How to Journal Scaled and Partial-Fill Trades

You entered a position in three clips and exited in two. Your broker's order history shows five separate fills at five different prices. Your journal has one row with one entry field and one exit field. Something has to give, and usually what gives is the accuracy of the number you type in.

This is the scaling problem. A trading journal built for one entry and one exit quietly falls apart the moment you scale in and out of a position. If you record the wrong average, every downstream metric inherits the error: your P&L, your R-multiple, your win rate on that setup, your whole read on whether the trade was good. This post covers how to journal scaling in and out cleanly, so the blended averages stay honest and your trade history stays queryable.

TL;DR

• A scaled trade is still one trade. Log the individual fills (executions), then let the trade roll up to one blended average.
• The correct blend is a volume-weighted average, not a simple average of the prices. Weight each fill by its size.
• Compute a separate volume-weighted average for the entry side and the exit side. Realized P&L is the difference between them, times the closed quantity.
• Capture each execution's price, quantity, time, and fees at the moment of the fill. You cannot reconstruct those later without pulling broker records.
• The blended average is a summary, not the story. Keep the fills so you can see whether you added into strength or averaged down into a loser.

Why one-entry-one-exit logging breaks

A basic journal assumes a clean shape: one price in, one price out, one position size. That shape holds for a single-clip trade. It stops holding the instant you scale.

When you scale in, you take a partial position, then add layers as the trade develops. When you scale out, you close the position in pieces to bank profit or cut risk gradually. Both are standard risk-management techniques, and Warrior Trading describes them as core tools for managing exposure rather than committing to a full size at a single price.

The catch is on the record-keeping side. Your broker does not hand you a tidy summary. As one breakdown of the problem puts it, your order history gets "flooded with Partial Fills and Fractional Executions," and to log the trade accurately you have to calculate the volume-weighted blended average price for both your entry and your exit (Pics Totally). A single-row journal has nowhere to put four entry fills, so traders do one of two things: they eyeball an average, or they log only the first fill. Both corrupt the data.

Before you decide what to log, it helps to settle what to log in a trading journal at all, because a scaled trade needs the same fields as any other, just recorded per execution instead of once.

The core model: executions roll up to one trade

The clean mental model has two layers.

The execution is a single fill: a price, a quantity, a timestamp, and its fees. Every clip you enter and every clip you exit is one execution. A trade you scaled into in three parts and out of in two has five executions.

The trade is the whole position, from first entry to full close. It carries the blended entry average, the blended exit average, the total P&L, your thesis, your tags, and your setup. This is the unit you review and the unit your analytics count.

Keeping both layers is what makes a scaled trade honest. The blended average tells you the result; the executions tell you the shape of how you got there. You need both to review the decision, not just the outcome.

How to calculate the blended average correctly

The one mistake that quietly poisons scaled trades is averaging the prices instead of weighting them by size. If you buy 100 shares at $50 and 50 shares at $52, the naive answer is $51. The correct answer is not.

The right method is a volume-weighted average. The Motley Fool lays out the steps plainly: list the price and share count for each transaction, multiply each price by its share count, add those products together, then divide by the total shares. In their worked example, four buys of different sizes totaling $9,825 across 100 shares produce a weighted average of $98.25 per share, which is not the simple average of the four prices.

Applied to the two-clip entry above:

• Fill 1: 100 shares times $50.00 = $5,000
• Fill 2: 50 shares times $52.00 = $2,600
• Total cost = $7,600, total shares = 150
• Blended entry = $7,600 / 150 = $50.67

The simple average would have said $51.00. That is 33 cents too high, and on a 150-share position it invents $50 of P&L that never happened. Multiply that error across a year of scaled trades and your journal stops describing your trading.

The same formula runs on the exit side. Do the volume-weighted blend for your sells, and realized P&L is the difference between the blended exit and blended entry, multiplied by the quantity actually closed. This weighted-average approach is exactly how brokers compute a position's cost basis. Blofin's execution primer notes that when you receive a partial fill, "the cost basis for your position is calculated as the weighted average price of all executed portions" rather than tracking each piece separately (Blofin Academy).

Scaling out: partial exits and what P&L to book

Scaling out adds a second wrinkle. When you close part of a position, you realize P&L on the closed shares while the rest stays open. The question is which cost basis the closed shares carry.

For a position built at one blended average, the answer is simple: the closed shares carry the blended entry average, and the remaining shares keep it too. Alpaca's documentation shows this with a worked example under the weighted-average method. After buying 100 at $10, 50 at $12, and 30 at $15 (total cost $2,050 across 180 shares, a blended average of about $11.39), selling 120 shares deducts 120 times the blended average from the cost basis. The remaining 60 shares keep the same $11.39 average entry. Your realized P&L on the scale-out is the sell price minus $11.39, on those 120 shares.

Two practical rules follow from this:

• Do not reset your average when you scale out. Selling shares does not change what you paid for the ones you still hold. A journal that recalculates entry on every exit will drift.
• Book realized P&L only on closed quantity. The open remainder stays unrealized until you close it. If your journal mixes the two, your realized numbers stop matching your broker.

