Keeping Records That Survive a Full Year of Premium Selling
The record-keeping problems that only appear after a few months of selling premium, and how to design around them.
Worth reading first: Cash-Secured Puts, From Cash to Assignment
Course contents73 lessons · 28 questions
Tracking options trades is easy for about three months. You sell a put, you write down what you collected, it expires, you write down that you kept it. A spreadsheet handles this perfectly well and there is nothing to learn.
Then something rolls. Then something gets assigned and becomes shares. Then you write calls against those shares and one of them gets called away, and you want to know what the whole sequence earned — and you discover that no row in your spreadsheet knows the answer.
This is the pillar page for that problem. Every lesson in this track deals with one specific version of it.
The four things that break a tracking system
In roughly the order they arrive:
1. Rolls create a chain, not a trade. When you close one leg and open another as a single decision, you now have two rows describing one continuous position. Both show a credit; only one of them is income. This is the roll-recording problem, and it is the most common source of overstated returns in retail options logs.
2. Assignment changes the instrument. A put position becomes a stock position. The premium you collected does not disappear — it reduces the basis of the shares. Most systems have no way to express that, so they record a closed option trade and a separate stock purchase, and the connection is lost.
3. Open positions are not income. The credit is in your account on day one and none of it is earned. Conflating open premium with realised P&L makes every dashboard flattering.
4. Several accounts hide the real exposure. Two brokers, two spreadsheets, and concentration in one name becomes invisible.
What to record, and what is decoration
The instinct is to record everything. In practice most journal templates capture twenty fields and use four, and the fields people skip are frequently the load-bearing ones.
Which fields actually earn their place
Toggle a field off to see what becomes uncomputable. The ones marked Core are the minimum that lets every metric in this course be reconstructed later.
Toggle a field to see what recording it buys you — and what its absence costs.
These ten fields are enough to reconstruct every metric in this course, including blended returns across a rolled chain. Everything beyond them is optional.
Record inputs, compute outputs
A rule that saves a great deal of pain: store what happened, derive what it means.
Record the fill price, the date, the strike, the fees. Do not record “annualised return” as a typed number, because the moment you roll the position that number is wrong and nothing will tell you.
| Stored as a number | Computed from inputs | |
|---|---|---|
| Position rolled twice | Still shows day-one figure | Recomputes across the chain |
| Fee schedule changes | Historic rows stay wrong | Correct once inputs updated |
| You find an error | Every derived cell is suspect | Fix the input, everything follows |
The same position, recorded two ways.
A cadence that actually gets done
At the fill. Record the trade with the actual price, not the mid you aimed for. This is the only step that has to happen immediately, because it is the only information that disappears.
Monthly. Reconcile against the broker. Missed assignments and unlogged fees are cheap to fix in the month they happen and expensive to fix in April.
Quarterly. Review outcomes by strategy and by delta band. This is the point of tracking at all — not compliance, but noticing that your 0.30 delta trades have been assigned four times this year while your 0.16s have not.
Annually. Close the year properly before the 1099-B arrives.
Is a spreadsheet enough?
For a simple book, genuinely yes — and anyone telling you otherwise is selling something. The honest boundaries are covered in where a spreadsheet stops working, but the short version is that it copes fine until you have roll chains, assignment hand-offs and more than one account, and then it needs formulas complex enough that they become their own source of error.
What can go wrong
Recording intentions rather than fills. The mid you wanted is not the price you got, and the difference compounds into a log that flatters you.
Storing computed values. They rot the moment anything changes.
Logging in batches. Reconstructing last month from memory produces confident, wrong records.
Tracking without reviewing. A log nobody reads is a filing system, not a feedback loop.
Key takeaways
- Tracking is trivial until the first roll or assignment, and then it needs actual structure.
- Four things break naive systems: roll chains, assignment hand-offs, open-versus-realised, and multiple accounts.
- Ten fields are enough — including the actual fill price and a link from a replacement leg to the one it replaced.
- Store inputs and compute outputs. A stored 'annualised return' is wrong the moment you roll.
- The point of a log is the quarterly review, not the record itself.
Check your understanding
1. Which single field most improves a manual options log?
2. Why should you record the fill price rather than the mid you aimed for?
3. Why store inputs and compute metrics rather than storing the metrics?
Everything in Tracking & Records
- Cost Basis After Assignment, Worked Through a Whole Cycle— Premium reduces your basis, and the adjustment compounds across cycles. Here is the full arithmetic.
- Recording a Roll Without Counting the Buyback Twice— The most common bookkeeping error in premium selling, and the linkage that prevents it.
- The Fields an Options Journal Actually Needs— Which fields change a future decision, and which just make the spreadsheet look thorough.
- Tracking a Wheel From First Put to Called-Away Shares— One continuous position across puts, shares and calls — and the single number that ties it together.
- Where an Options Spreadsheet Stops Working— An honest account of what a spreadsheet does well, and the four things that reliably break it.
- Reconciling Your Records Against the Broker— A monthly routine that catches the errors which otherwise surface at tax time.
- One Book, Several Brokers— Aggregating several accounts without double-counting capital or hiding concentration.
- Open Premium Is Not Income Yet— The credit hit your account on day one. It is still not yours, and here is the accounting reason why.
- Three Different Numbers All Called Premium— Gross, net and kept are three separate figures, and most tools quietly report whichever flatters.
- The Hand-Off From Option to Shares and Back— Keeping the lineage intact when a position changes form twice in one cycle.
- Closing the Year Cleanly— What to verify in December so that January is filing, not archaeology.
- Reconstructing a Year You Did Not Record— It is recoverable. Here is what broker exports contain, what they omit, and how to fill the gaps.
Related lessons
- Cost Basis After Assignment, Worked Through a Whole Cycle— Premium reduces your basis, and the adjustment compounds across cycles. Here is the full arithmetic.
- Recording a Roll Without Counting the Buyback Twice— The most common bookkeeping error in premium selling, and the linkage that prevents it.
- Where an Options Spreadsheet Stops Working— An honest account of what a spreadsheet does well, and the four things that reliably break it.
- One Book, Several Brokers— Aggregating several accounts without double-counting capital or hiding concentration.