Daily P&L is today's valuation minus yesterday's, so it reports every difference between the two: price moves, but also new, amended and late-entered trades, transfers to realised and changes in how prices were obtained. The work that makes the figure reliable is explaining each day's change and keeping each valuation as it was produced.
Marking to market values every open position at the price the market gives it today, and the daily profit and loss (P&L) is the change in that value since yesterday. That makes the daily figure a comparison rather than a measurement. It is the value of the book at today's cut-off minus its value at yesterday's, plus whatever was realised in between, so it reports every difference between the two valuations: a price that moved, but also a trade entered today, a quantity corrected, a trade agreed yesterday and booked this morning, a cargo that moved from unrealised to realised, a price taken from a different source. The market is one cause among several. A figure can be correct and still unexplained, and the part of the work that decides whether anyone can rely on it is splitting each day's change into its causes and keeping each day's valuation exactly as it was produced.
This article sets out how a daily P&L is actually made: what a mark is for each kind of position, why the daily number is the difference between two valuations, how changes to the book arrive in the figure, what happens when a position prices or settles, how a day's change is explained, and what has to be kept each day so that the explanation can be rebuilt later. It is written for the people who produce the number and answer for it: the risk and middle office who run the valuation, the traders whose positions it values, and the finance teams who receive a figure and are later asked why it moved.
If the subject is new to you, start with A mark values what is still open at today's price, and the result is a gain or loss nobody has taken yet. If you already produce a daily P&L and want the part that goes wrong quietly, start with Daily P&L is today's valuation minus yesterday's, so it reports every difference between the two.
A mark values what is still open at today's price, and the result is a gain or loss nobody has taken yet
Marking to market puts a current price on each open position and compares it with the price in the contract. The difference, multiplied by the quantity still open, is a gain or loss that exists on paper, will be different tomorrow, and has not been paid or received by anyone.
What counts as today's price depends on the position. A futures position is marked at the exchange's settlement price for its contract month. A fixed-price physical purchase or sale is marked at the forward price for its delivery period, taken from the curve the book is valued on. If you bought at a price below today's forward price, the position shows a gain; if you sold below it, a loss. The arithmetic is the same in every case: today's price against the contract price, times what is still open.
A quantity whose price has not yet been fixed is marked differently, and this is easy to miss. A purchase priced on a future average of a published quotation plus an agreed premium has little flat-price value to mark, because until the average is set its price moves with the market on both sides. What it carries is the difference between the premium agreed and the premium the market would set for the same goods today. As each pricing day passes and a slice of the price is fixed, that slice becomes a fixed-price quantity and starts to carry a flat-price mark. The same cargo therefore carries a different kind of mark at each stage of its life, and a valuation that treats it as one number throughout is valuing something it is not.
A mark is also not cash. On an exchange, the daily change in a futures position's value is paid or received as variation margin, so for that leg the mark turns into money every day. A physical forward's mark does not, until the goods are delivered and invoiced. And a mark is not the price at which the position could be closed. It is a reference price, such as a settlement price or the middle of a quoted market, applied to a quantity that might take time and a worse price to exit.
A mark is a price nobody has offered you. What it would take to actually close the position is a separate question, worked through here. Market Risk: Why Your Price Exposure Starts Before the Trade Exists
Daily P&L is today's valuation minus yesterday's, so it reports every difference between the two
A daily P&L figure is not observed. It is computed as the value of the book at today's cut-off, minus its value at yesterday's cut-off, plus what was realised in between. Anything that differs between those two valuations appears in it, whether or not the market moved.
The two valuations can differ in more ways than price. The trades in the book can differ, because some were added and some cancelled. The fields on a trade can differ, because a quantity, a price or a delivery period was corrected. The source of a price can differ, because a point that was quoted yesterday was derived today. The exchange rates can differ. Reference data can differ, such as the dates assigned to a delivery period or the conversion between two units. Each of these changes the value of the book, and each one arrives in the day's P&L in the same column as the market move.
So a change in P&L is a statement that two valuations differ. It is not, by itself, a statement about the market. This is why a figure that is arithmetically right can still be one nobody can explain, and why "the P&L moved" is the start of a question rather than an answer to one.
It also means that yesterday's valuation has to be kept as it was produced. If yesterday is recomputed this morning, using a trade corrected overnight or a price revised after the close, then yesterday's value changes, the correction no longer appears in today's figure, and the number that was reported for yesterday can no longer be reproduced from the system. The difference has not gone away. It has been moved into a day that was already reported, where nobody will look for it.