There is a separate, stricter version of this question for taxes. The IRS lets you elect an average-basis method for mutual fund shares, where you "add up the cost of all the shares," divide by the total number of shares to get an average basis per share, then multiply by the shares sold (IRS Publication 550). But average basis is only one allowed method, and for individual stocks the default is often FIFO or specific identification. Your journal's job is to track your trading performance accurately; your broker's 1099-B is the record that governs your taxes, and the two can legitimately use different accounting. Do not treat your journal's blended average as tax advice.

The fields you must capture per execution

The single hardest part of journaling scaled trades is that the fill-level data is perishable. Price, quantity, time, and fees exist cleanly at the moment of the fill and get muddy afterward. Capture them then, not later.

For each execution, record:

• Side: buy or sell (or the short-side equivalent).
• Price: the actual fill price, not your limit order price.
• Quantity: the shares or contracts in that specific fill.
• Timestamp: when it filled. This is what lets you see the sequence, whether you added into a move or knife-caught a drop.
• Fees and commissions: per fill. Scaling multiplies your fill count, and each fill can carry its own commission, so fees on a five-clip trade are not trivial.

From those five fields, the blended averages and P&L compute themselves. What you should never do is store only the computed average and throw away the fills. The average is a summary; the fills are the evidence. If you tag this trade "added into strength" versus "averaged down," the timestamps and prices are what verify the label. A clean trading journal tagging system depends on that underlying fill data being intact, because the tag is only as trustworthy as the record it describes.

Reading scaled trades back on review

The reason to keep all this structure is that scaled trades hide two very different behaviors under one P&L number, and only the fill sequence tells them apart.

Adding to a winner as it confirms is disciplined pyramiding. Adding to a loser to lower your average is averaging down, and it is often a rule-break dressed up as conviction. Both show up in your journal as "scaled in." Both might even be green. But they are not the same trade, and if your journal collapsed them into a single entry price you would never see the difference.

This is where scaling connects to how you rate your own decisions. If you log a confidence score on each entry clip, you can check whether your later adds were higher-conviction or just hope. A profitable average-down that you would not repeat is a lucky loss of process, and the fill-level record is the only thing that catches it. On review, run three checks on every scaled trade:

• Did each add improve the trade's expected value, or did it just increase risk?
• Was the scale-out planned, or a panic trim after the position moved against you?
• Did fees on the extra fills meaningfully eat the edge?

Frequently Asked Questions

Should I log each fill separately or just the final average?

Log each fill, and let the average compute from them. If you store only the average, you lose the timing and sequence that let you tell disciplined adds from averaging down, and you cannot audit the number later. Capture the executions; treat the blended average as a derived summary.

Do I use a simple average or a weighted average for scaled entries?

Always weighted. A simple average of the fill prices is only correct when every fill is the same size, which is rarely true. Weight each fill price by its quantity, sum the products, and divide by total quantity, as the volume-weighted method requires.

Is a scaled trade one entry in my journal or several?

One. A position you built and closed in multiple clips is a single trade with multiple executions underneath it. Counting it as several trades inflates your trade count and distorts win rate and every per-trade metric.

How do I handle P&L when I only scaled out of part of the position?

Realize P&L only on the quantity you actually closed, using the blended entry average as the cost basis. Keep the remaining shares at that same average as an open position. Do not reset the average when you sell part of the position.

Does scaling change my cost basis for taxes?

Possibly, but your journal is not the authority on that. Brokers report tax basis using a specific accounting method (often FIFO or specific identification for stocks, or an elected average-basis method in some cases). Use your broker's 1099-B for taxes and your journal for measuring your trading. They can differ legitimately.

Final Thoughts

Scaling is a legitimate way to manage risk, and it is not going away just because a spreadsheet finds it inconvenient. The fix is to stop forcing a multi-fill trade into a single-fill shape. Record the executions, let the volume-weighted averages compute themselves, book realized P&L only on closed quantity, and keep the fill-level detail so review can see the shape of the decision and not just its result.

Do that consistently and a scaled trade stops being the entry that breaks your journal. It becomes one of the more honest records you keep, because it shows exactly how the position was built and unwound. TradeReveal models trades this way by design: each trade holds its individual executions, and the blended average and P&L roll up from them, so scaling in and out does not force you to hand-average anything.

Sources

• The Motley Fool, "How to Calculate Average Trade Price": fool.com
• Alpaca Docs, "Position Average Entry Price Calculation": docs.alpaca.markets
• IRS Publication 550, "Investment Income and Expenses" (Average Basis): irs.gov/publications/p550
• Blofin Academy, "Partial Fills Explained: Average Fill Price, Liquidity, and Latency Basics": blofin.com
• Warrior Trading, "Beginners Guide to Scaling In and Out of Trading Positions": warriortrading.com

Start your free TradeReveal account today

Happy Trading,

The TradeReveal Team