Partial Fills and Unfilled Orders: Review the Whole Execution
An order for 100 shares can become a position of 40 shares, several executions at different prices, or no position at all. A journal that records only the final entry and exit can lose that sequence. Reviewing the whole order helps explain what exposure you actually held and whether your instructions matched your intention. This article offers educational recordkeeping suggestions, not personalised investment advice. The regulatory sources concern U.S. stock orders; other instruments and jurisdictions can work differently. Trading involves risk, and better documentation does not guarantee better returns.
1. Keep the order separate from its fills
An order records an instruction; a fill records an execution. Preserve both when reviewing a trade. Suggested order fields include submission time, direction, requested quantity, order type, limit price where applicable, and duration. Alongside them, retain each executed quantity, price and time. Use an order identifier to connect the records rather than relying on memory.
This distinction matters because one order can execute at several prices. Investor.gov explains that parts of a large market order may fill at different prices when sufficient liquidity is unavailable at one price. It also explains that a market order's execution price is not guaranteed. A single entry price can therefore conceal the sequence behind the position. [1](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins-14)
Choose a consistent way to group fills into a position for performance reporting, while retaining the underlying detail. Three fills from one entry instruction need not become three independent strategy observations. Otherwise, execution fragmentation can inflate your apparent sample size without adding three separate trading decisions.
2. Reconstruct the position you actually held
Hypothetical arithmetic: suppose a buy order for 100 shares receives 40 shares at $25.00 and 20 at $25.05, with the remaining 40 later cancelled. The executed quantity is 60 shares. Its weighted average entry is ((40 × $25.00) + (20 × $25.05)) ÷ 60, or approximately $25.0167 before charges. Averaging the two prices without weighting them would give a different answer.
If all 60 shares subsequently sell at $25.20, the hypothetical gross gain is $11: sale proceeds of $1,512 minus purchase cost of $1,501. The unfilled 40 shares contribute nothing to realised trading profit or loss. Any estimate of what they might have earned belongs in a clearly labelled hypothetical analysis.
For risk review, compare the intended quantity with the quantity held at each stage. A partial entry followed by an additional fill changes exposure over time. Suggested notes should also identify how exit instructions were checked against the actual position. A final flat position does not explain whether quantities matched during the trade.
3. Record the final status of the remainder
After a partial fill, give the unexecuted balance an explicit status: still working, cancelled, expired or otherwise unresolved in your records. Retain the broker's status and timestamps where available. Treat a cancellation request and a confirmed cancellation as separate events in your notes; an intention to cancel is not execution evidence.
Order duration and instructions affect this history. Investor.gov explains that an immediate-or-cancel instruction cancels any portion that cannot execute immediately, while a fill-or-kill instruction permits no partial execution. These are descriptions of order mechanics, not suggestions to choose either instruction. Available instructions and broker policies differ. [1](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins-14)
At reconciliation, compare requested quantity, cumulative filled quantity and remaining quantity. Investigate unexplained differences using broker records. If your journal cannot represent working orders, a separate order note or spreadsheet can preserve the history. Label unresolved information as unknown instead of filling the gap with an assumed sequence.
4. Choose an execution benchmark before judging the fill
Calling a fill 'bad' requires a defined comparison. Suggested benchmarks include the intended entry recorded before submission or a timestamped quote captured when the order was sent. Name the benchmark and its source. A later chart price answers a different question and should not silently replace the original reference.
The SEC explains that execution takes time, prices can change, and displayed quotes cover a specific quantity. Consequently, the price seen on screen may differ from the eventual execution price. A screenshot can document what you observed, but it does not prove that your entire order could have executed there. [2](https://www.sec.gov/reportspubs/investor-publications/investorpubstradexechtm.html)
Hypothetical arithmetic: against a recorded buy benchmark of $25.00, the 60-share purchase above cost $1 more than buying those same shares at the benchmark. Record that as an adverse benchmark difference. It is already reflected in the actual purchase cost; subtracting it again from realised profit would double count it. Explicit commissions and other charges require separate reconciliation, and journal software should not be assumed to adjust them automatically.
5. Give unfilled orders their own review
An unfilled order can reveal a process question even though it creates no executed position. Suggested fields include the intended setup, submission time, order terms, final status and any amendments. Investor.gov states that a limit order specifies its execution price or better but is not guaranteed to execute. An unfilled limit order alone therefore does not establish a broker error. [1](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins-14)
Keep these records separate from completed-trade outcomes. Do not add an unfilled order as a zero-profit trade merely to include it somewhere: doing so changes averages and trade counts. Instead, report submitted orders and their execution statuses alongside, but separately from, the results of positions actually opened.
Avoid reviewing only missed moves that later looked attractive. A suggested review should include every submitted order within a defined period, including those whose non-execution preceded an adverse move. If you estimate hypothetical outcomes, specify entry and exit assumptions and acknowledge that neither execution was observed. A favourable later chart cannot supply the missing fills.
6. Turn execution observations into a narrow process check
Summarise the review with counts of fully filled, partly filled and unfilled orders, plus unresolved records. You might calculate executed quantity divided by requested quantity, but state the denominator and order types included. A partly filled order followed by cancellation carries a different history from one left working; one aggregate percentage can hide that difference.
Compare reasonably similar observations and show the sample size. Differences between order types may reflect different instruments, sizes or trading conditions rather than the instruction itself. Measured historical execution results describe that recorded sample. They do not establish future execution quality or future strategy expectancy, and hypothetical missed-trade results should remain visibly separate.
One suggested next step is a documentation check: for the next review period, require every submitted order to have a final status and every partial fill to reconcile to the position record. Define completion by whether the records reconcile, rather than whether profits improve. FreeTradeLedger can hold your manually entered trade records and screenshots; retain supporting broker documents for the execution detail your review needs.