The prices in yesterday's valuation came from a curve built yesterday. Why that curve cannot be rebuilt from today's is worked through point by point. Forward Curve: Why It Is a Price List, Not a Forecast
New, amended and late-entered trades change the book between valuations, and their effect lands on the day they are entered
Every change made to the book since the last valuation is part of today's P&L. A trade done at a price away from today's close shows a result on its first day. A corrected field shows as a step on an old trade. A trade agreed yesterday but entered today brings two days' effect into one.
A new trade is marked at today's price from the moment it is entered. If it was done at a price different from where the market closed, it shows a gain or loss on its first day. That result is not a market move. It is the difference between the price the trade was done at and the price the book is marked at, and it is worth looking at, because a mistyped price produces exactly the same shape.
An amended trade is revalued with its new fields. A quantity adjusted after a nomination, a price corrected after the confirmation is matched, a delivery period moved: each changes the trade's value, and the change appears today. On a report that shows only the total, it is indistinguishable from a market result on that trade.
A trade entered late is the hardest case. Suppose it was agreed yesterday and booked this morning. Yesterday's valuation did not include it, so yesterday's reported P&L missed its first-day result, and nothing reported its exposure overnight. Today's P&L now contains both its first-day result and yesterday's market move on it. There are two ways to deal with this: restate yesterday, or report the effect today and label it as late. Either can be defended. Which one applies is a policy decision. What cannot be defended is being unable to tell, afterwards, which of the two happened.
Telling them apart needs more than a trade date. It needs the time each trade was entered and each version of its fields, with the time of each change. A trade record that holds only its current state can say what the book looks like now. It cannot say what the book looked like at yesterday's cut-off, and that is the question every explanation of a day's P&L begins with.
As a position prices or settles, its value moves from unrealised to realised, and the total should not move with it
When a slice of a position is priced, a futures position is closed or a cargo is invoiced, the gain or loss on it stops moving with the market. It leaves the unrealised figure and enters the realised one. On that day the total P&L should not change because of the transfer. If it does, the two figures were computed from different numbers.
Unrealised P&L is the mark on what is still open. Realised P&L is the result on what has been closed, priced or settled. A position moves from one to the other in steps: a futures position when it is closed, a priced cargo as each pricing day fixes a slice, a physical trade when the final invoice is issued. On the day of each step, the unrealised figure falls by the value of the slice and the realised figure rises by the same amount. The transfer itself is not a gain.
The total jumps when the two figures come from different places. The unrealised value may be computed in the trading system on the contract quantity and the forward price. The realised value may be taken from an invoice in the finance system, on the bill of lading quantity, with a quality adjustment and an exchange rate fixed on a different day. Each is correct by its own rules. Placed side by side on the day of transfer, they do not agree, and the difference appears in P&L as if it were a result. It is a reconciliation finding that has been reported as performance.
Physical trades add a second, later step. A first invoice can be provisional, issued before the final quantity, the quality analysis or the full pricing is known. When the final invoice follows, the realised figure changes again, on a trade that may have left the open book weeks earlier. That change belongs to the trade, and it arrives in the P&L of the day the final invoice is entered, long after the market move it reflects.
A day's change has to be split into its causes, and what is left over is the figure to look at first
Explaining a day's P&L means removing each known cause in turn. What remains after all of them is the residual. A residual close to nothing says the causes were captured. A residual that is not says something changed that nobody recorded, or that something is wrong.
One way to do it is in a fixed order. Start from the change in value since yesterday. Remove new, amended and late-entered trades. Remove transfers to realised, which should net to nothing in the total. Remove price moves on yesterday's book, by revaluing the positions held at yesterday's cut-off at today's prices. Remove time and currency effects. What is left is the residual. If it is close to nothing, the day is explained. If it is not, find the cause before reporting the figure, because an unexplained residual is either a change nobody captured or an error in one of the two valuations.
The order matters at the edges. A trade amended on a day the price moved has an effect that belongs partly to each cause, and the split depends on which is removed first. Choosing an order and keeping it is enough. Changing it from day to day makes the causes incomparable across days.
The valuation itself is a calculation the trading system performs. The explanation needs more: yesterday's book as it stood, today's book, the list of what changed between them and when, and the prices used on each day. Where the system keeps only the current state of each trade, those inputs are rebuilt from two exports and a comparison, in a workbook, by whoever has to answer the question before the report goes out. The work is then repeated every day, and its quality depends on that person's knowledge of what happened in the book.
