiment and the decisions of the largest central banks. Check that on your own data instead of assuming it.

Start with three events that genuinely move your instrument, and extend the list only when your journal shows that the new positions contribute something. You will find a broader overview of data in the lesson on macro data.

Common mistakes when working with the calendar

• Checking the calendar only on the morning of the publication day. Too late to measure the spread and to close a position that should not survive the event.
• Ignoring the platform’s time zone. The calendar time and the platform time can differ by an hour during the period when one side follows a different time than the other.
• Treating all events as equal. A low-impact reading and a central bank decision are two different volatility scales, and the calendar shows them side by side.
• No entry after the publication. Without comparing your plan with how the session unfolded, you do not know whether your A/B/C rule works at all.
• Chasing the first move. Entering on the first jump after publication is usually the most expensive moment of the day and the worst possible spread.

Measurable exercise

Pick the three nearest high-impact events and run the full protocol on each: calculate Buffer and Relative_spread, record the consensus with its source, and assign rule A, B or C.

After a week, check two things: whether the number of open positions at publication times fell, and whether every row of the worksheet has the I kept to the plan column filled in. Keep the summary in the same spreadsheet as your trading journal, so that the result in R is comparable across weeks.

Check yourself

Common questions

How many minutes before publication should I close positions?

There is no universal number. Set your own rule based on the Buffer value and the spread, write it into the playbook, and check in your journal whether it actually improves the result.

Are company results also a calendar?

Yes. Periodic reports and current announcements are volatility events just like macro readings. If you trade stocks or ETFs, list them alongside macro data.

Where should a news playbook start?

With a small size, a pre-set maximum R and a condition stating that you do not chase the first spike. Without those three elements it is better to stay out of the event.

Why calculate Buffer and Relative_spread when the chart shows it anyway?

Because an assessment "by eye" changes with your emotions, while a number does not. Buffer reduces the distance to your stop and the typical range of the event to a single common measure, so you can compare events and instruments with each other. Relative_spread shows when the transaction cost stops fitting inside the plan. Only from recorded numbers can you work out whether your rejections were accurate or merely cautious — otherwise you have nothing to compare. Match the thresholds to your instrument and confirm them in the instrument description at your broker.