The way it degrades is gradual. The residual is given a line called "other" and carried. Once that line exists, it absorbs whatever could not be explained on a busy day, and its size from one day to the next stops being watched. A residual that was a warning becomes a column.
Eight things that move a day's P&L, and which of them is the market
| Source | How it shows on the report | Is it a market result? | Where to confirm it |
|---|---|---|---|
| Price moves on positions held yesterday | A change in proportion to each position and its price move | Yes | The price used for each point on each of the two days |
| A change in how a price was obtained | A move on points nobody traded | Partly. It is a change of source or method as much as of market | Which points changed source or rule between the two days |
| New trades | A result on the first day, from trade price against the close | No. It is the price the trade was done at against the price it is marked at | Trades entered since the last cut-off |
| Amendments and cancellations | A step on an existing trade | No | Each version of the trade's fields, with the time it was changed |
| Late-entered trades | Two days' effect arriving at once | Partly. Yesterday's market move on a trade nobody saw yesterday | Entry time against trade date |
| Transfers to realised | Unrealised down, realised up | No. The total should not change | Both figures from the same quantity, price and rate |
| Final invoice adjustments | A change on a trade already closed | No. Quantity, quality or pricing has been finalised | Provisional invoice against final invoice |
| Exchange rates | A move on positions valued in another currency | Yes, of a different market | The rate used on each of the two days |
Read down the third column. Two of the eight are the market the report is assumed to describe. The others are changes in the book, in its records or in how it is measured, and each needs a different person to confirm it.
A daily valuation has to keep six things before its change can be explained later
| The day's record keeps | Which question it answers | What happens when it is missing |
|---|---|---|
| The positions as they stood at the cut-off, not as they stand now | Which book was valued | Yesterday is recomputed with today's corrections, and the reported figure cannot be reproduced |
| The price used for each point, and where each price came from | Which market the book was valued on | A change of source cannot be told apart from a price move |
| The cut-off time | Which trades and which prices belong to the day | Trades near the cut-off fall into either day |
| Each version of each trade, with the time it was entered | What changed in the book | Amendments look like market results |
| Unrealised and realised values from the same calculation | Whether a transfer was only a transfer | The total jumps when a position settles |
| The explanation of the day's change, with its residual | Whether the figure was understood before it was reported | The question comes back weeks later and is answered from memory |
The first five describe the valuation. The sixth is the record that the valuation was understood, and it is the one that cannot be produced afterwards, because by then the person who knew what happened that day is answering from what they remember.
Questions people ask about this
What does mark-to-market mean in commodity trading?
It means valuing each open position at the price the market gives it today rather than at the price in the contract. A futures position is marked at the exchange settlement price, a fixed-price physical trade at the forward price for its delivery period. The difference between that price and the contract price, multiplied by the open quantity, is the position's unrealised gain or loss. The daily P&L is the change in the total of those values since the previous day, plus anything realised in between.
What is the difference between realised and unrealised P&L?
Unrealised P&L is the gain or loss on positions that are still open, measured at today's prices, and it will change tomorrow. Realised P&L is the result on positions that have been closed, priced or settled, and it no longer moves with the market. A position moves from one to the other when a futures contract is closed, a slice of a priced cargo is fixed or a trade is invoiced. On that day one figure falls and the other rises by the same amount, and the total should be unchanged.
Why does a trade booked today already show a profit or loss?
Because it was done at a price different from the price it is marked at. A trade entered during the day is marked at the close like every other position, so if it was bought below the close it shows a gain and if above, a loss, on its first day. That result is real in the sense that the trade was done at that price. It is not a market move, and a large one is worth checking, because a mistyped price or quantity produces the same result.
Why did our P&L jump on the day a cargo was invoiced?
Because the transfer from unrealised to realised cannot change the total on its own, a jump that day means the value before invoicing and the value after it were computed from different numbers. The trading system may value the open trade on the contract quantity and the forward price, while the invoice uses the delivered quantity, a quality adjustment and an exchange rate fixed on a different day. The transfer itself should not change the total. A jump on the day of invoicing is a difference between two records of the same trade, and it is worth reconciling rather than reporting as a result.
What is a P&L explain, and do we need one every day?
A P&L explain, also called P&L attribution, splits the day's change in value into its causes: market moves on existing positions, new trades, amendments, late entries, realisations, time and exchange rates, and a residual that is left when all of them are removed. Whether it is needed every day depends on how many of those causes occur on a typical day. A book that changes daily needs a daily explain, because an unexplained change is much harder to trace once several days have been added on top of it.
How should we treat a trade that was entered a day late?
Either restate the previous day's P&L to include it, or report its effect in today's P&L and label it as a late entry. Both approaches are used, and the choice is a policy for finance and risk to agree. What matters more is that the record shows which one was applied, and that the trade carries both its trade date and the time it was entered, so that the late effect can be separated from today's market move on the same trade.
Is mark-to-market profit real money?
Not until the position is closed, priced or settled. An unrealised gain is the value the position would have at today's price, and it changes every day. For futures, the daily change is paid or received as margin, so cash moves even though the trade is still open. For physical forwards, nothing is paid until delivery and invoicing, and the final amount can differ from the last mark because of quantity, quality and pricing adjustments.
Can we produce the daily P&L in a spreadsheet if the trades are in the trading system?
The valuation can be computed in the system and the explanation assembled in a spreadsheet from two exports. The calculation is not where this fails. What fails is that the spreadsheet holds the knowledge of what changed between the two days, which trades were amended, which were late, which prices changed source, and that knowledge is rebuilt by hand each day. The point at which this becomes a problem is the first time someone asks why a figure from several weeks ago moved, and the sheet has been overwritten since.
Where this lands in a trading system
Nothing above is an argument for buying software. It is an argument about what a daily P&L is made of and where each of its parts is recorded. Some of that maps onto system capabilities, and it is worth being exact about which parts.
The first conclusion is that a day's P&L is only explainable if physical trades, the derivatives that hedge them and the results of settling them are valued from the same records. That is the job of a CTRM/ETRM system providing complete trading lifecycle management for commodity traders, integrating physical trade, financial hedging, risk control, and settlement in one platform; Time Dynamics' Fusion is one, and the parts of it that bear on this article are complete physical trade lifecycle management and execution, advanced trading management for futures and derivatives markets, comprehensive risk assessment and monitoring capabilities and automated financial settlement and payment processing. Holding all four in one place is the condition under which the unrealised value of a trade and its realised value at settlement can be drawn from the same quantity and price, rather than from two systems that are each right by their own rules.
The second conclusion is that, wherever the system does not keep the book's history, the explanation is assembled outside it, from exports and a comparison kept by whoever answers for the number. That is what X-Ray addresses: a non-invasive data processing and analysis platform whose collection toolkit XDK performs non-invasive automated data collection from databases, Excel files, and web interfaces, and which is built to collect data without disrupting existing systems or workflows. The last property matters here, because the person running the explanation is the one who knows what happened in the book each day, and the first step is to read their record rather than replace it.
The third conclusion is that the explanation has to be seen by cause and by day, so that a growing residual is noticed while the day is still fresh. One-click generation of personalized reports with Excel-like interface, X-Sheet, is that kind of work.
One conclusion of this article is deliberately not mapped. Whether a late trade restates yesterday or is reported today, which prices the book is marked on, and in what order the causes are removed are decisions for risk and finance. They do not correspond to anything on the published capability list, and the honest thing to do is leave them unmapped.
And here is where the rest stops. None of this decides how a position is marked or which curve it is marked on. It does not explain a residual by itself, and it does not make a mark into a price anyone will pay. What a system can do here is smaller and easier to check than the problem in this article: whether each day's valuation is kept as it was produced, whether each change to a trade carries the time it was made, and whether the realised and unrealised figures come from the same numbers. That covers a real part of the gap. It leaves the judgement where it was.
A note on the evidence in this article
There are no statistics in this article, and that is a decision rather than an oversight. The observations behind it, about how daily P&L is produced and how often its change goes unexplained, come from material we are not in a position to publish as a citable figure. A number nobody could check would make the argument look stronger and make it worth less.
So the argument is made from the mechanism, and each step can be tested against your own book. Take three days from the last month on which the P&L moved most. For each, try to split the change into the eight sources in the table above using only what is stored today. Then pick one reported day and try to reproduce its figure exactly from the system as it stands now. Finally, find the day a large cargo was invoiced and check whether the total moved. The causes you could not separate, the figure you could not reproduce and the jump you could not account for are the finding. If your book behaves differently from what is described here, your book is the better evidence.
Practice also varies by company and by market. Cut-off times, marking conventions, the treatment of late entries and the point at which a physical trade counts as realised are set by your own policies, your auditors and the markets you trade in. Where this article describes a mechanism, it describes its shape, not the rules of any particular company or framework.
The P&L tells you that the book changed. The explanation tells you how.